English Language Teaching and Linguistics Studies ISSN 2640-9836 (Print) ISSN 2640-9844 (Online) Vol. 5, No. 3, 2023 www.scholink.org/ojs/index.php/eltls 199 Original Paper A Study on Identity Construction of First Person Pronouns in Academic Papers from the Perspective of Evidentiality Ziqi, Zhao 1* & Xuefei, Wang 1 1 Faculty of Humanities and Social Sciences, Beijing University of Technology, Beijing, China * Ziqi, Zhao, Faculty of Humanities and Social Sciences, Beijing University of Technology, Beijing, China Received: June 22, 2023 Accepted: August 02, 2023 Online Published: August 07, 2023 doi:10.22158/eltls.v5n3p199 URL: http://dx.doi.org/10.22158/eltls.v5n3p199 Abstract The first person pronoun plays an important role in identity construction, however, there is few study on it from the perspective of evidentiality. This paper took the first person pronouns as evidentials, and conducted a comparable analysis on the frequency of them and the identities they constructed in academic papers between soft and hard sciences, aiming to find the differences between different discourse communities and explore their preferences for academic identity construction. The results showed that both fields prefer to use plural and subjective cases of first person pronouns, and they both prefer to construct the authorial identity of “researcher”, but scarcely construct the authorial identity of “responsible person”. Researchers in hard science use less evidentials than researchers in soft science, and they prefer to use evidentials “we” and “statement”, which weaken the authorial identity. Evidentials that embody authorial identity, including singular first person pronouns and “participation” evidentials, account for higher proportion in soft science than those in hard science. Keywords Evidentiality, Evidentials, First Person Pronoun, Identity Construction 1. Introduction Identity construction refers to the process in which speakers or writers use language resources to construct tenor of discourse and realize interpersonal function of discourse (Zhao et al., 2012). With the rise of post-structuralism, researchers no longer regarded identity as a fixed feature (Chen, 2013), but as a production produced in a dynamic interacting process. In other words, “identity” is generated intersubjectively, not individually (Bucholtz & Hall, 2005). Therefore, identity construction is a dynamic process realizing through communication. www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 200 Published by SCHOLINK INC. Discourse is an important carrier of identity. The writer communicates through discourses and constructs certain identities through a combination of his or her various discoursal choices in communication (Chang, 2018). And the reader also plays an essential role in the identity construction process. Before the reader reads the text, the writer’s identity exists in a “vacuum”. Only after being received and understood by the reader, will the ideas expressed by the writer through specific language resources can come into play and construct an authorial identity (Burgess & Ivanič, 2010). In short, the construction of identity is a process in which the writer and the reader work together to construct the authorial identity. Academic paper is the major place for scholars to construct authorship (Tang & John, 1999). When the authors objectively present their research results in their academic papers, the language they use will conveys their attitudes and ideas, then a specific authorial identity is constructed (Hyland, 2002a). Ivanic (1998) pointed out that authorial identity contains three dimensions: autobiographical self, discoursal self and self as author. Among them, “self as author” carries the author’s ideas and is the most effective interaction means between the author and the reader, but it was often ignored in academic context before. In general, traditional academic writing aims at being ‘objective’ and “scientific” in its expression of ideas. Subjective words such as “I think” and “we conclude” should be avoided as much as possible, because self-mentions will add a tinge of subjectivity to academic papers, which goes against the paper’s objectivity (Arnaudet & Barrett, 1984, p. 73). However, some authors argued that not all discourse communities follow the same conventions, and that readers in different fields have different expectations for argumentative discourse. Therefore, the argument of “avoiding the use of self-mentions” is too general because it ignores the differences across disciplines. In addition, in the 21st century, more and more researches have shown that the authorial identity is an essential element in academic writing, which contributes to establishing the interactive relationship between the author and the reader, realizing the communicative function of the academic papers (Lou & Wang, 2020). Besides, author’s self-mentions are helpful to presents author’s academic achievements in the academic fields, promoting the authority of the paper and helping the author establish an academic identity in the academic discourse community to which the author belongs. In academic papers, the author’s self-mention is an important discourse strategy for constructing authorial identity (Ivanic, 1998, p. 255; Hyland, 2002b). Different personal pronouns can present different attitudes and positions of the author and construct various authorship. Among these pronouns, the strongest manifestation of the authorial identity is the first person pronoun, which is crucial in expressing meaning and establishing credibility (Tang & John, 1999). They help to establish credibility of authority and increase the accuracy of the paper (Mur Dueñas, 2007). Previous studies on identity construction of first person pronouns were conducted quantitatively or qualitatively from the perspective of different disciplines or different language backgrounds of authors. These studies have found that differences in disciplinaries and linguistic backgrounds will influence the identity construction. Li and Xiao (2018) found that linguistic background will influence authors’ www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 201 Published by SCHOLINK INC. choices of first personal pronouns, and the authorial identities constructed. Yang (2015) found that authors of different linguistic background have different preferences for discourse roles. And some researchers compared the self-mentions in academic papers of different disciplines to explore the identities constructed (Harwood, 2005; Hyland, 2001; Kuo, 1999). However, few previous studies have studied the identity construction from the perspective of evidentiality. Evidentiality is a discourse strategy that reflects the speaker or writer’s commitment to the source of knowledge and the reliability of information. Evidentials refer to the evidences that reflect evidentiality and source of information. Academic papers aim to convey authors’ views to readers, then whether the readers accept them or not, to some extent, depends on the author’s credibility, and the sources of information and knowledge in the papers. The study on the evidentials in academic papers is helpful to understand how authors use discourse strategies to convince readers of their ideas. Hyland (2012, pp. 23-46) found that authors will use discipline-related language resources and discourse practices to get their academic identities recognized by readers from the same academic filed, and the academic papers of different disciplines tend to construct different identities. Therefore, a comparative study of the evidentials in academic papers of different disciplines can help people know what kind of identity each discipline tends to construct and what kind of discourse strategies authors of different disciplines tend to use. Previous studies seldom treated first person pronouns as evidentials. Taking the first person pronouns as evidential can help people more clearly understand the function of discourse strategy of the first person pronouns, so as to master the use of such evidentials and better construct the appropriate authorial identity in academic papers. From the perspective of evidentiality, this paper will compare the use of evidentials of first person pronouns in academic papers of different disciplines, aiming to explore the differences and similarities of evidentiality and identities constructed in soft and hard sciences, and revealing their preference for the use of the evidentials of first person pronoun, in order to provide references for academic writings in different academic fields. The corpora analyzed by this paper is relatively new, which can reflect the modern international academic writing norms, and help scholars understand the preference of first person pronouns in international journals with high citation. 2. Research Framework 2.1 Evidential Theory Evidentials originally referred to the grammatical suffixes in some language that signifies the source of information, and evidentiality refers to this linguistic phenomenon. Boas was the first to identify evidentiality in a language and to use the term “evidential” to refer to the “evidence” of the source of information in a statement. Later studies expanded the definition of evidentiality. Chafe and Nichols (1986) regarded evidentiality as a set of means to express different attitudes towards information, that is, evidentiality includes not only the reliability of information but also the speaker or writer’s own attitude towards information mentioned in the discourse. Thus, evidentiality can be regarded as a www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 202 Published by SCHOLINK INC. linguistic strategy used to convey reliability and express attitude, while evidentials are defined as a linguistic resource used to mark evidentiality. 2.2 Classification of Evidentials After furthering the research on evidentials, scholars have proposed different classifications. Chafe and Nichols (1986) classified evidentials into belief evidence, induction evidence, hearsay evidence and deduction evidence. Yang (2015) classified evidentials into four categories: sensory evidential, reporting evidential (including other-reporting and self-reporting evidentials), inferring evidential and belief evidential. Based on Yang’s classification, Liao and Weng (2022) subdivided inferring evidentials into inferring evidential of high-uncertainty and that of low-uncertainty. Yang et al. (2019) divided evidentials into belief, inference, hearsay, reliability and expectation. Taking Yang’s (2015) classification as an example, the definitions and usage of the evidentials are illustrated in Table 1. Table 1. Categorizations of Evidentials by Yang (2015) Evidential types Definitions Examples Sensory evidential The information of proposal comes from first-hand experience. The source of information includes many sensory channels, like vision or auditory sense. Thus, in Example 1, we see the speaker using …… Reporting evidential Self-reporting evidential The information source is related to the author or something in the paper As we have discussed …… As shown in the figure I got Other-reporting evidential The information comes from other people rather than the author, or the resource outside the paper According to Francis (1994: 89), retrospective labels are much more …… Inferring evidential The information is based on conjecture process like induction and deduction It is very likely both… It would certainly be interesting to know …… Belief evidential The information comes from the authors’ beliefs or opinions. It can be subjective or objective I think, I believe, I argue, we suppose, it can be suggested This paper studies the evidentiality of academic papers on the basis the first person pronouns, so the first person pronouns will be regarded as the evidentials. In consideration of the functions of first person pronouns in academic papers, this study will propose a new framework fit for evidentials of first person pronouns. Previous researchers also categorize the constructed identities. Tang and John (1999) divided the identity constructed by researchers into representative, guide through the essay, architect of www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 203 Published by SCHOLINK INC. the essay, recounter of the research process, opinion-holder and originator. Wu (2013) divided authorial identities into researcher, discourse constructor, opinion holders and evaluators. Liao and Weng (2022) classified authorial identities into four categories: authorial identities of credibility, respect, authority and responsibility. In consideration of the role of first person pronouns play in academic papers, this study classifies authorial identities into responsible person, guider and researcher (see Table 2). Table 2. Classification of Evidenitals of First Person Pronouns Evidential types Corresponding identities Example Engagement evidential Responsible person I think, in my view, we all know Deductive evidential Guider so that I/we can conclude, according to my/our research/data, I have argued that, so we can see Declarative evidential Researcher I mark this as, I will study/ argue, our approach In this framework, the evidential types include engagement evidential, deductive evidential and declarative evidential, which are all used to indicate sources of information or express opinions. Engagement evidential means the author explicitly engage in the articles to express his or her own opinions and judgments, such as “I think” and “in my opinion”, which is highly subjective, or the author directly informs the reader of certain information, like “we all know that” or “we cannot deny that”, implying that the reader should also know the information. Such evidentials correspond to the identity of “responsible person”, which is used by the author to introduce his or her own opinions and declare that he or she is responsible for the discourse. Deductive evidential, on the basis of logic, preceding part of the text, common sense, etc., guides the reader to connect the context and explains the source of information to the reader, such as “so we can conclude”. This kind of evidential corresponds to the identity of “guider”, which aims to guide the reader to follow the development and understand the idea of the paper. Declarative evidential is used to present the research process or summarize the results, such as “I will conduct a research” or “I computed the data”, which presents to readers how the author get the information, that is, the results or conclusions. This evidential show the readers the research process objectively and construct the identity of “researcher”. 3. Research Methods 3.1 Research Questions The first person pronouns are regarded as evidentials. This part will compare the differences in the use of such evidentials between academic papers of soft and hard sciences, and answer the following two questions: (1) What are the differences in the frequency of three types of evidentials between academic www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 204 Published by SCHOLINK INC. papers of soft and hard sciences? (2) Which identity do hard and soft sciences prefer respectively? 3.2 Research Data The research corpora in this study were selected from DEAP Baby (V1.0) (Sun & Xu, 2022). Its full name is Database of English for Academic Purposes Baby version 1.0. The complete version DEAP contains academic papers with high citation from international journals published from 2011 to 2021. It covers a total of 25 disciplines. The DEAP Baby (V1.0) was re-sampled from the DEAP corpus and still covered 25 disciplines, with complete sentences of about 50,000 words randomly selected for each discipline. Academic papers of biology, chemistry, linguistics and sociology were selected from DEAP Baby (V1.0) as the research objects. There are 49,082 tokens in BioDEAP (biological) and 48,455 tokens in ChemDEAP (chemistry). LinDEAP (linguistics) contains 49,290 tokens and SociDEAP (sociology) contains 49,096 tokens. These four corpora are comparable in terms of volume. In this study, the hard science academic paper corpus HDEAP with a total of 97,537 tokens was established by combining the BioDEAP and ChemDEAP, while the soft science academic paper corpus SDEAP with a total of 98,386 shapes was established by combining the LinDEAP and SociDEAP. The two corpora are also comparable. 3.3 Research Methods This paper takes the sentence as the analysis unit and analyzes the function of each first person pronoun in line with the context, so as to judge and annotate the evidentials of first person pronouns. The software AntConc3.5.9 was used to carry out the quantitative analysis of evidentials in the corpora, which was supplemented with qualitative analysis. The “concordance” function of software was used to search and count the frequency of the evidentials of first person pronouns. The count included only those first person pronouns which present evidentiality. The detailed procedures are as follows: Firstly, the concordance function of Antconc3.5.9 was used to search the corpora for the first personal pronouns such as “I”, “we” and “me” successively, and then the collocations found were read one by one to determine whether the pronoun belongs to evidentials, and then the collocations that do not meet the conditions are deleted. For example, “my” in example 1 plays a narrative role instead of serving as an evidential, so it is excluded from the evidentials list. In addition, there will be some reporting sentences in papers, like example 2. Such sentences are direct quotations from others, and the first personal pronouns in such sentences are not used by authors to express evidentiality, so such first person pronouns were also out of the counting list. Finally, first personal pronouns in translational sentences (example 3) were also excluded. (Example 1) ... , one of my best friends came three months later… (Example 2) He said, “We began the survey yesterday.” (Example 3) ... , for example, j’m interesser means “I am interested in”,... After this procedure, the functions of the remaining evidentials in the context were analyzed according to their collocations. Then these evidentials were classified and conducted according to the www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 205 Published by SCHOLINK INC. classification framework in Table 2. Last, the comparative analysis was carried out based on this result. 4. Results and Discussion The frequency of evidentials of first person pronoun in each discipline was firstly counted (see Table 3), and then the frequency was summed up and compared according to soft and hard sciences (see Table 4). Table 3. The Frequency of Evidentials of First Person Pronouns in Each Discipline I me my we us our Total number LinDEAP Declarative evidentials 21 3 7 164 17 63 275 Deductive evidentials 1 0 0 9 0 1 11 Engagement evidentials 10 2 1 38 6 0 57 Total number 32 5 8 211 23 64 343 SociDEAP Declarative evidentials 28 0 3 127 8 68 234 Deductive evidentials 5 0 0 8 1 3 17 Engagement evidential 12 0 1 24 0 5 42 Total number 45 0 4 159 9 76 293 ChemDEAP Declarative evidentials 0 0 0 96 3 13 112 Deductive evidentials 1 0 0 4 1 4 10 Engagement evidential 0 0 0 4 0 1 5 Total number 1 0 0 104 4 18 127 BioDEAP Declarative evidentials 6 0 0 152 1 29 188 Deductive evidentials 1 0 0 2 0 1 4 Engagement evidential 0 1 1 13 0 3 18 www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 206 Published by SCHOLINK INC. Total number 7 1 1 167 1 33 210 Total number 85 6 13 641 37 191 973 Table 4. The Frequency of Evidentials of First Person Pronouns in Soft and Hard Sciences SDEAP HDEAP Declarative evidentials Deductive evidentials Engagement evidentials Total number Declarative evidentials Deductive evidentials Engagement evidential Total number I 49** 6 22** 77** 6** 2 0** 8** me 3* 0 2 5 0* 0 1 1 my 10** 0 2 12** 0* 0 1 1** we 291 17* 62** 370** 248 6* 17** 271** us 25** 1 6** 32** 4* 1 0** 5** our 131** 4 5 140** 42** 5 4 51** Total number 509** 28* 99** 636** 300** 14* 23** 337** Note: *p<.05 **p<.01. 4.1 Comparative Analysis on the Frequency of Each Evidential The first person pronouns got from the research corpora include nominative personal pronouns (I, we), accusative personal pronouns (me, us) and adjectival possessive pronouns, but do not include nominal possessive pronouns (mine, ours). By comparing the data in Table 3, it can be seen that there are few differences in the frequencies between linguistics and sociology, and between biology and chemistry. So academic communities in the same field have similar preferences in terms of evidential. By comparing the data between soft and hard sciences in Table 4, it can be seen that most of the frequencies have significant differences, which indicates that there exist significant differences between soft and hard sciences. In terms of quantity, the total number of frequencies of evidentials of first personal pronouns used in academic papers in soft science is about twice that in hard science. The reason is that hard science is based on experimental data and usually uses the third person pronoun “it”. In order to enhance the objectivity of research, hard science papers often use passive sentences and avoids using of first person pronouns, so in the results, the frequency of hard science authors using first person pronouns is very low. As can be seen from Table 4, the frequency of the plural pronouns is higher than that of their corresponding singular pronouns. For example, the frequency of “we” is greater than that of “I”, and the frequency of “our” is greater than that of “my”. This result indicates that the authors of international journals prefer to use plural first person pronouns in both soft and hard sciences. The reasons are as follows: firstly, this preference is influenced to a certain extent by traditional academic view, which www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 207 Published by SCHOLINK INC. claims that academic papers should state research facts and convey information in a neutral and objective manner, and authors should avoid elaborating their personal identities and positions (Arnaudet & Barrett, 1984, p. 73). The plural first person pronouns can weaken the author’s identity. By using “we” and “our”, the author can realize the interaction with the reader and stand on the same point with the reader, weakening his or her own authorial identity. Hyland (2001) also pointed out that the plural form of the first person pronoun emphasizes the group attribute and highlights the shareability of the mentioned knowledge, alleviating the threat of face. Secondly, many current academic papers in international journals are co-authored by multiple authors, which also increases the frequency of plural first personal pronouns to some extent. By further comparing the proportion of each evidential (see Figure 1-2), it is found that the evidential “we” accounts for a large proportion of evidentials used in hard science, while the singular forms “my” and “me” accounts for very small proportion. In soft science, the evidential “we” accounts for the largest proportion, but it is less than that in hard science, while the proportions of singular forms are higher than those in hard science. It shows that the hard science prefers to construct a weaker authorial identity by using first person pronouns that weaken authorial identity, while the soft science constructs a more obvious authorial identity than the hard science. Figure 1. The Proportion of Each Evidential in Soft Science www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 208 Published by SCHOLINK INC. Figure 2. The Proportion of Each Evidential in Hard Science Although it was previously assumed that academic paper should be objective and neutral in its expression of ideas and avoid specific reference to personal identity and opinions, more and more studies have pointed out that the first person pronouns contribute to enhancing the interactivity and persuasive of discourse by constructing overt and authoritative self-identities. Wu and Pang (2009) randomly selected 10 academic papers published from 2004 to 2009 in SCI- and SSCI-indexed scholarly journals, and searched the first person pronouns in the papers, finding that the frequency of singular first person pronouns in engineering papers was zero, while the use of plural first person pronouns accounted for 1.49‰ of the total words. This paper selects academic papers published between 2011 and 2021 for research, and finds that the frequency of evidentials of singular first person pronouns in hard science is also much lower than that of plural first person pronouns, but there is still a very small amount of evidentials of singular first person pronoun. Besides, the frequency of evidentials of plural first person pronouns accounts for 3.35‰ of the total words in hard science papers, which is greater than the value of 1.49‰ in Wu and Pang’s study (2009). These suggest that international authors have become more accepting of the use of first person pronouns in hard science academic papers in recent years, and are able to establish a clearer authorial identity in their papers. In terms of cases of first person pronouns, it can be seen from Table 4 that, whether in soft or hard science, authors prefer to use evidentials of nominative first person pronoun “I” and “we”, while other evidentials are relatively less frequently used, because compared with nominative forms, the accusative and possessive forms are hard to convey the author’s thoughts to readers and to the manifest the authorial identity (Mur Dueñas, 2007). In addition, the collocations of accusative personal pronouns which can express evidentiality are also relatively fixed, such as “it seems to me” and “allow us to find”, so the frequency is lowest. www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 209 Published by SCHOLINK INC. 4.2. Comparative Analysis on the Frequency of Each Evidential Type From the perspective of evidential type, the most frequently used evidential in soft science and hard sciences is declarative evidential, and the least used is deductive evidential. Declarative evidential is mainly used to describe the research content and procedure, so as to show the source of information and the reliability. This type of evidential corresponds to the identity of researcher, which is relatively objective, but simply gives narration instead of persuasion. Such authorial identity is weak in authority. The total frequency of evidentials of first person pronouns elements in academic papers has increased compared with that in papers in the past, and the awareness of constructing authorial identity has been enhanced, however, the increased evidential type is mainly declarative evidential which construct a relatively weak authorial identity. Comparing Figure 3 with Figure 4, it can be seen that declarative evidentials account for a larger proportion in hard science than those in soft science, because hard science mainly abides by logic and practice, and takes discovery and explaination as target. Authors need to introduce the research process, showing the reproducibility of the research, so as to increase the scientificity and credibility of the paper in the discourse community of the discipline. Figure 3. The Proportion of Each Evidential Type in Soft Science www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 210 Published by SCHOLINK INC. Figure 4. The Proportion of Each Evidential Type in Hard Science Deductive evidential accounts for the least in both soft and hard sciences, probably because when academic authors tend to use deductive evidentials, they prefer other expressions, like “according to the data” or “the table in the previous part shows,” rather than expressions that include evidentials of first person pronoun. Therefore, what kind of evidentials the academic authors use to construct the identity of guider needs to be further explored. Finally, the engagement evidentials are used to express the author’s attitudes and opinions. The evidentials of first person pronouns, especially, present higher subjectivity and can be used to construct the identity of the responsible person. By comparing Figure 3 and Figure 4, it is found that engagement evidentials account for a larger proportion in soft science than those in hard science. It can be seen that soft science authors are more likely to engage in the academic papers and express opinions, and the subjectivity in soft science academic papers is higher than that in hard science. When an academic author chooses the participative engagement evidential of first person pronoun, it means that the author claims to take full responsibility for the ideas and information in the paper. Compared with hard science, soft science is a knowledge system built on the basis of ideology or interests, focusing on understanding and explanation. Therefore, it is easier for soft science authors to express their own opinions in the paper, showing more obvious authorial identity. 5. Conclusion Previous studies have found that disciplinary characteristics affect the use of self-mentions (Hyland, 2012, pp. 23-46; Hyland & Jiang, 2017). Each discipline has its own discourse community with its corresponding thesis norms, so there are different preferences for the evidentials of first person pronouns. This paper collected academic papers from four disciplines, taking soft science and hard science as discourse community units, and makes statistics on the frequency of evidentials of first person pronouns in them, exploring their preferences for different evidential types of first person www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 211 Published by SCHOLINK INC. pronouns and the authorial identities they constructed. It is found that there exist significant differences in the frequency of evidentials of first person pronouns between soft and hard sciences, and the total frequency in soft science is much higher than that in hard science. Compared with the soft science, the proportion of evidential “we” in hard science is very large, while the proportion of “me” and “my” is very small. There are also similarities between the two fields that authors of international journals in both fields prefer the plural form of the first person pronouns to the singular form, and the nominative form to the accusativeform. From the view of the constructed identity, the hard science often uses declarative evidentials and constructs the identity of “researcher”. Compared with the hard science, engagement evidentials account for a higher proportion in soft science, constructing the identity of “responsible person”. Both disciplinary fields use deductive evidentials and construct the identity of “guider” least frequently. The analysis of these results shows that the contributing factors to the differences are the discourse characteristics of disciplinary community and the characteristics of academic papers. Hard science community presents more objectivity in its academic discourse that its papers, based on experimental data, usually try to highlight the replicability of research and the generality of research results. Therefore, hard science seldom uses evidentials of singular first person pronouns, but often uses evidentials that weaken the author’s identity, such as the evidential “we”. It can reduce the degree of engagement and avoid strong impersonality by hiding authorial identity, while implicitly conveying the author’s own view at the same time. However, the frequency of evidentials of first person pronouns is relatively high in soft science, for the authors need to show authority to expect the readers’ recognition of their academic arguments based on their own subjective interpretations. In addition, the engagement evidentials imperceptibly place the author’s views within the readers’ knowledge framework, such as “as we know that”, or directly express the author’s views, such as “in my opinion”. Both of these methods help the author show authority and construct the identity of the “responsible person”, so the proportion of engagement evidentials in soft science academic papers is also higher than those in hard science academic papers. In terms of declarative evidentials, which are used to describe the research procedures, hard science uses more than soft science, since the authors of the former need to show the research procedures to the readers of the same discourse community in order to increase the credibility of the article. In conclusion, compared with hard sciences, soft sciences use more evidentials, and prefer to construct identities that manifest engagement. However, authors of hard sciences prefer invisible authorial identities. The same knowledge system and thinking mode in the same discipline together constitute a discourse convention, so evidentiality will also show corresponding characteristics in academic papers of different disciplines. Understanding the differences in the use of evidentials is helpful for academic authors to better learn the writing norms and conventions of the international discourse community and establish appropriate authorial identities. It is also helpful for teachers to cultivate students’ ability to use evidential discourse in academic writing while take into account the writing norms and conventions www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 212 Published by SCHOLINK INC. of discourse community. References Arnaudet, M. L., & Barrett, M. E. (1984). Approaches to Academic Reading and Writing. Hoboken, NJ: Prentice-Hall. Bucholtz, M., & Hall, K. (2005). Identity and interaction: A sociocultural linguistic approach. Discourse Studies, 7(4-5), 585-614. https://doi.org/10.1177/1461445605054407 Burgess, A., & Ivanič, R. (2010). Writing and Being Written: Issues of Identity Across Timescales. Written Communication, 27(2), 228-255. https://doi.org/10.1177/0741088310363447 Chafe, W., & Nichols, J. (1986). Evidentiality: The Linguistic Coding of Epistemology. Norwood: Ablex. Chang, X. L. (2018). A Review of Research on Writer Identity Construction in English Writing: Concepts, Theories and Methods. Foreign Language and Literature, 34(02), 138-144. Chen, X. R. (2013). Pragmatic Identity: Dynamic Choice and Discursive Construction. Foreign Language Research, (04), 27-32. Harwood, N. (2005). Nowhere has anyone attempted … In this article I aim to do just that. Journal of Pragmatics, 37(8), 1207-1231. https://doi.org/10.1016/j.pragma.2005.01.012 Hyland, K. (2001). Humble servants of the discipline? Self-mention in research articles. English for Specific Purposes, 20(3), 207-226. https://doi.org/10.1016/S0889-4906(00)00012-0 Hyland, K. (2002a). Options of identity in academic writing. ELT, 56(4), 351-358. https://doi.org/10.1093/elt/56.4.351 Hyland, K. (2002b). Authority and invisibility: Authorial identity in academic writing. Journal of Pragmatics, 34(8), 1091-1112. https://doi.org/10.1016/S0378-2166(02)00035-8 Hyland, K. (2012). Disciplinary Identities: Individuality and Community in Academic Discourse. Cambridge University Press. https://doi.org/10.1017/9781009406512 Hyland, K., & Jiang, F. K. (2021). Our striking results demonstrate …: Persuasion and the growth of academic hype. Journal of Pragmatics, 182, 189-202. https://doi.org/10.1016/j.pragma.2021.06.018 Ivanič, R. (1998). Writing and identity: The discoursal construction of identity in academic writing. Amsterdam: Benjamins. https://doi.org/10.1075/swll.5 Kuo, C. (1999). The Use of Personal Pronouns: Role Relationships in Scientific Journal Articles. English for Specific Purposes, 18(2), 121-138. https://doi.org/10.1016/S0889-4906(97)00058-6 Li, M., & Xiao, Y. (2018). A Study on Interactiveness of English Research Atricles: Evidence from the Use of the First Person Pronouns and Their Identity Construction. Journal of Xi’an International Studies University, 26(02), 18-23. https://doi.org/10.16362/j.cnki.cn61-1457/h.2018.02.004. Liao, Q. Y., & Weng, X. (2022). A Comparative Study of the Authorial Identity Construction English Academic Writing from the perspective of Evidentiality. Journal of Xi’an International Studies www.scholink.org/ojs/index.php/eltls English Language Teaching and Linguistics Studies Vol. 5, No. 3, 2023 213 Published by SCHOLINK INC. University, 30(01), 24-29. https://doi.org/10.16362/j.cnki.cn61-1457/h.2022.01.006 Lou, B. C., & Wang, L. (2020). Self-mentions and Authorial Identity Construction in Chinese Learners’ English Academic Writing. Journal of PLA University of Foreign Languages, 43(01), 93-99. Mur Dueñas, P. (2007). “I/we focus on…”: A cross-cultural analysis of self-mentions in business management research articles. Journal of English for Academic Purposes, 6(2), 143-162. https://doi.org/10.1016/j.jeap.2007.05.002 Sun, M. C., & Xu, J. J. (2022). DEAP Baby Corpus V1.0. In National Research Centre for Foreign Language Education. Beijing Foreign Studies University Tang, R., & John, S. (1999). The `I’ in identity: Exploring writer identity in student academic writing through the ®rst person pronounp. English for Specific Purposes, (18), 23-39. https://doi.org/10.1016/S0889-4906(99)00009-5 Wu, G. Q. (2013). Self-mention and Authorial Identity Construction: A Corpus-based Contrastive Study between English and Chinese Research Articles. Journal of PLA University of Foreign Languages, 36(03), 6-11. Wu, Z. H., & Pang, J. X. (2009).A Study of Construction of Self as Author in Scholarly Journals. Foreign Language Education, (00), 78-83. Yang, L. X. (2015). Academic Identity in English Research Articles: From Evidentiality Perspective. Foreign Language Education, 36(02), 21-25. https://doi.org/10.16362/j.cnki.cn61-1023/h.2015.02.005 Yang, X. R. (2015). A Study of Authorial First Person Pronouns and Writer Identity Construction in EFL Academic Writing. Foreign Languages and Their Teaching, (04), 50-56. https://doi.org/10.13458/j.cnki.flatt.004104. Yang, Y. W., Xu, M. Y.Z., & Liu, Q. H. (2019). A Contrastive Study on Evidentiality of Academic Research Articles in English and Chinese. Journal of Xi’an International Studies University, 27(03), 48-52. https://doi.org/10.16362/j.cnki.cn61-1457/h.2019.03.010 Zhao, Y. Q., Chen, J., Huang, T., & Qin, L.L. (2012). EFL Public Speakers’ Identity Construct: A Study Based on Appraisal System. Foreign Language Education, 33(02), 22-26.