Use of Authorial Stance Markers in Different Rhetorical Moves of Dissertation Abstract by International Students in China: A Corpus-Based Comparative Study

: The abstract is a critical component of an academic genre as it enables readers with an overview of the contents. Based on two self-constructed corpora, this study examined the use of rhetorical move structure and stance markers in dissertation abstracts by international students in China and native Chinese speakers through a comparative analysis. The results show that the majority of dissertation abstracts by international students in China contain three essential moves (M1, M2, M3), while the majority of native Chinese speakers' abstracts included four essential moves (M1, M2, M3, M4). International students in China spend more space on "Background" Move while native Chinese speakers focus more on "Results" Move. Overall, the distribution of stance markers between the two corpora is quite similar, with hedges, self-mentions, boosters, and attitude markers appearing in descending order. However, the preference in the usage of stance markers varies from move to move in this study based on the different functions of each move. The findings of this study suggest that the importance of rhetorical move structures and stance markers in academic writing proficiency for international students in China, and it is necessary to provide appropriate teaching and guidance to help CSL students improve their academic writing proficiency.


Introduction
With the internationalization of higher education, there is an increasing influx of foreign students into China. In order to better participate in academic activities, international students attempt to express their academic ideas through Chinese, expecting to be accepted by a wider range of scholars. Consequently, the academic language proficiency of international students in China is of great importance for their scientific research work. Nonetheless, Li Lijun & Guo Qi [20] point out that studies targeting the linguistic features of Chinese academic papers have not received much attention from academics.
The academic paper is a complex genre of writing [29]. The traditional view is that academic papers should be objective, scientific, and theoretical, and intend to convey objective information to the reader. As such, when composing an academic paper, in addition to having accurate, understandable, and concise language expression, the author should also avoid using emotionally charged language [25] to prevent influencing the reader's comprehension and judgment of the paper. However, as research progresses, scholars have found that in academic discourse, authors also explicitly or implicitly express their opinions and attitudes through evaluative language materials [4,[22][23][24]27] and persuasive discourse to promote their academic results to readers [16,30]. These linguistic strategies, which reflect the author's attitude, personal feelings and value judgments about the pro, are referred to as "stance" [3]. Hyland (2005a: 178) points stance is a type of comment that can demonstrate an author's reliability of a proposition, reflecting the author's commitment to or attitude toward the proposition. Thus, an academic paper is not merely a transmission of objective facts as traditionally thought [31], but also a process in which International Students in China: A Corpus-Based Comparative Study the author engages and utilizes stance to express his or her attitude and opinion.
Conrad & Biber [5] proposed a framework for stance analysis from the perspective of semantic types, referring to linguistic resources that reflect the author's stance as "stance markers" and classifying them into three semantic types: epistemic, attitudinal, and style-of-speaking. From a perspective of functional linguistics, Hyland (2005a: 177) focuses on the author-reader interaction and classifies stance markers into: hedges, boosters, attitude markers and self-mentions. From a discourse-interaction perspective, Du Bois [9] views stance expressions as verbal interaction processes, examining the linguistic forms used by communicators in the course of performing social acts. The results of previous studies show that comparative studies of stance have received increasing attention [7,16,17,21,26], with a particular focus on how second language learners accurately express the author's stance in academic writing. However, there is a dearth of research on stance expression in Chinese corpus [11], and domestic studies largely compare English corpus of Chinese English learners with native speakers [34,37]. Xu Jingning [35] conducted a preliminary investigation of "I think" in spoken Chinese, pointing out the characteristics of its stance in semantics and discourse. Zheng Youdi and Luo Yaohua [39] investigated the subjective evaluation functions of "this" and "that" in spoken Chinese by applying the stance triangle method, and Le Yao [19] examined the expression of stance in different conversational sequence structures by the concession-like homonymy. However, all the above studies were limited to spoken Chinese and ignored the study of written discourse.
As a special genre of academic discourse, abstracts are used to concisely summarize the main content of the literature and facilitate inter-authorial and reader interactions to achieve scholarly consensus [14]. Abstracts have an independent discourse structure, in which the components with explicit rhetorical functions are known as "move" [12]. Therefore, the selection of stance markers is contingent on the communicative purpose of the moves, indicating the author's attempt to gain the reader's recognition [1,29]. Conversely, without a move-by-move analysis of the linguistic features of abstracts, the conclusions drawn are not sufficiently detailed or in-depth. Currently, the five-move model of discourse move analysis, which comprises background, purpose, method, result, and conclusion [8,15], is generally accepted in academia. Yet, few studies have been conducted to investigate the Chinese stance markers, and how Chinese academic abstracts are structured and linguistically organized through discourse moves.
The stance classification framework proposed by Hyland [16,17] is based on an analysis of 240 dissertations from advanced degree programs, which is better suited for the linguistic features of academic discourse. Therefore, based on Hyland's stance marker classification framework, this study adopts a combination of quantitative and qualitative methods to examine the distribution and linguistic features of stance markers in the academic abstracts of international students in China and native Chinese speakers. The aim of this study is to reveal the construction of stance in international students' academic writing in Chinese and to deepen our understanding of their academic discourse writing practices in Chinese. The main questions explored in this study are as follows: 1) What are the similarities and differences in the structural features of the abstracts of international students in China and native Chinese speakers? 2) What are the similarities and differences between the stances expressed in the different rhetorical moves of the abstracts of international students in China and native Chinese language authors?

Move Analysis Model in Abstract
Based on the "five-move structure" model proposed by Dos Santos [8] and Hyland [16,17], this study manually annotated the two corpora. However, since this model is based on the abstracts of research articles, and the abstracts of dissertations and research articles have genre differences, this study refers to Biber et al. [2] for move identification and Cotos et al. [6] for move analysis based on this model. The researchers conducted preliminary discourse move delineation on 30 randomly selected discourses from each of the two corpora in advance: taking a single discourse as a starting point, we read the text content carefully and determined the discourse moves based on the logical relations of the discourse linguistic context and the characteristic information of the discourse. In the identification process, if an element could not be classified into five moves, it was determined as a new separate rhetorical move based on the frequency of the element's occurrence and its capability to fulfill a complete function [1,36]. To guarantee the reliability, the researchers performed the identification again after an interval of one month and tested the results of the two identifications for consistency. The results showed that the structure of the dissertation abstract was "six moves", including Background (M1), Topic (M2), Method (M3), Result (M4), Conclusion (M5) and Thesis Structure (M6) moves.
The "Thesis Structure" move (M6) is an important part of the macrostructure of the dissertation. Background move (M1) describes the background information of the research topic, the problems, or indicates the need for the research; Topic move (M2) describes the object, purpose or content of the study; Method move (M3) describes the theoretical framework on which the study is based, or the subjects, instruments, progresses, or research tools; Results move (M4) describes the results or findings of the study; Conclusion move (M5) evaluates research findings (e.g., significance, value, innovation), inferences, derivations, or interpretations of the results; Thesis Structure (M6) describes the structure of the dissertation.

Classification Framework for Stance Markers
This study refers to the stance marker classification framework proposed by Hyland [16,17], which consists of four broad categories: hedges, boosters, attitude markers and self-mentions. 1) Hedges: used to indicate the author's uncertainty or lack of complete certainty of the statement assertion. Examples include may, probably, suggest in English; "ke neng (可能)", "da yue (大约)" in Chinese. The use of hedges allows the author to reduce the degree of commitment to the proposition, allowing the reader to think critically and form their own judgment. 2) Boosters: usually used to indicate the degree of certainty of the statement or argument about a proposition. Examples include "certainly" and "demonstrate" in English and "ken ding (肯定)" and "shi shi shang (事实上) "in Chinese. Boosters help express the author's subjective certainty about the evidence for his or her argument. 3) Attitude markers: used to express the author's personal feelings or attitudes. Examples include "importantly" and "hope" in English and "zhong yao (重要)", "xi wang (希望)"in Chinese [17,33]. Attitudinal markers clearly convey the author's emotion and attitude toward the propositional message to readers, making them the most direct and clear expressions of the four types of stance markers. 4) Self-mentions: a first-person pronoun or noun used to refer to the author. Examples include I, my, me in English; "wo (我)", "wo men (我们)", "ben ren (本人)" in Chinese. Self-mentions reflect the author's identity and emphasize the author's authority over propositional information and his or her contribution to the academic field [20]. Due to the absence of a comprehensive system for classifying Chinese stance markers in academia, we proposed a framework system for Chinese stance markers based on the classification framework of Hyland [16,17] and after referring to the classification framework of Wu Geqi, Pan Chunlei [32] on Chinese stance markers. And we modified it to fit the corpora used in this study. Through the search, four types of stance markers were found to exist in two corpora. The classification of stance markers and the corresponding examples are shown in Table 1.

Corpus Collection
In this study, we collected abstracts of academic dissertations from international students in China as a corpus for second-language Chinese learners. Based on the Chinese Master's Theses Full-text Database (CMFD), We exhaustively searched for dissertations with authors' names in parentheses (e.g., "Siarhei Shamko (Sergei)") between 2007 and 2021, using linguistics and applied linguistics as the discipline subject. In addition, the papers of Chinese minority students who speak Chinese as a second language were excluded by checking manually according to the author's profile and thesis acknowledgments. Proportional sampling was then adopted, i.e., the specific sampling number for each year was calculated in proportion to the number of dissertations per year. Random sampling was performed to select 50 CSL master's theses with a total word count of 32,122 (excluding titles). Finally, a plain-text corpus of Chinese abstracts of master's theses of international students was constructed as a representation of the group of international students in China.
To provide more convincing results, this study exhaustively searched all doctoral dissertations in linguistics and applied linguistics from 2007 to 2021 based on the CNKI Chinese full-text doctoral dissertation database (CDFD). After manually eliminating non-native Chinese authors, the same proportional sampling was conducted, and then random sampling was performed. 50 doctoral dissertations by native Chinese authors, with a total word count of 83,676 (excluding titles), were finally selected to construct a corpus of Chinese abstracts of doctoral dissertations by native Chinese authors. Two corpora were constructed as the corpus of the abstracts by international students in China (CAIS) and the corpus of abstracts by native Chinese students (CACS). In order to have comparable observation frequencies, the text lengths of the corpora were normalized at the ratio of words per 10000 [22]. The basic information contained in the two corpora is presented in Table 2.

Identification and Classification of Move Structure
With reference to the "six-move structure" outlined in Section 2.1, this study read each text of the self-constructed corpora, identified its move composition, and manually annotated it. To ensure the reliability of the move International Students in China: A Corpus-Based Comparative Study identification, the two researchers underwent systematic training and then divided the two sets of texts into moves separately. A third-party assessor, such as an expert in the field, was consulted to arbitrate any ambiguous content. The data identified by the two researchers were then checked for consistency to ensure the reliability of the intra-linguistic recognition. Upon completion of the test, the consistency of annotation in CAIS and CACS reached 97% and 94% respectively, with a high level of confidence.

Identification of Stance Markers
Based on the above theoretical framework of Chinese stance markers, we conducted specific identification of stance markers in CAIS and CACS. In order to ensure the consistency of statistical units, we first used CorpusWordParser to subdivide the two groups of corpora and remove the language items that did not meet the requirements, such as the appearance of foreign names and foreign proper nouns. Then, using the corpus search tool AntConc 3.5.9 and the word list of stance markers with the corresponding definitions, we manually ranked the discourse units to classify them into different categories. It is worth noting that in determining which category a linguistic unit belongs to, it must be placed in a specific context and determined according to its function and role. In this process, the researcher needs to observe whether the discourse unit reflects the author's attitude or reflects the interaction between the author and the potential readers. Furthermore, it is important to distinguish the expression of the author's own position from the paraphrasing of others' views; as well as to distinguish the actual semantics of a given linguistic unit, especially polysemous or homonymic words, as the same written form can convey different meanings and have different linguistic functions [13]. It should be noted that the self-mentions counted in this study must be exclusive and refer only to the author of the study or the ideology, excluding researchers related to groups of readers or other studies.

Characteristics of Rhetorical Move Structure
The results (Table 3) indicate that there are some differences in the structural integrity of the abstracts between CAIS and CACS. The former group has the highest percentage of four-move abstracts (38%) and the lowest percentage of six-move abstracts (12%), while the latter group has the highest percentage of five-move abstracts (50%), followed by four-move abstracts (34%), suggesting that the abstract rhetorical structure written by native Chinese speakers was relatively more complete than that of the group of international students in China. In addition, the statistical results demonstrate that the case of only one move is almost non-existent, and the case of two moves is only found in the group of international students, which may be due to the fact that such expressions may make it difficult for readers to fully comprehend the main idea of the contents. To further analyze the distribution of the six moves in the two corpora, the average move length and the frequency of the moves were counted. The average move length was measured by assessing the proportion of each rhetorical move structure in the total number of words in the abstract. The frequency of each move was computed by counting the number of each move's occurrences in the two corpora. According to Dos Santos [8], the greater the importance of a rhetorical move, the greater the proportion of it. Table 4 shows that M2 has the largest share in both groups with31.16% and 32.78%, respectively, while M4 has a higher share with 21.63% and 32.57% respectively. This finding is in line with the research conducted by El-Dakhs [10] which suggests this is a prominent feature of scientific discourse. In contrast, the percentages of M3, M5 and M6 were lower in both groups. Additionally, the comparison reveals a difference in the proportion of M1 and M4: the group of international students prefers to use more space to describe the "research background", while the group of native Chinese speakers favors the description of "research background". The frequency of each move reveals that M2 move appears in all chosen discourses, but M5 and M6 occur at a lower rate. Native Chinese speakers demonstrate a higher use of M3, M4 and M6 (especially M4 is used twice as often as international students). International students in China were more likely to employ M1 and M5. This disparity may be attributed to the varying levels of experience with the research topic and proficiency in Chinese language between the two groups of students. Kanoksilapatham [18] found that moves with a frequency of over 60% are regarded as essential moves.
A comparison of abstracts in CAIS and CACS revealed that the majority of international students' abstracts featured three essential moves (M1, M2, M3); the majority of native Chinese speakers' abstracts featured four essential moves (M1, M2, M3, M4). This implies a difference in understanding of features of rhetorical moves between the groups [28].

Statistics on the Frequency of Stance Markers
This section aims to explore the distribution and features of stance markers in each rhetorical move and how the two groups of authors use these four types of markers to achieve different communicative functions. The following tables are the statistical results of stance markers after using 10000 frequency standardization.  Table 5 reveals both similarities and differences of use of stance markers between CAIS and CACS. In the two corpora, the distribution of the four major types of stance markers has some similarity, and their frequencies are, in descending order, hedges, self-mentions, boosters, and attitude markers. However, according to the results of the Chi-square test, there are significant differences between CAIS and CACS in the use of hedges, attitude stance and self-mentions. This suggests that international students in China may have a preference for certain linguistic strategies in order to express their stance or attitude in academic writing. Further research is needed to assess the impact of rhetorical moves and stance markers in different rhetorical moves on academic writing performances between two groups.
Hedges are negotiable and can help authors reduce the threat of losing face to other members of the academic community. By using these hedges, authors reduce the degree of their own commitment to the proposition, giving readers room to consider their judgment and possibly increasing the chances of the proposition being accepted [32]. In academic discourse, the first person is generally used sparingly as it is seen as too subjective and could lead to unobjective conclusions that lack credibility. However, in the abstracts of both corpora, the first person is used to emphasize the author's contribution to the academic field. Native Chinese speakers, on the other hand, influenced by the traditional Confucian culture of "no-self", tend to highlight the author's identity in an indirect way and avoid expressing too much personal emotion or value judgments. This can explain why the use of attitude markers and self-mentions in the abstracts of Chinese native speakers is lower than that of international students. Attitude markers are intuitive and distinct expressions of the writer's feelings, clearly conveying the writer's attitude toward the propositional message. However, too strong emotional color can weaken the authority and seriousness of the contents; therefore, relative to other stance markers, we can see that its occurrence is very low in this study.

The Use of Stance Markers in M1
The role of M1 in the abstract is to introduce the background of the study, including identifying any shortcomings of previous studies and demonstrating the importance of the current study [15]. As shown in Table 6, the total number of stance markers used by international students in China is greater than that of Chinese native speakers, and the difference in the frequency of hedges and boosters employed was greater, with international students in China using three times and more of these types of stance markers than Chinese native speakers; both groups have a preference for hedges and use self-mentions the least. After the Chi-square test, the results show that there is a difference in the use of hedges and boosters on M1 (p=0.000 and p=0.006). This prominent use of hedges by international students in China implies that they are limiting the truth of their propositions while demonstrating an objective and cautious attitude. And this preference is likely to make their opinions more acceptable to the readers. For example.
(1) 目前，基于汉韩、汉日、汉英可能表达形式对比的 汉语可能补语教学研究成果较多，而基于汉俄可能表达形 式对比的汉语可能补语教学研究成果较少。 (CAIS-no. 4) (2) 在汉语作为第二语言教学领域内，关于案例和围绕 案例衍生的各种概念存在很多模糊认识。 (CACS-no. 11) In examples (1) and (2), the authors employ the hedges "jiao (较)" and "hen duo (很多)" to diminish the certainty of the proposition so as to attenuate the threat of loss of face to other members of the academic community in the field, but quickly and effectively define the stance of the study. The boosters in CAIS are higher than those used in CACS. Unlike the function of hedges, in M1, boosters clarify the higher degree of certainty that authors hold about the value of their research, aiding them in emphasizing their research area and augmenting the persuasive power of the discourse. For example.
(3) 为了让两个民族顺利地交往， 必须了解对方的语言。 (CAIS-no. 7) (4) 不同语言存在着很多深层的共性以及诸多表达差异， 而不同语言的称数法作为语言系统的一部分，也必然存在 着一些共性和差异。 (CACS-no. 14) In examples (3) and (4), the words "bi xu (必须)" and "bi ran (必然)" enhance the credibility of the proposition and thus the persuasiveness of the paper. In M1, international students tend to use stance markers to express their opinions, while native Chinese speakers tend to state their opinions in an impartial manner to emphasize the significance of the proposition.

The Use of Stance Markers in M2
The M2 mainly focuses on the description of the object, purpose or content of the research. In M2, there is a significant difference in the stance markers used by international students and native Chinese speakers. Results show that international students are more proficient in using stance markers, especially hedges (p=0.000) and self-mentions (p=0.000), which help make the discourse more objective and increase the reliability of their proposition. The difference in the use of hedges words between the two groups was substantial, with the former using them 20.55/10,000 times and the latter only 2.39/10,000 times.
(5) 第四章主要介绍目前在越南汉语俗语教学中存在的 问题以及对此大胆提出教学建议。 (CAIS-no. 14) The use of "zhu yao (主要)" and "chang (常)" in Examples 5 and 6 implies the author's uncertainty towards the proposition, thus highlighting his or her caution and modesty towards the academic field, and leaving some face for other scholars in the field. In examples (7) and (8), "bi zhe (笔者) ", "ben yan jiu (本研究) ", and "ben ke ti (本课题)" are presented as nouns, which can express the intention of self-reference more subtly than the direct use of first-person pronouns. In general, in M2, international students in China are more inclined to use hedges and self-mentions to present the research contents and make the discourse more objective and enhance the credibility of their proposition.

The Use of Stance Markers in M3
In M3, both groups show low usage of stance markers. This could be attributed to the fact that this move mainly necessitates an objective description of the research design, such as data, methods, and procedures, which does not involve much narrative expression. Results show a significant difference in the usage of hedges and self-mentions between the two groups, with p=0.02<0.05 for the former and p=0.000 for the latter. In M3, International students are more likely to use self-referential language strategies to emphasize their role in the research process and the authority of their opinions.
(9) 这一部分主要通过调查问卷的方式对俄罗斯留学生 汉语复合趋向补语引申义用法的使用情况进行调查。 (CAIS-no. 5) In example (9), the use of "zhu yao (主要)" highlights the emphasis placed on the research methodology in the article and constructs rigorous authorship. The author provides a detailed description of the methodology in a logical and scientific manner, demonstrating the main progress of the research.

The Use of Stance Markers in M4
M4 is generally used to present the main findings of the study accurately, avoid arbitrariness and leave room for discussion. Although the proportion of M4 in the abstracts of international students is relatively small, the frequency of stance markers is very high. The analysis of the data reveal that there was no significant difference between CAIS and CACS in terms of the frequency of stance markers used, with hedges, boosters, self-mentions, and attitude markers being the most frequently used, in descending order. In M4, both groups tend to use hedges to imply the uncertainty of the proposition, which weakens the author's responsibility for the risk, and at the same time expresses the author's respect for the reader's freedom to judge the acceptability of the argument.
Unlike other rhetorical moves, in M4, Chinese native speakers use stance markers more frequently. For both native Chinese speakers and international students, the presence of boosters can strengthen the author's certainty of his or her own opinion and confidence in the research results, which in turn can infect the reader, thus enhancing the persuasive power of the discourse. It can also highlight the author's confidence in the research results and show that he or she is the owner of the results and is responsible for them. However, international students and native Chinese speakers hardly use attitude markers at this step, 1.25/10,000 times and 2.99/10,000 times, respectively. This indicates that both do not tend to use personal emotions to reflect their attitudes toward the proposition and communicate with the reader potentially. members of both groups hardly use self-mentions in M4, which occur 5.29 and 5.98 times per 10,000 words, respectively, showing modesty and not using singular first-person pronouns to highlight the authorship and emphasize the credibility of the findings. Thus, it can be concluded that authors of both groups utilize stance markers in order to present their findings accurately and to allow room for discussion in M4.

The Use of Stance Markers in M5
The primary function of M5 is to facilitate the comprehension of research findings by summarizing, discussing, evaluating, inferring, and explaining them [15]. As Table 10 shown, a significant difference between the two groups is observed in terms of the use of stance markers, with the subcategories of booster and attitude markers differing most significantly. This suggests international students tend to prefer the use of stance markers in M5, likely to enhance the persuasiveness of the proposition, express hope for related future research, reduce the absolute nature of the proposition using hedges, and emphasize their contribution to the research results through self-mentions and they are more concerned with expressing emotions. In comparison, native Chinese speakers tend to discuss the research results relatively objectively and express fewer personal emotions.

The Use of Stance Markers in M6
The function of M6 assists readers in comprehending the organization of the thesis more effectively and facilitates them in reading the thesis in a more expedient way. CAIS has a higher frequency of stance markers than CACS, which was 26.46 and 8.96 times per 10,000 words, respectively. This study reveals that M6 often appears in combination with M2 to introduce the content of the research.  (13), the author employs the hedge "zhu yao (主要)" to blur the degree of the certainty towards the proposition and enhance the objectivity of the proposition's viewpoint. Example (14) utilizes the noun "ben yan jiu (本研 究)" to allude to the author's identity obliquely.
Although the frequency of occurrence of M6 in CAIS is International Students in China: A Corpus-Based Comparative Study low, the proportion of it in whole texts is high, and the number of stance markers used by international students in China is higher than that of native Chinese speakers. This may be due to the negative transfer effect of second language learning for international students in China.

Conclusion
In this study, based on Hyland's classification model of stance markers, we conducted a comparative analysis of the rhetorical moves and the use of stance markers in each move in the dissertation abstracts by international students in China and native Chinese speakers.
After analyzing the rhetorical move structures, both CAIS and CACS reveal that most international students' abstracts contain three essential moves (M1, M2, M3), while most native Chinese speakers' abstracts included four essential moves (M1, M2, M3, M4). Among these six moves, it is found that Chinese international students had significantly higher hedges in M1, M2, M3, M5, and M6 compared to native Chinese language speakers. In terms of boosters, international students have a significantly higher frequency of usage in M1 and M5. Additionally, international students employ more self-mentions in M2, M3, M5, and M6. Moreover, there is a clear difference between the two groups in their use of attitude markers, with international students having a higher frequency of usage in M5. The result suggests that Chinese international students pay more attention to pre-research preparation when constructing an academic discourse, whereas native Chinese language speakers focus more on research results and discussions. The difference may be due to various factors, such as their different thinking modes and the influence of academic culture discrepancies [31]. Therefore, to ensure effective academic information dissemination and respect the language expression of different cultural backgrounds, attention should be paid to the teaching of writing under the rhetorical move structure. The distributions of the four types of stance markers are similar in the two corpora, with hedges having the highest frequency of occurrence, followed by self-mentions, boosters, and attitude markers. And it is found that the stance markers distributed across the rhetorical moves are closely connected to the communicative functions carried by the rhetorical moves.
The findings of this study can bring some pedagogical implications for academic writing. It was found that international students studying in China consciously use different stance markers to promote their research results when writing academic abstracts in Chinese, but at the same time, they also show certain limitations. Therefore, they need to receive effective academic instruction [38] to improve their expressive skills in the Chinese academic language.