Writer-reader interaction in L2 learner academic discourse: Reader engagement in Czech students’ Master’s theses

Writer-reader interaction in L2 learner academic discourse: Reader engagement in Czech students’ Master’s theses Olga Dontcheva-Navrátilová (Masaryk University, Brno) ABSTRACT This article investigates writer-reader interaction in L2 (Czech) learner academic discourse with aspecific focus on reader engagement in English-medium Master’s theses in the humanities. The study draws on Hyland and Jiang’s (2016) model of engagement. It aims to reveal how Czech graduates use features of engagement (such as reader reference, appeals to shared knowledge, directives and questions) to establish solidarity with readers by acknowledging their presence and negotiating potential alternative views. The contrastive corpus-based analysis compares acorpus of Czech English-medium Master’s theses with two reference L1 corpora representing learner and published academic discourse to explore the impact of linguacultural background, expertise and discipline on the frequency of use and functions of engagement markers. The findings suggest notable variations in the realization patterns and functions of engagement markers across the corpora. It is observed that Czech graduates tend to underuse reader reference and questions, while overusing directives. Moreover, they generally struggle to approximate disciplinary patterns of engagement markers. This trend might reflect students’ limited awareness of academic rhetorical conventions, their efforts to blend L1 and L2 academic norms, and the unique context of addressing an audience within the examination framework of the Master’s thesis. KEYWORDS academic discourse, intercultural rhetoric, Master’s thesis, metadiscourse, reader engagement DOI https://doi.org/10.14712/18059635.2023.2.2 1 INTRODUCTION Academic writing is ahighly specialised use of language which aims to present new knowledge, negotiate social roles and persuade readers to accept the writer’s claims and views. In order to succeed in this persuasive endeavour, academic writers need to construct acredible persona and develop an appropriate relationship with their readers via the use of various rhetorical devices, typically referred to as interactional metadiscourse (Hyland 2001; 2005b; 2008). Interactional metadiscourse encompasses stance resources, i.e. attitudinal, writer-oriented features casting the writer’s voice into the text and expressing his/her judgements, opinions and commitments, and engagement resources, i.e. alignment, reader-oriented features recognising the readers’ presence, involving the audience in the argument and guiding them to intended interpretations (Hyland 2008: 5). Learning to use interactional metadiscourse appropriately is amajor challenge for university students as they are socialised into an academic discourse community. This task is particularly demanding for L2 students OPEN ACCESS 118 LINGUISTICA PRAGENSIA 2/2023 writing in English, as apart from learning to build an academic argument, they are writing in alanguage different from their mother tongue and accommodating Anglophone academic writing conventions, which are bound to differ from those of their original academic literacy. As aresult, the rhetorical strategies L2 learners opt for to convey authorial stance and engage with readers tend to differ considerably from those employed by L1 British students and published Anglophone academic research (cf. Hyland 2002a; Park and Lee 2022). Previous studies into interactional metadiscourse features in L2 learner academic writing are primarily quantitatively oriented. They compare tendencies in the frequency of use of the whole spectrum of interactional metadiscourse markers between L2 learners’ texts and international journal articles (Liu and Zhang 2022), between L1 and L2 Master’s theses (Lee and Casal 2014), between high-rated and low-rated learner essays (Ho and Li 2018; Lee and Deakin 2016), and in successful L2 learner writing across disciplines (Yoon and Römer 2020). Some investigations have explored stance markers, focusing primarily on self-mention (Çandarli, Bayyurt and Marti 2015; Samraj 2008; Zhao and Liu 2021) and hedges and boosters (Malá 2022; Qiu and Ma 2019; Wang and Zeng 2021; Wu and Paltridge 2021). However, engagement has attracted considerably less scholarly attention, and very few studies explore functions of different types of engagement markers (reader-reference, appeals to shared knowledge, questions and directives) in learner discourse. Even so, previous research into published academic discourse (e.g. Dontcheva-Navrátilová 2021; Hyland 2001; Hyland and Jiang 2016; Myers 1989) has shown convincingly that engagement features reflect politeness considerations and play akey role in enhancing persuasion in academic writing. For instance, Swales et al. (1998) and Hyland (2002c) have noted the rhetorical potential of directives to engage readers and direct their attention to aspects relevant to the progress of the argument. Hyland (2002b) has indicated that questions highlight the dialogic nature of academic writing as they allow writers to invoke explicitly the involvement of their readers in the discourse, addressing the perceptions, interests and needs of apotential audience. Several studies have shown that by binding together the writer and the reader as members of the same community the inclusive we creates afeeling of agreement, collegiality and solidarity and thus increases the persuasiveness of the text (Harwood 2005; Hyland 2008; Vassileva 1998). This points to aneed for investigation into engagement markers in L2 learner academic discourse in order to deepen our understanding of the rhetorical development of L2 student writers, so allowing us to assist them more effectively in their efforts to make their claims and arguments more convincing. This study explores engagement features in L2 (Czech) learner discourse as represented by Master’s theses in the field of humanities. Engagement in English-medium discourse by Czech graduates is anew subject of study. The genre of the Master’s thesis was chosen as it is “the most sustained and complex piece of academic writing” that graduate students undertake (Swales 2004: 99) and marks an important milestone in the process of students’ partial socialisation into the academic writing conventions of their field. The purpose of the investigation is to study engagement in Czech students’ English-medium Master’s theses along three dimensions of analyOPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 119 sis: realisations, rhetorical functions and disciplinary variation. Previous research into individual engagement categories in Czech academic discourse (Sudková 2012; Dontcheva-Navrátilová 2020) and in English-medium research articles by Czech authors (Dontcheva-Navrátilová 2021, Kozubíková Šandová 2019) has shown that Czech authors tend to overuse or underuse some categories of engagement markers in comparison to L1 English writers. This has to alarge extent been explained by differences between Czech academic discourse, which is characterised by apreference for awriter-oriented, depersonalised style (Čmejrková and Daneš 1997; DontchevaNavrátilová 2014), and Anglophone academic conventions, which favour amore reader-friendly attitude and ahigher level of interactivity (Thompson 2001; Hyland 2005b). Yet Czech graduates’ choices of engagement markers are likely to be affected not only by linguacultural variation, but also by the level of their socialisation into disciplinary academic writing conventions. The aim of this study is to contribute to intercultural rhetoric research by exploring the influence of linguacultural background and level of expertise on the frequency of use, realisations and functions of engagement resources in Czech graduates’ Master’s theses. Metadiscourse categories will be compared across three disciplines— linguistics, literary studies and English language teaching (ELT) methodology— to establish whether students show awareness of the conventions for the use of these rhetorical features in their own discipline. This investigation combines quantitative and qualitative analysis to answer the following research questions: 1. What are the differences and similarities in the frequency of use, realisations and functions of engagement markers in English-medium L2 (Czech) Master’s theses, L1 argumentative essays and L1 published research articles? 2. Is there variation in the use of engagement markers in academic texts representing the fields of linguistics, literary studies and ELT methodology? 2 METHOD AND DATA 2.1 ANALYTICAL FRAMEWORK The analysis of engagement is based on Hyland’s (2005a) interactional metadiscourse framework. It draws on the modified model of engagement suggested by Hyland and Jiang (2016) which classifies engagement markers into two broad categories according to how they modulate the distance between the participants in the interaction. The first category, ‘proximity’, refers to rhetorical devices creating asense of alliance and solidarity between the writer and the audience. The second category, ‘positioning’, comprises rhetorical devices helping the writer to create involvement with readers, suggests intended interpretations and anticipates potential objections and alternative views. The proximity category comprises two types of engagement markers: — Reader mentions are typically realised by first-person inclusive pronouns and possessives (we, us, our, one), second person pronouns and possessives (you, your), OPEN ACCESS 120 LINGUISTICA PRAGENSIA 2/2023 and the impersonal forms one(’s) and (the) reader. They indicate reader involvement by representing the reader as adiscourse participant sharing common values, views and positions with the writer. This study considers the two functions of reader reference identified by Dontcheva-Navrátilová (forthcoming): (1) community member, i.e. the writer refers to established practices and knowledge shared within agroup, and (2) the ‘reader-in-the text’ (Thompson and Thetela 1995), i.e. the reader is positioned as an interested colleague following the argument presented in the text. — Personal asides are writer comments, commonly included in parenthesis, which represent the writer’s subjective views on what has been said, thus creating abond between the writer and the reader. However, personal asides are excluded from this study due to their low occurrence in academic discourse in general and learner writing in particular (cf. Zhao 2012; Yoon and Römer 2020) and atendency to overlap with self-mention. The positioning category encompasses three types of engagement markers: — Appeals to shared knowledge assume in-group membership presupposing shared previous knowledge of circumstances, processes, practices and beliefs, which helps writers to seek the reader’s agreement with suggested claims and opinions. They comprise three categories: logical reasoning (e.g. of course, obviously), routine conditions (e.g. normally, routinely), and familiarity with tradition (traditionally, commonly). — Directives convey direct address to the reader realised by imperatives (suppose, note), obligation modals (e.g. must, should) and predicative adjectives (e.g. it is important to). They instruct readers to perform three types of acts (Hyland 2002c): (i) textual acts invite readers to refer to another text or adifferent part of the text; (ii) physical acts give instructions to perform an action in the real world; and (iii) cognitive acts guide readers towards an intended interpretation of the argument. — Questions open adialogue with the reader; their rhetorical function is to focus the reader’s attention on the focal points in the writer’s argument. The analysis of the selected four engagement features (reader reference, appeals to shared knowledge, directives and questions) in this study combines quantitative and qualitative methods. Quantitative analysis was used to identify the frequency of occurrence of engagement markers across the three corpora. Qualitative analysis was employed to establish the context-specific rhetorical functions of the various types of engagement markers. 2.2 CORPORA This investigation into engagement features in L2 academic discourse is carried out on aspecialised learner corpus comprising 48 English-medium Master’s degree theses (MT corpus) by Czech university graduate students majoring in English Language and Literature at the Faculty of Arts or Faculty of Education at Masaryk University in Brno, Czech Republic. All Master’s theses were defended in the period 2010–2018, OPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 121 the L1 of their authors is Czech and they were graded ‘A’ (‘Excellent’/‘Merit’), which allows me to assume that they represent ahigh standard of Czech graduates’ writing. The MT corpus represents equally three disciplines: linguistics, literature studies, and ELT methodology (16 theses per discipline). The texts of Master’s theses were cleaned, i.e. all citations, examples, tables, charts and references were removed, so that only the author’s own discourse was subjected to linguistic analysis. The modified corpus totals 948,000 words; the individual disciplines are represented as follows: linguistics: 260,000, literature studies: 328,000, ELT methodology: 360,000. Two reference corpora were used for the purposes of the contrastive analysis. The first reference corpus— aselection from the British Academic Written English Corpus (BAWE_LLC)— was used to identify cross-linguacultural differences in learner academic discourse (native British-English students vs non-native Czech students). The second reference corpus— the Research Articles Corpus (RA_LLM)— comprises research articles published in peer-reviewed international academic journals by Anglophone authors (judging from their names and affiliations). Contrastive analysis of the MT and RA_LLM corpus was employed to reveal variation along the expertise dimension (novice writers vs experienced writers). To ensure comparability, maximally similar criteria were applied in the compilation of all three corpora following the principles of tertium comparationis (Connor and Moreno 2005). The texts of the reference corpora have undergone the same cleaning procedure as the MT corpus. The BAWE_LLC corpus comprises 197 assignments graded ‘Merit’ and ‘Distinction’ written by British university students in the period 2004–2007. Since only students with L1 British English were selected, these essays can be seen as representing L1 learner discourse at asimilar level of proficiency as that of Czech Master’s theses writers. Comparability between the MT and the BAWE_LLC corpus along the discipline dimension was ensured by selection of 197 argumentative essays representing similar disciplines as in the MT corpus, i.e. linguistics, English literature and comparative American studies. Thus, the BAWE_LLM corpus, which totals 491,000 words, may be seen as an ‘analogue’ corpus, i.e. acorpus which is as near as possible in terms of genre and discipline (Tribble 2002; Flowerdew 2015) to the MT corpus. The RA_LLM corpus includes single-authored research articles by Anglophone scholars published in the period 2010–2018. Comparability with the MT corpus was ensured by selection of 36 articles from journals representing the same disciplines as those included in the MT corpus (12 articles per discipline, four per journal): for linguistics, Discourse & Communication, Journal of Pragmatics and Applied Linguistics; for literature, Eighteenth-Century Fiction, New Literary History and SEL: Studies in English Literature; and for ELT methodology, Language and Education, Language Teaching Research and Language Learning. Despite the differences in genre, length and intended audience, research articles and Master’s theses seem to show “significant areas of overlap in lexico-grammar and rhetorical functions” (Flowerdew 2015: 60); thus, the RA_LLM corpus may be seen as an adequate reference corpus for the purposes of this research. The overall corpus wordcount is 243,000 words, split across the disciplines as follows: linguistics: 77,000, literature studies: 83,000, ELT methodology: 83,000. The composition of the learner MT corpus and the two reference L1 corpora is presented in Table 1. OPEN ACCESS 122 LINGUISTICA PRAGENSIA 2/2023 Corpus Texts Wordcount Disciplines MT corpus 48 948,000 linguistics, literature studies, ELT methodology BAWE corpus 197 491,000 linguistics, literature studies, comp. American studies RAs corpus 36 243,000 linguistics, literature studies, ELT methodology Table 1.Composition of the MT, BAWE_LLC and RA_LLM corpora The SketchEngine software (Kilgarriff et al. 2004) was used to build and annotate the corpora. As is common in contrastive corpus-based studies, differences in wordcount between the three corpora were resolved by normalisation to occurrences per 10,000 words. While differences in size and genre between the MT and the reference corpora might have acertain impact on the results, Ibelieve that the analysis will reveal tendencies that could be verified by future research. 2.3 PROCEDURE The SketchEngine corpus tool was used to search for occurrences of engagement markers, i.e. reader reference, appeals to shared knowledge, directives and questions. The hits in all three corpora were checked in context to verify that they perform the target metadiscourse function. It was verified that only inclusive first-person plural pronominal forms (we, our, us) were taken into consideration as reader reference markers (cf. Harwood 2005). Obligation modals (must, should, need, have to, ought to) were also considered in context to verify that they perform the function of directives. As to questions, close reading was necessary to eliminate research questions which do not function as metadiscourse markers. All engagement markers were coded according to the taxonomies of realisations and functional classifications as presented in Section 2.1 above. While comparison of realisation types was carried out across the three corpora, analysis of cross-disciplinary differences was limited to the MT and RA_LLM corpora, as the BAWE selection comprises similar, but not identical disciplines. The statistical significance of differences was determined by application of the non-parametric log-likelihood statistical test (Rayson et al. 2004), where the significance level was set at the standard value of <0.05 (very low p-values are represented as <0.001). 3 RESULTS AND DISCUSSION The results of the quantitative analysis of engagement features across the three corpora are summarised in Table 2.1 The significance of the differences between the corpora as identified by the log-likelihood statistical test are presented in Table 3. The results show that overall variation in the frequency of occurrence of engagement features is significant across the MT and the RA_LLM corpora (p-value <0.001) and the RA_LLM and the BAWE_LLC corpora (p-value <0.001), while the difference between 1 Starting from Table 2 on, the terms ‘raw’ and ‘pttw’ stand for absolute and relative frequencies (per ten thousand words). OPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 123 the MT and BAWE_LLC corpora is not statistically significant (p-value 0.09). This suggests that in academic discourse level of expertise seems to be amore important variable than linguacultural background. Two engagement categories— reader reference and directives— show significant variation across all three corpora. Reader reference shows the lowest frequency of occurrence in the MT corpus, which may stem from advice to opt for impersonality and objectivity as provided by academic writing instruction and style manuals (e.g. Bennett 2009). Other factors affecting the less frequent use of reader reference by Czech graduates may include their rhetorical immaturity and the specificity of writer-reader interaction in the examination context of the Master’s thesis, where the primary reader is an examiner, whom students may not easily position as an ingroup member sharing the same values and understandings. Unlike novice writers, expert writers rely extensively on inclusive we to build aproximity relation with the reader, which is confirmed by the very high frequency of reader reference in the RA_LLM corpus. The BAWE corpus is somewhere between the MT and the RA_LLM corpora in terms of frequency of reader reference, as L1 English novice writers differ from L1 publishing scholars in terms of expertise and from MT writers in terms of linguacultural background. By contrast, directives are most frequent in the MT corpus, while the BAWE corpus shows by far the lowest rate. As the subsequent analysis will show, the high rate of directives in the MT corpus is due to alarge extent to use of the textual directive see to indicate intertextual reference. This seems to be asign of transference from Czech academic discourse, where the use of the imperative viz (see) in parenthetical refEngagement markers MT corpus BAWE_LLC corpus RA_LLM corpus raw pttw raw pttw raw pttw Reader reference 1,489 15.7 1,087 22.1 650 26.7 Shared knowledge 1,770 18.6 842 17.0 418 17.1 Questions 87 0.9 137 2.8 126 5.1 Directives 1,205 12.6 393 8.0 271 11.2 Total 4,551 47.8 2,459 50.1 1,465 60.1 Table 2.Frequency of engagement categories in the MT, BAWE_LLC and RA_LLM corpora Corpora MT vs BAWE MT vs RA BAWE vs RA LL-G2p-value LL-G2p-value LL-G2p-value Reader reference 72.5116 <0.001 118.9419 <0.001 14.3343 <0.001 Shared knowledge 4.1749 0.04 2.3140 0.128 0.0026 0.95 Questions 1.0341 0.30 152.1923 <0.001 24.6126 <0.001 Directives 68.2377 <0.001 1293.638 <0.001 17.2357 <0.001 Engagement markers 2.8607 0.09 55.5445 <0.001 31.2252 <0.001 Table 3.Significance of differences across the corpora (log-likelihood statistical test) OPEN ACCESS 124 LINGUISTICA PRAGENSIA 2/2023 erences to previous research is standard practice (cf. Dontcheva-Navrátilová 2021). The low frequency of directives in BAWE_LLC may stem from the less frequent use of interand intra-textual reference, as well as from the limited length of the text, where the need to guide the reader through the argumentation chain is less pressing. These results differ somewhat from the findings of Lee and Casal’s (2014) analysis of reader reference and directives in English and Spanish theses in engineering: they found that Spanish graduates use these engagement features significantly more frequently than L1 English authors. Following Lafuente-Millán (2014), Lee and Casal explain this intercultural difference by apreference on the part of Spanish authors for positive politeness strategies, in contrast to apreference on the part of Anglophone authors for negative politeness. Moreover, the divergence between my results and those of Lee and Casal (2014) may be due to disciplinary variation. The second positioning feature, questions, shows significant variation when the RA_LLM corpus is compared to the learner corpora, where its incidence is rare. This suggests that this explicitly dialogic feature allowing writers to bring key points of the argument to the attention of the reader tends to be associated with rhetorical maturity and expertise. Due to their relatively low rate and lack of functional variation, questions will not be discussed in the subsequent detailed analysis of engagement features. The only engagement marker that does not show significant variation across any of the corpora is appeals to shared knowledge. It is the most frequent type of engagement marker in the MT corpus and the second in frequency in the L1 corpora. This highlights the centrality of this positioning feature in academic discourse, as it allows writers to assume shared knowledge, values and beliefs with the reader and thus to seek acceptance for suggested arguments and claims. The lack of differences across the corpora may also be due to the relatively small number of adverbs, adjectives and set phrases that typically realise shared knowledge appeal, which helps learners to notice and use them. In the following sections Iwill present and discuss consecutively the results of the contrastive analysis of the engagement categories included in my analytical framework. 3.1 READER REFERENCE Acontrastive analysis of reader reference features indicates that in all the corpora proximity is established primarily by the inclusive we in the agentive subject position and the possessive form our (Table 4). However, L1 writers use this strategy reflecting politeness considerations (Myers 1989; Harwood 2005) considerably more extensively than Czech graduates— in the case of expert authors the inclusive we has twice as many instances than in the MT corpus. This difference is not only quantitative. As Table 5 shows, the corpora also differ in the distribution of the functions of inclusive we: in the MT corpus the community member and the reader-in-thetext functions are equally represented (49.6% vs. 50.4%), while in the L1 corpora the reader-in-the-text function is more prominent (77.5% in the RA_LLM corpus and 63% in the BAWE_LLC corpus). OPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 125 Reader mention MT corpus BAWE_LLC corpus RA_LLM corpus raw pttw raw pttw raw pttw we/our/us 1,005 10.6 889 18.1 537 22.1 you/your 47 0.5 54 1.1 24 1.0 one/reader 437 4.6 144 2.9 89 3.6 Table 4.Reader reference features across the sub-corpora (pttw) Inclusive we/us/our Community member Reader-in-the-text raw %raw % MT corpus 499 49.6 506 50.4 BAWE_LLC corpus 329 37.0 560 63.0 RA_LLM corpus 121 22.5 416 77.5 Table 5.Rhetorical functions of inclusive we/us/our across the sub-corpora (in per cent) The more extensive use of the community member function, i.e. the less powerful function (cf. Tang and John’s (1999) representative role), suggests that Czech graduates tend to create the feeling of solidarity and agreement with the reader by claiming shared in-group membership; however, the group with which the reader is invited to identify is typically fuzzy and general, such as people in general (1), readers of literary texts or speakers of English. This is also indicated by the most frequent collocates of our, which in the MT corpus are rather general, e.g. minds, culture, everyday language, experience, understanding, while in the RA_LLM corpus they are primarily research oriented, e.g. data, research, research questions/findings, understanding, work. (1) Sure, we experience our memories as limited— we do not and physically cannot remember everything that flows through our consciousness, but this is not because there is acertain space that can “house” only agiven amount of content. (MT_LIN_04) The use of the second person (you, your) and indefinite reference forms (one, reader) to refer to people in general, as in (2) and (3), has asimilar function, although general reference forms lack the solidarity dimension prominent in the inclusive firstperson forms. Similarly to experienced writers, Czech graduates tend to use considerably more indefinite reference forms than second-person forms, thus showing awareness of the expected level of formality in academic writing. An additional reason for the low incidence of second-person pronouns may be their potential to imply adistance between the reader and the writer rather than creating abond between them (Hyland and Jiang 2016; Dontcheva-Navrátilová 2020). The BAWE_LLC corpus displays the lowest rate of indefinite forms and the highest rate of second-person pronouns, which seems to indicate alower degree of formality. This may stem from the composition of the corpus, as apart from graduate assignments it comprises the work of undergraduate students, who might lack full awareness of academic writing conventions. OPEN ACCESS 132 LINGUISTICA PRAGENSIA 2/2023 linguacultural background and level of expertise dimensions, overall level of expertise seems to be amore important variable than linguacultural background. Czech graduates’ Master’s theses employ the whole spectrum of engagement markers; however, it was found that they tend to underuse or overuse significantly most types of engagement resources in comparison with the L1 English writers. The only category of engagement which Czech students employ similarly to L1 writers is appeals to shared knowledge, which attests its key role in establishing solidarity with readers. Reader reference and directives are the engagement categories showing the most significant differences. Czech graduates tend to underuse reader reference, which may stem from an effort to display impersonality and objectivity as aresult of academic writing instructions combined with the influence of the distanced style of writing characteristic of Czech academic discourse (Čmejrková and Daneš 1997). The overuse of directives may be tentatively interpreted as a‘compensation’ strategy students use to enhance reader engagement by positioning the reader as an in-group member sharing the writer’s assumptions and beliefs. Overall, it can be concluded that Czech graduates prefer to rely on positioning features and underestimate the role of proximity as they strive to make their discourse more persuasive. An additional factor affecting students’ choices might be the examination context in which the Master’s thesis is set, as the primary reader of the text is an examiner, whom the students might struggle to position as an in-group member sharing the same values and understandings. The analysis has also shown that the students have an awareness of disciplinary conventions in academic writing as the use of engagement resources varies across the linguistics, literary studies and ELT methodology theses. However, ways in which Czech students use engagement resources differ considerably from their use in L1 English research articles in the same disciplines. While linguistics Master’s theses overuse all the categories of engagement, probably as adevelopment stage in the acquisition of these rhetorical devices, literary studies and ELT methodology underuse by far the most of the engagement categories. This indicates that raising students’ awareness of engagement resources as one of the linguistics aspects of academic writing may help Czech graduates improve their rhetorical skills. To conclude, this study has shown that writer-reader interaction in L2 Master’s theses is affected by complex contextual factors, comprising the students’ original literacy, Anglophone writing conventions, previous writing experience and academic writing instruction, as well as the intended audience of the text. The findings of this study may inform the design of resources that can be used in university writing courses to draw the attention of graduate students to the use of engagement resources across disciplinary and cultural contexts. Of course, the results of this investigation should not be generalised, as they take into consideration only one L2 linguacultural background and three disciplines. They should be verified by further intercultural rhetoric studies in different linguacultural contexts and disciplines. Acknowledgement This work was supported by the Czech Science Foundation grant 21-12150S Intercultural variation in writer-reader interaction in English-medium academic discourse by Czech and Anglophone novice writers. OPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 133 REFERENCES Bennett, K. (2009) English academic style manuals: Asurvey. Journal of English for Academic Purposes 8, 43–54. https://doi. org/10.1016/j.jeap.2008.12.003. Brown, G. and S.C.Levinson (1987) Politeness. Some universals in language usage. Cambridge: Cambridge University Press. Çandarli, D., Y.Yasemin Bayyurt and L.Marti (2015) Authorial presence in L1 and L2 novice academic writing: Cross-linguistic and crosscultural perspectives. Journal of English for Academic Purposes 20, 192–202. http://dx.doi. org/10.1016/j.jeap.2015.10.001. Čmejrková, S. and F.Daneš (1997) Academic writing and cultural identity: The case of Czech academic writing. In: Duzsak, A. (ed) Culture and Styles of Academic Discourse, 40–62. Berlin: Mouton de Gruyter. https://doi. org/10.1515/9783110821048.41. Connor, U. and A.Moreno (2005) Tertium comparationis: Avital component in contrastive rhetoric research. In: Bruthiaux, P., D.Atkinson, W.Eggington, W.Grabe and V.Ramanathan (eds) Directions in applied linguistics: Essays in honor of Robert B.Kaplan. Multilingual Matters, 153–164. England: Clevedon. Dontcheva-Navrátilová, O. (2014) The changing face of Czech academic discourse. In: Bennett, K. (ed) The Semiperiphery of Academic Writing, 39–61. Basingstoke: Palgrave Macmillan. https://doi.org/10.1057/9781137351197_3. Dontcheva-Navrátilová, O. (2016) Rhetorical functions of citations in linguistics research articles: Acontrastive (English-Czech) study. Discourse and Interaction 9/2, 51–74. Dontcheva-Navrátilová, O. (2020) Persuasion in academic discourse: Metadiscourse as ameans of persuasion in Anglophone and Czech linguistics and economics research articles. In: Persuasion in specialised discourses. Basingstoke: Palgrave Macmillan. https://doi. org/10.1007/978-3-030-58163-3_3. Dontcheva-Navrátilová, O. (2021) Engaging with the reader in research articles in English: Variation across disciplines and linguacultural backgrounds. English for Specific Purposes 63, 18–32. https://doi.org/10.1016/j. esp.2021.02.003. Dontcheva-Navrátilová, O. (forthcoming) Academic writing conventions in Czech English-medium linguistics journals: Continuity and change over the last 30 years. Flowerdew, L. (2015) Using corpus-based research and online academic corpora to inform writing of the discussion section of athesis. Journal of English for Academic Purposes 20, 58–68. https://doi.org/10.1016/j. jeap.2015.06.001. Harwood, N. (2005) ‘We do not seem to have atheory… The theory Ipresent here attempts to fill this gap’: Inclusive and exclusive pronouns in academic writing. Applied linguistics 26, 343–375. https://doi.org/10.1093/ applin/ami012. Ho, V. and C.Li (2018) The use of metadiscourse and persuasion: An analysis of first year university students timed argumentative essays. Journal of English for Academic Purposes 33, 53–68. http://dx.doi.org/10.1016/j. jslw.2016.06.004. Hyland, K. (2001) Bringing in the reader: Addressee features in academic articles. Written Communication 18/4, 549–74. https:// doi.org/10.1177/0741088301018004005. Hyland, K. (2002a) Authority and invisibility: Authorial identity in academic writing. Journal of Pragmatics 34, 1091–1112. https://doi. org/10.1016/S0378-2166(02)00035-8. Hyland, K. (2002b) What do they mean? Questions in academic writing. Text 32/4, 529–557. https://doi.org/10.1515/ text.2002.021. Hyland, K. (2002c) Directives: Argument and engagement in academic writing. Applied Linguistics 23/2, 216–239. https://doi. org/10.1093/applin/23.2.215. Hyland, K. (2005a) Metadiscourse: Exploring interaction in writing. London: Continuum. Hyland, K. (2005b) Stance and engagement: Amodel of interaction in academic discourse. OPEN ACCESS 134 LINGUISTICA PRAGENSIA 2/2023 Discourse Studies 7/2, 173–192. https://doi. org/10.1177/1461445605050365. Hyland, K. (2008) Persuasion, interaction and the construction of knowledge: Representing self and others in research writing. IJES 8/2, 1–23. https://doi. org/10.6018/ijes.8.2.49151. Hyland, K. and K.Jiang (2016) “We must conclude that…”: Adiachronic study of academic engagement. Journal of English for Academic Purposes 24, 29–42. https://doi. org/10.1016/j.jeap.2016.09.003. Kilgarriff, A., P.Rychlý, P.Smrž, and D.Tugwell (2004) The Sketch Engine. In: Proc Eleventh EURALEX International Congress. Lorient: France. Kozubíková Šandová, J. (2019) Proměny českého akademického diskurzu během posledních padesáti let. [Changing Czech academic discourse in the past 50 years]. Časopis pro moderní filologii 101, 54–71. Lafuente-Millán, E. (2014) Reader engagement across cultures, languages and contexts of publication in business research articles. International Journal of Applied Linguistics 24, 201–223. https://doi.org/10.1111/ijal.12019. Lee, J.J. and J.E.Casal (2014) Metadiscourse in results and discussion chapters: Acrosslinguistic analysis of English and Spanish thesis writers in engineering. System 46, 39–54. http://dx.doi.org/10.1016/j. system.2014.07.009. Lee, J.J. and L.Deakin (2016) Interactions in L1 and L2 undergraduate student writing: Interactional metadiscourse in successful and less-successful argumentative essays. Journal of Second Language Writing 33, 21–34. http:// dx.doi.org/10.1016/j.jslw.2016.06.004. Li, Q. and X.Zhang (2021) An analysis of citations in Chinese English-major Master’s theses and doctoral dissertations. Journal of English for Academic Purposes 51, 1–14. Liu, G. and J.Zhang, (2022) Interactional metadiscourse and author identity construction in academic theses. Journal of Language Teaching and Research, 13/6, 1313–1323. DOI: https://doi.org/10.17507/ jltr.1306.20. Malá, M. (2022) Hedging like aprofessional: Acorpus-driven approach to interactional metadiscourse in English learner academic writing. In: Hůlková, I., R.Povolná and R.Vogel (eds) Patterns and variation in English language discourse. 9th Brno conference on linguistics studies in English, 72–86. Brno: Masaryk University. https://doi.org/10.5817/ CZ.MUNI.P280-0212-2022-5. Myers, G. (1989) The pragmatics of politeness in scientific articles. Applied Linguistics 10, 1–35. https://doi.org/10.1093/applin/10.1.1. Park, H.I. and S.Lee (2022) Interactional metadiscourse in English teaching articles: Adiachronic perspective (1980–2021). English Teaching 77/2, 3–23. https://doi.org/10.15858 / engtea.77.2.202206.3 Qiu, X. and X.Ma (2019) Disciplinary enculturation and authorial stance: Comparison of stance features among master’s dissertations, doctoral theses, and research articles. Iberica 38, 327–348. Rayson, P., Berridge, D., and B.Francis (2004) Extending the Cochran rule for the comparison of word frequencies between corpora. In: Purnelle, G., C.Fairon and A.Dister (eds) Le poids des mots: Proceedings of the 7th International Conference on Statistical analysis of textual data (JADT 2004), 926–936. Louvain-la-Neuve: Presses universitaires de Louvain. Samraj, B. (2008) Adiscourse analysis of Master’s theses across disciplines with afocus on introductions. English for Academic Purposes 7, 55–67. https://doi.org/10.1016/j. jeap.2008.02.005. Sudková, M. (2012) Vyjadřování interpersonální funkce včeských aanglických odborných textech. Kontrastivní studie. [The expression of interpersonal metadiscourse in Czech and English Academic texts. Acontrastive study]. Časopis pro moderní filologii 94/1, 31–42. Swales, J. (2004). Research Genres. Cambridge: Cambridge University Press. https://doi. org/10.1017/CBO9781139524827. Swales, J., U.K.Ahmad, Y.-Y. Chang, D.Chavez, D.Dressen, and R.Seymour (1998) Consider this: The role of imperatives in scholarly OPEN ACCESS OLGA DONTCHEVA-NAVRáTILOVá 135 writing. Applied Linguistics 19/1, 97–121. https://doi.org/10.1093/applin/19.1.97. Tang, R. and S. John (1999). The ‘I’ in identity: Exploring writer identity in student academic writing through the first person pronoun. English for Specific Purposes 18, 23–39. https:// doi.org/10.1016/S0889-4906(99)00009-5. Thompson, G. (2001) Interaction in academic writing: Learning to argue with the reader. Applied Linguistics 22/1, 58–78. https://doi. org/10.1093/applin/22.1.58. Thompson, G. and P.Thetela (1995) The sound of one hand clapping: The management of interaction in written discourse. Text 15/1, 103–127. https://doi.org/10.1515/ text.1.1995.15.1.103. Tribble, C. (2002) Corpora and corpus analysis: new windows on academic writing. In: J.Flowerdew (ed.) Academic discourse, 131–149. London: Longman. Yoon, H.-J. and U.Römer (2020) Quantifying disciplinary voices: An automated approach to interactional metadiscourse in successful student writing. Written Communication 37/2, 208–244. Vassileva, I. (1998) Who am I/who are we in academic writing? International Journal of Applied Linguistics 8/2, 163–192. https://doi. org/10.1111/j.1473-4192.1998.tb00128.x. Wang, J. and L.Zeng (2021) Disciplinary recognized self-presence: Self-mention used with hedges and boosters in PhD students’ research writing. SAGE Open, 11/2, 1–13. https://doi.org/10.1177/21582440211005454. Wu, B. and B.Paltridge (2021) Stance expressions in academic writing: Acorpusbased comparison of Chinese students’ MA dissertations and PhD theses. Lingua 253, 103071, https://doi.org/10.1016/j. lingua.2021.103071. Zhao, C.G. (2012) Measuring authorial voice strength in L2 argumentative writing: The development and validation of an analytic rubric. Language Testing 30/2, 201–230. Zhao, Y. and J.Liu (2021) Adevelopmental view of authorial voice construction in master’s thesis: Acase study of two novice L2 writers. SAGE Open 11/4, 1–15. https://doi. org/10.1177/21582440211054483. Olga Dontcheva-Navrátilová Department of English Language and Literature Faculty of Education, Masaryk University Poříčí 623/7, 603 00 Brno ORCID ID: 0000-0002-0378-7975 navratilo[email protected].cz OPEN ACCESS