Rhetorical Style Across Cultures: An Analysis of Metadiscourse Markers in Academic Writing of Thai and Malaysian Students

RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 19 LSP International Journal, Vol. 6, Issue 1, 2019, 19–37 © Universiti Teknologi Malaysia E-ISSN 2601–002X DOI: https://doi.org/10.11113/lspi.v6n1.76 Rhetorical Style Across Cultures: An Analysis of Metadiscourse Markers in Academic Writing of Thai and Malaysian Students Faridah Hayisama & Mohamed Ismail Ahamad Shah International Islamic University Malaysia, P.O. Box 10, 50728 Kuala Lumpur, Malaysia Wan Nur Asyura Wan Adnan Language Academy, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia Submitted: 8/11/2018. Revised edition: 18/03/2019. Accepted: 01/04/2019. Published online: 19/06/2019 ABSTRACT It is believed that students from different societies and cultural background have their own preferred rhetorical style of interaction. In writing, such distinctive preference is usually exhibited through the use of linguistic features, of which metadiscourse (MD) markers are considered as one of the signposts to the interpretation of writing style preferences. This study aims to investigate the use of interactional metadiscourse (MD) features and its relevance to the rhetorical style preferences in academic writing of Thai and Malaysian master’s students. Using Hyland’s (2005) taxonomy, their thesis discussions were manually analysed in terms of interactional MD markers to determine their frequency of occurrence and to relate the results to the rhetorical styles of writing preferred by each group of students. The analysis revealed that of all five types of interactional MD features, hedges were the most frequently used device followed respectively by boosters, attitude markers, engagement marker and self-mention. In terms of the rhetorical style of writing, the frequency of MD features suggests that tentative and indirect statements, reader-responsibility, distant-relationship between writer and readers, and less writer-involvement in the texts were the preferred rhetorical styles of interaction of both Thai and Malaysian students. The writing conventions and rhetorical styles of the students can be explained from a sociocultural point of view that they are relevant to the oriental style of interaction from which both groups of students originated. The study draws attention to the pedagogical implications that students in Thai and Malaysia should be given more instructional focus on how to utilize MD features in making academic writing more persuasive and interactive. Keywords: Rhetorical style, cultures, metadiscourse markers 1.0 INTRODUCTION Language and writing are regarded as cultural phenomena with each community having its own rhetorical conventions and different preferred discourse style that shape structures of composition and argumentation to appear differently across cultures (Connor 2002, Hyland 2005). Being direct and assertive in presenting an argument may be favorable to some cultures, while people in other cultures *Correspondence to: Wan Nur Asyura Wan Adnan (email: asyura.kl@utm.my) 20 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan may deem tentative statements that leave some space for readers to agree or disagree as appropriate and rhetorically persuasive. Adel (2006) and Hyland (2004) assert that the style of discourse organization, the degree of rhetorical uncertainty and assertiveness of the writers are associated with the cultural conventions of a particular community. In the same vein, Abdollahzadeh (2011) also suggests “the way writers present themselves, negotiate an argument, and engage with their readers is closely linked to the norms and expectations of the particular cultural and professional communities” (p. 296). The choice of metadiscourse (MD) employment in a particular text is also an indication of assorted preferences resulting from a diversity of people’s socio-cultural perceptions towards the definition of persuasiveness. In the same direction, Hyland (2005) postulates that measuring interactional metadiscourse usage in a text can be one of the ways to study the rhetorical style of writing preferred differently by people of different cultures. He further elaborates that metadiscourse markers such as hedging devices, boosters, or engagement markers are used as indicators of a writer’s interactional strategies in negotiating or persuading the readers to concur with a statement. Similarly, the rhetorical identity behind the influence of L1 culture is also revealed in the discourse of ESL learners. The degree of interpersonal interaction between the writer and readers in L2 writing may vary due to different cultural preconceptions of the writer’s L1 background (Hyland, 2005). Hinkel’s (2002) study was among the extensive comparative studies to prove such a phenomenon. In her study, she compared the use of metadiscourse features and rhetorical styles in English essays written by ESL students of six groups of different national backgrounds. The result of the study revealed that the employment of metadiscourse markers indeed varied from cultures to cultures, despite the fact that all six groups shared mutual oriental style of discourse pattern as their commonality. Although the study did not set the ground on neither the writing convention of Malaysian nor Thai ESL students, nonetheless, it paves ways to the idea that the rhetorical style of writing through the use of metadiscourse markers performed by Thai and Malaysian students could be varied from one another too. A recent study by Loan and Pramoolsook (2016) similarly looks at rhetorical structures in master’s theses but approaching it from Chen and Kuo’s (2012) framework with a focus on Vietnamese students only. In similar vein, Zakaria, (2016) explores the metadiscourse markers in academic essays of Arab and Malaysian students. He identifies that there are differences and forms of metadiscourse markers used by them which could be attributed to the different culture background between the two countries. Metadiscourse carries important social meaning. It significantly implies the author’s personality and identity and at the same time indicates how he expects his readers to respond to his propositions (Toumi, 2009, p.64). In ESL context, studying metadiscourse pedagogically assists teachers to better understand the variation of writing conventions of ESL students from different L1 backgrounds. However, only a limited number of studies in the past have been conducted on metadiscourse marker employment behavior of Malaysian and Thai ESL students particularly in academic texts such as thesis. It was not until very recently that the studies of Chan and Tan (2010), Tan (2012) and Anwardeen, Luyee, Gabriel, and Kalajahi (2013) from a local university in Malaysia were published in the field and with these, metadiscoursal studies in the Malaysian context began to gain attention. However, none of these available studies focuses on cross-cultural factor; rather they are more towards highlighting Malaysian students’ proficiency in using metadiscourse markers in academic RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 21 writing. Meanwhile, studies on metadiscourse marker preference among Thai ESL writers are even more sparing in number. Due to the inadequacy of pedagogical references in this area, this study therefore aims at filling the research gap in intercultural rhetoric studies, as well as to provide ESL teachers a deeper insight into the field of metadiscourse for further intercultural research in the future. Therefore, this study sets forth to find frequency to compare the interactional metadiscourse features employed by Thai and Malaysian students and analyze the differences and discuss the rhetorical styles of interaction and preferences of Thai and Malaysian postgraduate students from sociocultural perspective. Specifically, it aims at answering these two research questions, notably, (i) What are the differences in the interactional metadiscourse features employed by Thai and Malaysian students? and (ii) What are the rhetorical styles of interaction preferred by Thai and Malaysian postgraduate students with relevance to their respective sociocultural backgrounds? 1.1 Theoretical Framework The theoretical framework for this study is based on two well-known theories, namely, Hyland’s (2005) concept of interpersonal metadiscourse and Connor’s (2002) theory of intercultural rhetoric, formerly known as Contrastive Rhetoric. Intercultural Rhetoric theory supports the idea that there are relative ties between writing style and culture (Connor, 2002). It values the notion of culture behind language performance of individuals significantly, particularly in terms of rhetorical preferences that are believed to differ from language to language (Kaplan, 2000 in Hyland 2005). It is also believed that people of different cultural background may have differing expectations about logical organization of written texts and this, therefore, may have an effect on their literacy development in second language acquisition. In other words, culture is said to be one of the factors that shapes our background understandings and the way we weave our language in spoken and written discourse (Hyland, 2005; Rasekh & Amiryousefi, 2010). This theory of Intercultural Rhetoric suggests that the study of discourse linguistics that aims to analyze metatextual features of texts can be one of the ways to approach language studies from a cross-cultural perspective (Connor, 1996). Therefore, the analysis of metadiscoursal strategies is considered as another area to study the cultural significance behind the variety of rhetorical preferences across cultures. In fact, the study of different preferred rhetorical styles through metadiscourse use across cultures is one of the popular areas among intercultural rhetoric scholars (e.g. Crismore et al., 1993; Hinkel, 2002; Abdi, 2009; Mur-Duenas, 2011; Uysal, 2012). In short, the broad scope of theoretical framework to be fundamentally grounded for this study was thus a combination of two theories, i.e. Hyland’s (2005) taxonomy of interpersonal metadiscourse and Connor’s (2002) linguistic approach of intercultural rhetoric theory. Hyland’s MD taxonomy was used as the analytical framework for analyzing the rhetorical styles of interaction. Connor’s theory was referred to in parts where the rhetorical styles are analyzed from a sociocultural point of view. Figure 1.1 depicts how both theories were used to identify the frequency and preferred rhetorical styles and its relevance to sociocultural background of the students. 22 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan Figure 1.1 Theoretical Framework 2.0 METHODOLOGY 2.1 Research Design All linguistic features such as conjuncts, adverbials, or paraphrasing expressions have a potential to carry both propositional and metadiscoursal meanings (Hyland, 2005). Furthermore, each linguistic item can also express both interactive and interpersonal meanings. Since MD is such a ‘fuzzy’ term (Hyland, 2005, p.16) in the sense that we cannot simply label its metadiscoursal quality by its manifested appearance, thus, latent content analysis is needed. Other than employing manifest analysis based on MD item lists suggested by Hyland (2005), this research also requires a latent analysis at sentential co-text level to identify the underlying meaning of those MD items. Apart from referring to Hyland’s MD items as its manifest coding, it is essential to analyze the metadiscoursal quality that each word carries based on the sentential underlying meaning as well. Babbie (2013) suggests that using both manifest and latent content analysis is the best solution to overcome the issue of validity and reliability in conducting content analysis research. In light of that, Hyland’s (2005) taxonomy is, thus, primarily used as the conceptual framework for the analysis. His five types of interactional MD markers were employed as the manifest coding frame of this research. 2.2 Corpus of the Study Material in this study was drawn from two sources. The first group is the Thai master’s students’ theses taken from Prince of Songkhla University’s archive that are available online. The second group is Malaysian master’s students’ theses taken from International Islamic University Malaysia which are available in the form of hard copies. Both groups are English-major students specializing in RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 23 Applied Linguistics. The topics chosen are also of equal issue, i.e., if the selected Thai student’s thesis is on phonetics then the matched thesis from the Malaysian counterpart should be on phonetics as well. It is widely suggested that the discussion and introduction sections of the articles are the most rhetorical parts due to their persuasive and convincing nature of the text (Mauranen, 1993; Faghih and Rahimpour, 2009; Khedri et al., 2010). The discussion part of the thesis is the floor where the writers would normally exhibit their art of crafting words by using the rhetorical style that they deem most persuasive to convince the readers. It is also where the interpersonal interactions between the writer and readers are notably seen. Therefore, the discussion section of master’s theses written by Thai and Malaysian postgraduate students was chosen as the data of this study. However, after collecting samples of data, it was found that not all the available theses consist of discussion section as a separate segment. Some students had combined the discussion part with their findings section—the section which was meant to inform rather than persuade and convince the readers. Since this study seeks for the rhetorical persuasiveness through the use of MD devices, thus, the data selection should be very specific of which only the theses with the discussion section as a separate part from the finding report were chosen. Besides that, due to the dissimilarity in terms of lengths that each student may have written in his/her discussion section, it is decided that the word limit of each discussion was restricted to 1800-2000 words to maintain the validity of the results. Fourteen discussion sections from each group of participants with a total of twenty-eight scripts of discussion segments (approximately 50000-56000 words in total) were used as the corpus of this analysis. 2.3 Subject The nationality of the students was ensured by the name written on the cover page of the thesis. The Malaysian subjects were identified by the arrangement of the name that comprises of ‘bin’ or ‘binti’ and typical Malay names were specifically chosen. Similarly, in the case of Thai subjects, those with typical Thai names were particularly selected to ensure the authenticity of nationality. All participants were ESL students whose L1 background is not English but who are native Thai and Malay, accordingly. 2.4 Data Analysis Since the contextual nature of MD is rather fuzzy in the sense that one particular marker can be interpreted in terms of either propositional or interpersonal meaning, a careful manual search for interpersonal markers was conducted repetitively line by line to ensure they were not wrongly labeled. The propositions in which MD markers exist were highlighted in order to analyze the contextual function of that particular MD marker. This is in response to the suggestion made by Adel (2006) who postulates that metadiscursive expressions are in fact multifunctional and context dependent. Sentences with the occurrence of interactional MD markers were examined at the contextual level to determine their syntactic and pragmatic functions. Hyland’s (2005) interactional MD taxonomy was employed as the instrument for the analysis. Each corpus was scanned manually by careful reading through the text and highlighting where the devices were found. The occurrence of each device was recorded and classified according to each five sub-types, i.e. hedges, booster, attitude markers, self- 24 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan mention, and engagement markers. The number of MD occurrence for each device was then totaled up and contrastive analysis was repeatedly conducted. In this analysis, the frequency of the preferred MD devices employed by Thai and Malaysian students was compared in order to see the similarities and differences in their preferences. This numerical measurement was aimed to fulfill the first aim of the research. The writer’s degree of assertiveness and tentativeness, the writer’s degree of interaction with the reader and his/her level of involvement occurring in the texts were the rhetorical styles of interaction sought to answer RQ 2. 2.5 Analytical Framework This research employed Hyland’s (2005, p.49) interpersonal model of MD as its analytical framework to investigate the metadiscoursal use and preferred rhetorical styles of the two groups of students from different cultural backgrounds. According to Hyland, interpersonal MD is divided into two categories, namely, interactive and interactional under which five types of markers are found in each category. Each marker has an exclusive function which significantly plays its role in written texts in a distinctive way. Interactional markers work as a platform for the writers to express or intrude their personal ideas and attitudes into the propositional text. Each category of MD can signify the rhetorical style of interaction portrayed in the written discourse by the writers. The categories, functions, and examples are presented in the Table 2.1: Table 2.1 A Model of Interactional Metadiscourse CATEGORY FUNCTION EXAMPLES 1. Hedges Withhold commitment and open dialogue Might; perhaps; possible; about In fact; definitely; it is clear that Unfortunately; I agree; surprisingly 2. Boosters Emphasize certainty or close dialogue 3. Attitude markers Express writer’s attitude to proposition 4. Self-mentions Explicit reference to author(s) I; we; my; me; our 5. Engagement markers Explicitly build relationship with reader Consider; note; you can see that (Adapted from Hyland, 2005, p.49) 2.6 Validity and Reliability of the Study The issue of validity and reliability is one of the concerns in this qualitative study. According to Merriam (2009, p.220), a reliable qualitative study may not yield the same result when replicating, but the result it yields should be ‘consistent’ with the data collected. Merriam (2009) further suggests that using multiple methods of data collection or what is known as triangulation strategy can ensure consistency and dependability on the reliability of the study. For the purpose of this study, one of the ways to do so is to have a peer review by means of having other individual field experts to run the RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 25 analysis on the same data. This is in order to avoid any bias or misinterpretation that might have occurred during the data analysis. In response to the suggestion, the researcher in this current study, thus, conducted a peer examination by a lecturer from a Malaysian public university who is an expert in linguistic studies to re-analyze a random portion of the same data and compare the findings. The analysis from peer review was finally revealed that there was a high correlation in the frequency counts of MD markers, which was 90% consistent with what discovered previously by the researcher. Such consistency, therefore, suggested the reliability and validity of this particular study. 3.0 FINDINGS AND DISCUSSIONS The analysis was conducted on 28 thesis discussions of which 14 samples came from each group i.e. Thais and Malaysians. The overall frequency of MD marker occurrence counted by items and percentage is presented in Table 2.2: Table 2.2 Comparison of Frequency of MD Markers in the Thesis Discussions of Thai and Malaysian Students Frequency of occurrence (items) Type of MD features 1. 2. 3. 4. 5. Hedges Boosters Attitude Markers Engagement Markers Self-mention Total of all MD found in text Thais 392 290 148 119 4 953 Malaysians 385 314 154 127 22 1002 Frequency of occurrence (percentage) Thais Malaysians 38.42% 41.13% 30.43% 31.33% 15.36% 15.52% 12.48% 12.67% 0.42% 2.19% 100% 100% 3.1 Findings 3.1.1 The Differences in the Interactional Metadiscourse Features by Thai and Malaysian Students The analysis revealed that there is a slight difference in the frequency of MD occurrence in the discussion texts of Thai and Malaysian students. Hedges were the MD markers found most heavily used in the Thai students’ theses (41.13%), whereas boosters, engagement markers, and self-mention were found slightly more frequent in the Malaysian students’ theses (31.33%, 12.67%, and 2.19% respectively). However, the percentage of occurrence did not reveal much difference from each other. The slightly higher number of hedges in the Thai students’ theses suggests that Thai students were identified to be more tentative in proposing opinions than their Malaysian counterparts. In terms of attitude markers and engagement markers, the frequency revealed no significant difference from each other. Thai students’ theses outnumbered the Malaysians’ very slightly for attitude markers while the latter showed a slightly higher usage of engagement markers. Such similar results indicated that there 26 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan was no remarkable difference between the Thai and Malaysian students when it comes to giving attitudinal comments or catching reader’s attention with metadiscoursal remarks in the discourse. Nonetheless, a significant difference was shown in the frequency of self-mention where Malaysian students were found to have employed this technique relatively more heavily than the Thai students (2.19% versus 0.42%). Some of the examples will be shown in the discussion section. In the aspect of keeping the balance of rhetorical tone when discussing the arguments, Malaysian students were seen to observe the balance of assertive and tentative tone slightly better than the Thai students. This can be noticed from the distinction of gap between hedges and boosters. The discussion texts of Thai students showed a much distinct gap between the use of hedges and boosters (41.13% versus 30.43%), while Malaysian students’ texts showed lesser distinction between hedges and boosters (38.42% versus 31.33%). The smaller distinction between these two markers showed a finer balance of assertive and tentative tone. Therefore, it can be said that the Malaysian students observed a balanced use of MD markers better than the Thai students, which renders a more refined interactional style to their thesis discussions as a result. Despite the fact that Thai and Malaysian students showed prevalent use over each other in each type of markers; the number of percentage, however, showed insignificant difference. That is to say, both groups constituted almost a similar amount of frequency for each type of markers (e.g. 30.43% vs. 31.33% for boosters, 15.52% vs. 15.36% for attitude markers, and 12.48% vs. 12.67% for engagement markers). The comparison between the two groups is presented in Figure 1.2 (See Appendix for individual results). Figure 1.2 Comparison of Frequency of MD Markers in the Thesis Discussions of Thai and Malaysian Students 3.1.2 Discussion Figure 1.2 shows there is not much difference in the use of MD markers between the two groups. It indicates that Thai and Malaysian students shared similar preferences in the use of MD markers in their thesis discussions. As for hedging markers, the result was consistent with the findings of Hyland RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 27 (2004) who found that hedges were the most frequently used devices in postgraduate writings. Both groups preferred to hedge their arguments rather than convince readers with assertive statements. To illustrate better, examples are shown below: Thai Students a) “This could occur in relation to L1 interference in which Thais rather use context and adverbs of time to signify tenses.” b) “One possible reason was that the student did not know the word and might have thought that leaving it out would not affect the main idea of the sentence.” In sample (a) the writer used hedging device such as could to explain a possibility which is not a 100% fact. Meanwhile in the sample (b) hedges such as possible and might were employed as a hedging tool to address the writer’s personal opinion towards the outcome of the study. The hedging technique was used to withhold a full commitment towards the statement being uttered so that if it were to be otherwise, the writer would have a safe space to stand. Another scenario in which hedging markers were seen to be employed was when a recommendation towards a statement was being proposed by the writers, as in (c): c) “Therefore, the teachers should utilize the teaching technique making students relaxed in English class and attentive in English learning.” Herein, should was used as a remark for proposing the writer’s personal recommendation. The epistemic verb should in the sample sentence helps the proposed statement to sound subjective and less assertive. Compared to the epistemic verbs must, should gave a meek tone to the proposed recommendation and made it sound more tentative and less forceful. Malaysian Students a) “It appears that these subjects’ overall scores in the Morphological Awareness Test, to a certain extent, signified a rather low or limited morphological awareness.” b) “It is possible that the proficient users used indirect strategies significantly more than direct strategies probably because they had better knowledge of the language.” c) “Hence, awareness of intonation’s role in communicative intelligibility should be taken seriously.” Hedging markers in sample (a) were used as a means to propose a result of the study. By choosing words that conveyed tentative connotation such as appear and to a certain extent, the writer attempted to leave a space for disagreement that may possibly arise from the reader’s counterpart. This, relatedly, renders the proposition to appear more reliable and as a result, tentatively persuasive. Hedges in sample (b) were seen to have been employed to address a possibility. The words possible and probably were used to signify that the statement is not a 100% fact and it could be arguable. The modal verb should in sample (c) was another hedging device employed to address tentativeness. Compared to other modal verbs such as must, using should to propose a suggestion to the readers made 28 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan the statement sound optional and less forceful. It seemed that the readers were given a choice to decide whether to take the recommendation or otherwise. However, the predominance of hedges over boosters was not consistent with the research finding of Hinkel (2002) who found that ESL students (i.e. Asian students) preferred to employ boosters to strengthen their claims and emphasize the truth of a statement. Thai and Malaysian ESL students in this study were shown to hedge their statements rather than using assertive markers such as boosters to convince readers. However, such discrepancy could possibly be due to the distinction of the students’ language proficiency rather than cultural factor per say. Hinkel’s ESL subjects were reported to be undergraduate students; whereas this study’s ESL subjects were postgraduate students. The level of exposure to English language could be the reason behind such dissimilarity in the findings. As for the case of Malaysian students, the result of this study was also not consistent with Chan and Tan’s (2010) research who found that engagement markers and boosters were the markers most preferred by Malaysian students. Hedging devices were less evident than boosters in Malaysian students’ argumentative essays in Chan and Tan’s study. Again, this could more possibly be due to the factor of student’s proficiency in English and their collective experience in written discourse. The examples are as illustrated below: Thai Students a) “Besides, sample English sentences of a target word in dictionaries can provide learners more clarity about the use of a particular word in context, so that learners can gain more knowledge about choosing proper words to fit in the right contexts.” b) “This was evident in their writing of run-on sentences and lack of sentence and paragraph divisions.” The booster can in sample (a) was employed twice and it boosted the proposed statement to sound even more convincing. The modal verb can in the sentence functioned as the reliability strengthener that gave the readers more sense of confidence towards the proposition. Likewise, boosters such as evident in sample (b) also helped increase credibility to the study outcome. The writer intentionally used the particular adjective as a boosting marker to highlight his confidence towards the result and therefore made the whole discussion sound more convincingly persuasive. Besides that, boosters were also found frequently employed in sentences where the writers brought about other researchers’ statements into the text to affirm and support the congruence of their own findings, as in the sample below: c) “This finding supported those of Hengsakul and Jai-Ngam (1990) that the students in the area of Physical Education had a low level of English proficiency”. The verb support in sample (c) acted as a certainty marker to strengthen the writer’s findings from the study. Herein, other researchers’ study was used to enhance the truth-value of the text with the help of a booster marker to make the statement sound more assured. RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 29 Malaysian Students a) “The same situation can also be seen in well-formed onsets of the novel words like ‘styib’.” b) “The above examples discuss Islamic terrorism and the way it is reported obviously shows that it has caused trouble and fear.” Boosters in sample (a) and (b) were strategically used to affirm the results of the study. The writers deliberately chose epistemic verb can to strengthen the result and denote a high level of certainty. Similarly, adverb obviously in sample (b) also helped to boost the confidence of the readers towards the statement. It explicitly showed that the writer was very certain with the outcome of the study and so should the readers. Another observable difference in the interactional MD markers employed by Thai and Malaysian students is the use of self-mention. The majority of Thai students (11 out of 14 students) did not use self-mention as the interactional marker in their discussions at all. This implied that presenting the writer’s identity in the arguments was rarely practised by the Thai students. As opposed to their counterparts, self-mention was commonly found to be employed by many Malaysian students in their thesis discussions. This echoes the findings of Loi and Lim (2019) whereby they posit that selfmention was one of the most frequently employed hedges among Malaysians’ in their academic writing. This indicates that, for Thai students, self-presenting in the arguments was not considered a common trend although a number of current studies (Hyland, 2004; Faghih, 2009; Mur-Duenas 2011) observed that self-mention has become a common practice among academic writers these days. Both of the examples from Thai and Malaysian students’ use of self-mention are illustrated below: Thai Students a) “Surprisingly, the high proficiency students used non-verbal language strategy more often than those of the low proficiency. This was in contrast to the researcher’s expectation” b) “This lack of distinction in the use of cohesive devices between the good and poor essays encourages the researcher to make three points of discussion.” The writers identified themselves in the discussion as the researcher instead of overtly using the first pronoun I. A possible reason to explain the avoidance of self-mention is that the writers may have attempted to make the texts sound less conversational. The distance between the writer and readers was thus strictly observed. This result coincides with what Scollon (1994 cited in Hyland 2005) pointed out that Asian students preferred to avoid self-mention in the text in order to disguise the direct involvement and their views as a writer. Presenting themselves in the texts was seen as uncommon among Thai student writers. As a result, the number of self-mention employment was found rare in practice. 30 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan Malaysian Students a) “Therefore, in this study, the researcher considers avoiding difficult words, sentences or concepts as topic avoidance.” b) “Thus, in general, we could say that the anxiety level was quite alarming.” Self-mention in sample (a) was marked by the writer addressing herself as the researcher. Instead of using the personal pronoun I the writer chose to depersonalize herself by presenting her status as ‘the researcher’. Hyland (2005) explained that the common purpose of adopting less personal writing by the writers in academic discourse is that they prefer to remove themselves from exerting an influence on the finding of the study, and so using personal pronoun is avoided. In sample (b), on the other hand, the use of first pronoun we in the sentence was seen, which could possibly signify that the writer may have meant to highlight her existence and involvement in the discourse. It, thus, seems that selfmention technique in this context could somewhat render a sense of interactiveness and persuasiveness more to the discussion. Both groups of students (i.e. Thai and Malaysian) showed a predominant use of hedging markers over boosters in their thesis discussions. The frequency of attitude markers and engagement markers also showed to differ very slightly from one another. Self-mention was reported to be the least employed type of interactional MD markers. The ranking of these five types of interactional markers was seen equivalent in both Thai and Malaysian students. This phenomenon hence indicates that there is not much difference in terms of MD marker preference in the thesis writings of Thai and Malaysian students. 3.1.4 Discussion Many studies (e.g. Abdi, 2002; Milne, 2003; 2008; Faghih & Rahimpour, 2009; Noorian & Biria, 2010 and Kashiha, 2018) discovered that hedges were the markers mostly preferred by ESL writers. Similarly, Thai and Malaysian students in this study were prone to use hedges as the primary interactional MD markers in their thesis discussions. Abdollahzadeh (2011) affirms that hedging statements were commonly used in academic writing in attempts to make the conveyed message sound more like an opinion which is open for discussion rather than a solid fact. Instead of presenting statements in a clear-cut and affirmative manner, both Thai and Malaysian students in this study preferred to hedge their arguments in order to make them sound more subjective. In relation to that, Hyland (2005; 2010) also suggests that the frequency of hedging markers is an indication of the writer’s degree of tentativeness. Since the use of hedges is evident in the Thai and Malaysian students’ writings, it could be said that one of the rhetorical styles of interaction preferred by both Thai and Malaysian students in this study was to make the arguments sound more tentative rather than assertive in manner. Hedges mitigate the certainty level of an argument, and they allow readers to present any possible refutations or disagreements toward the texts. In this study, the tentative style of writing through the prevalent use of hedges practiced by Thai and Malaysian students was deemed to require involvement from the reader as an active participant in the discussions. Readers were offered a chance to have RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 31 alternative voices or viewpoints that may differ from the proposed argument or information in the discourse. As such, commitment from the writer’s side towards the statements being proposed in the text was thus relatively allocated to the readers’ own interpretation as well. The fact that writers keep a distant from the text through the use of hedging MD markers proved their writing style to be tentative, as a result. From this point of view, it is agreeable to say that one of the rhetorical styles preferred by Thai and Malaysian master’s students in writing thesis discussions was to present the statements in a tentative and indirect manner. Also, the fact that readers were allowed to take charge of construing the texts at their own costs therefore allowed their written discourse to appear to be “reader-responsible”. There are a few possible reasons to explain the frequent occurrences of hedges in the thesis discussions of Thai and Malaysian master’s students. On the one hand, it is due to the fact that hedges are a common device in academic writing as Hyland (2004, 2005) suggested. On the other hand, however, it could also be due to the sociocultural fact behind these students’ L1 background. As for the latter reason, it is a known norm that Asian style of interaction is commonly seen to be indirect. According to Husain et al. (2012) who studied communication norms of Malaysian community and Centre for Intercultural Learning of Canada from which comments on Thai style of interaction were taken, Malaysian and Thai people prefer to be indirect in their discussions. This observation could lend support to any possible sociocultural explanation for the preference of hedging among Thai and Malaysian subjects in this study. Moreover, based on a slightly higher frequency of hedges found in the Thai students’ writing, it could be said that the rhetorical style of Thai students is shown to be slightly more tentative than the Malaysian students. Nevertheless, such sociocultural explanation should be given less weight than the fact that hedges are commonly preferred in academic genre. While Thai students utilized a higher number of hedging markers, Malaysian students were seen to have a finer balance between the number of hedges and boosters in their discussions. This was proven by a smaller gap between both markers (hedge 38.42% versus boosters 31.33%) used by Malaysian students. Such fine balance is considered ideal for a rhetorical style of interaction as postulated by Hyland (2004). A fine number of booster markers in the arguments helps to strengthen a proposed standpoint and convince the readers while hedging markers help to make it sound not overtly confident. In this way, Malaysian subjects were seen making their rhetorical style sound more persuasive in writing thesis discussions. Attitude markers ranked as the third most frequently employed markers by Thai and Malaysian students in this study. Both groups showed a very close percentage with each other (Thai 15.52% vs. Malaysian 15.36%). Basically, the use of attitude markers signified the writer’s level of involvement in the text. It is a platform for the writer to express his/her affective attitudes towards the propositions (Hyland, 2005) and interact with the readers. The degree of writer’s involvement in the text through the use of attitude markers is a measurement of boundary between the writer and the readers. It was shown in this study that both Thai and Malaysian students exhibited a distant rhetorical style of interaction with their readers when it comes to proposing attitudinal remarks. The small percentage of MD attitude markers seemed to suggest that both groups of students showed a small level of involvement in the discussions. Compared to Malaysians, Thai students were seen to have higher use of this MD device. This could relatedly indicate that not only their writing was meant to carry out a solid fact but it was also aimed to weave personal interaction with the readers and 32 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan establish a fine rapport in the discussion, yet it did so not very outstandingly. The numerical count of attitude markers also suggested that, compared to Malaysian students, the rhetorical style of Thai student writers was considered more interactional and reader-friendly. Among the lexical choices found in use as MD attitude markers were adjectives and adverbs such as surprisingly, interesting, reasonable, importantly, and significantly. The limited occurrence of attitude markers in the thesis discussions of Thai and Malaysian students in this study can be explained from a possible sociocultural perspective. Park and Kim (2008) explore the communication styles of Asian and American community and discover that Asian people preferred to have a less open communication style due to their higher level of emotional self-control. It is a cultural value of communication of many Asian societies to suppress their emotional expressions and also to avoid proposing personal opinions or attitudes towards a proposition. Expressing one’s personal attitude towards a proposed statement is not encouraged among Asian communities and, therefore, an indirect type of interaction was found to be largely preferred by them. Likewise, in the case of written discourse, employing attitude marker as an interactional MD feature is deemed to be less usual among Asian people. Thus, such sociocultural factor is a possible explanation for the small percentage of attitude markers found in the thesis discussions of Thai and Malaysian students in this particular study. Engagement marker is another interactional MD marker that helps to measure the level of solidarity between the writer and readers. It is typically used to highlight or downplay the presence of readers in the text and used as a tool to engage the readers with statements presented in the discussion (Hinkel, 2002). In this study, the use of engagement markers was ranked fourth among the frequent markers employed by Thai and Malaysian students. A small percentage of occurrence (Thai = 12.48% and Malaysian = 12.67%) seems to suggest that employing engagement marker was not a regular practice among Thai and Malaysian writers in this study. This is consistent with the study by MurDuenas (2011) in which engagement markers were found to be the least used compared to other types of interactional MD markers. However, the result was somehow not consistent with Chan and Tan (2010) who discover that engagement marker was the most highly employed device preferred by Malaysian students. Chan and Tan found that Malaysian subjects in their study often used we inclusive pronoun to mark solidarity with the readers. Kashiha (2018) also identified a high propensity on the use of we among Malaysian students in his studies when presenting their arguments in academic writing. Interestingly, in the current study, we was marked as self-mention for it was employed as a pronoun rather than a signal to address inclusiveness with the readers. Since an engagement marker metaphorically functions as a magnetic device that draws readers into the text, having less of it therefore resulted in the rhetorical tone of Thai and Malaysian writings seems to appear less engaging. It could be said that the persuasiveness in their writing style was somewhat alleviated and less catchy. The fact that readers were expected to take their own efforts to draw themselves into the texts instead of the writers taking responsibility bridging the readers into the discussions; it seems to consequently suggest that Thai and Malaysian students in this study were prone to produce a discourse which rhetorically was reader-responsible type of writing. In other words, the discussion texts written by the Thai and Malaysian students somehow seemed to require more effort and responsibility from the reader counterpart in order to understand the intended message conveyed in the texts. In most sentences, readers were deemed to make their own effort in RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 33 guiding themselves into the text because not much of MD devices such as engagement markers were seen to do the task. The less frequent use of engagement markers in the writing could be due to two possible reasons. Having a poor exposure to lexical terms for engagement markers is one possibility; whereas a sociocultural practice from the students’ L1 background is another. It is not common for Eastern culture of communication to persuade the interlocutor in an explicit manner. Due to facesaving concerns, Asian speakers or writers tend to avoid completely dwelling themselves on the discourse for it might cause them to take a full responsibility should there be any miscommunications (Asma & Pederson, 2009). Writer-responsible type of rhetorical interaction is thus hardly found in practice among Asian people in which Thais and Malaysians are included. Self-mention was the least used MD feature in the thesis discussions of Thai and Malaysian students. Based on the analysis, a number of students, particularly Thai students, were not inclined to use this type of marker at all in their discussions. This MD tool basically determines the degree of writer presence in the text (Hyland, 2005; 2010). It also shows the level of writer’s involvement in the text and the measurement of how interactional the discourse is. Despite the fact that self-mention nowadays has become a common MD device employed by many Anglo-American academic writers as a tool to establish credentials to the texts (Hyland, 2004; Mur-Duenas, 2011), the majority of Thai and Malaysian students in this study were not used to employing it in their thesis writing. The lesser use of self-mention by Thai and Malaysian students as an MD feature in the discourse could signify that they are not in favor of explicating authorial identity to their readers. Withdrawing the writer’s stance from the text was their preferred rhetorical style of interaction especially for the Thai students. The attempt to avoid mentioning themselves in the text could indicate that the Thai and Malaysian students preferred to present the discussions in a less dialogic manner. To them, keeping a small distance from the readers may have deemed to render more academic style to the discourse, although the American academic writers in Mur-Duenas’ (2011) study proved it otherwise. Particularly to the Thai subjects, presenting oneself with the use of first-person pronoun was not a common scenario in their academic discourse. As such, the Thai subjects who used self-mention as an MD marker, therefore, tended to address themselves as the researcher instead of using plainly first -person pronoun like I or we to discuss their findings. This seems to suggest that they would rather prefer to disguise their appearance as a reliable someone in order to preserve the factual tone of the discourse than to openly address themselves in a dialogic manner. Compared to the Malaysians, Thai students were seen to have employed less of self-mention in their arguments to the extent that their rhetorical style of interaction consequently sounded less writer-involved and less reader-friendly than the Malaysian counterpart as a result. Employing self-mention in an academic genre is not a common cultural practice among non-native writers. Many Eastern writers, students in particular, were trained to avoid mentioning themselves in writing academic discourse so as to maintain a professional tone (Hyland, 2004). Thai and Malaysian students in this study, too, were found not an exception of such sociocultural dogma. However, this is seen to contradict to the Anglo-American writers commonly practise in their academic writing. Mauranen (1993) observes that Anglo-American writers frequently prefer to signal their personal presence in academic texts. Personal markers and inclusive expressions play important role as the MD devices that help them to create solidarity and build rapport with their readers in the discourse. This observation was also supported by many studies in metadiscoursal field (i.e. Abdi, 2009; Faghih and Rahimpour, 2009; Noorian & Biria, 2010; Mur-Duenas, 2011; Zakaria, 2016). 34 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan 4.0 CONCLUSION In general, the similarity of MD frequency counts acquired by both groups seems to suggest that Thai and Malaysian master’s students in this study shared similar rhetorical preferences in writing thesis discussions. Based on the analysis of each type of interactional MD markers, the rhetorical styles of interaction can be construed in several ways. Firstly, the prevalent use of hedges over boosters from the results implies that both groups preferred to interact with their readers in a tentative, indirect manner. Secondly, the small percentage of occurrence of attitude markers and engagement markers indicates that the writer’s relationship with the readers was somewhat less-involved and distant. Reader-responsible type of interaction was seen to be their rhetorical preference. Lastly, a scant number of self-mention devices in the texts seemed to suggest that the level of writer’s involvement in the text was very insignificant. Thai and Malaysian students were seen not in favor of building explicit relationship with their readers, therefore, resulted in their rhetorical tone of writing to be less dialogic and distant. There were a few possible reasons to explain the rhetorical preferences of Thai and Malaysian students in thesis writing. Apart from perceiving it from a linguistic proficiency point of view, sociocultural perspective was also another possible factor contributing to the analysis of rhetorical style of writing performed by Thai and Malaysian students. It is postulated that Asian styles of communication are commonly seen to be indirect, distant, tentative and less-open (Park & Kim, 2008; Asma & Pederson, 2009; Husain et al., 2012) and it is such rhetorical style of interaction that was seen in the written discourse of Thai and Malaysian students in this study. Hence, it could be said that sociocultural factor of the students’ L1 background can be another relevant explanation to the rhetorical performance of Thai and Malaysian masters’ students in writing their thesis discussions as seen from the analysis mentioned above. REFERENCES Abdollahzadeh, E. 2011. Poring Over the Findings: Interpersonal Authorial Engagement in Applied Linguistics Papers. Journal of Pragmatics. 43: 288-297. Abdi, R. 2002. Interpersonal Metadiscourse as an Indicator of Interaction and Identity. Discourse Studies. 4: 139-145. Abdi, R. 2009. Projecting Cultural Identity through Metadiscourse Marking: A Comparison of Persian and English Research Articles. Journal of English Language Teaching and Learning. 212: 1-15. Adel, A. 2006. Metadiscourse in L1 and L2 English. Philadelphia: John Benjamins. Anwardeen, N. H., Luyee, E. O., Gabriel, J. I., & Kalajahi, S. A. R. 2013. An Analysis: The Usage of Metadiscourse in Argumentative Writing by Malaysian Tertiary Level of Students. English Language Teaching. 6(9): 83-96. Asma, A & Pederson, P. 2009. Understanding Multicultural Malaysia: Delights, Puzzles, and Irritations. Kuala Lumpur: Pearson, Prentice Hall. Babbie, E. R. 2013. The Practice of Social Research. Belmont, CA: Wadsworth Cengage Learning. RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 35 Chan, S., & Tan, H. 2010. Extracting and Comparing the Intricacies of Metadiscourse of Two Written Persuasive Corpora. International Journal of Education and Development using ICT. 6(3): 124146. Chen, T. Y., & Kuo, C. H. 2012. A Genre-based Analysis of the Information Structure of Master's Theses in Applied Linguistics. The Asian ESP Journal. 8(1): 24-52. Connor, U. 1996. Contrastive Rhetoric: Cross-cultural Aspects of Second Language Writing. Cambridge: Cambridge University Press. Connor, U. 2002. New Directions in Contrastive Rhetoric. TESOL Quarterly, 36: 493-510. Crismore, A. Markkanen, R. and M. Steffensen. 1993. Metadiscourse in Persuasive Writing: A Study of Texts Written by American and Finish University Students. Written Communication. 10(1): 39-71. Faghih, E., & Rahimpour, S. 2009. Contrastive Rhetoric of English and Persian Written Texts: Metadiscourse in Applied Linguistics Research Articles. Rice Working Papers in Linguistics. 1: 192-197. Hinkel, E. 2002. Second Language Writers’ Text. Mahwah, NJ: Erlbaum. Husain, S., Rohany, H., & Noor, N. 2012. Unblock Arguments in Malaysian Sopoblogs. International Journal of Social Sciences and Humanity Studies. 4(1): 347-356. Hyland, K. 2004. Disciplinary Interactions: Metadiscourse in L2 Postgraduate Writing. Journal of Second Language Writing. 13: 133-151. Hyland, K. 2005. Metadiscourse: Exploring Interaction in Writing. London, NY: Continuum. Hyland, K. 2010. Metadiscourse: Mapping Interactions in Academic Writing. Nordic Journal of English Studies. 9(2): 125-143. Kashiha, H. 2018. Malaysian ESL Students’ Perception of Metadiscoure n Essay Writing. Online Journal of Communication and Media Technologies. 8(3): 193-201. Khedri, M., Ebrahimi, S. J. and Heng C. S. 2010. Interactional Metadiscourse Markers in Academic Research Article Result and Discussion Sections. The Southeast Asian Journal of English Language Studies. 19(1): 65-74. Loan, N. T. T., & Pramoolsook, I. 2016. Master’s Theses Written by Vietnamese and International Writers: Rhetorical Structure Variations. The Asian ESP Journal. 12(1): 106-127. Loi, C. K. & Lim, J. M. H. 2019. Hedging in the discussion sections of English and Malay Educational Research Articles. GEMA Online Journal of Language Studies. 19(1): 36-61. Mauranen, A. 1993. Contrastive ESP Rhetoric: Metatext in Finnish-English Economics Texts. English for Specific Purposes. 12: 3-22. Merriam, S. B. 2009. Qualitative Research: A Guide to Design and Implementation. 2nd ed. San Francisco: Jossey-Bass. Milne, E. D. 2003. Metadiscourse Revisited: A Contrastive Study of Persuasive Writing in Professional Discourse. Estudios Ingleses de la Universidad Complutense. 11: 29-52. Mur-Duenas, P. 2011. An Intercultural Analysis of Metadiscourse Features in Research Articles Written in English and in Spanish. Journal of Pragmatics. 43: 3068-3079. Noorian, M., & Biria, R. 2010. Interpersonal Metadiscourse in Persuasive Journalism: A Study of Texts by American and Iranian EFL Columnists. Journal of Modern Languages. 20: 64-79. 36 Faridah Hayisama, Mohamed Ismail Ahamad Shah & Wan Nur Asyura Wan Adnan Park, Y. S., & Kim, B. S. 2008. Asian and European American Cultural Values and Communication Styles Among Asian American and European American College Students. Cultural Diversity and Ethnic Minority Psychology. 14(1): 47. Rasekh, A. E. & Amiryousefi, M. 2010. Metadiscourse: Definitions, Issues and Its Implications for English Teachers. English Language Teaching. 3(4): 159-167. Tan, H. 2012. A Proposed Metadiscourse Framework for Lay ESL Writers. World Applied Sciences Journal. 20(1): 1-6. Toumi, N. 2009. A Model for the Investigation of Reflexive Metadiscourse in Research Articles. Language. 1: 64-73. Uysal, H. H. 2012. Argumentation across L1 and L2: Exploring Cultural Influences and Transfer Issues. Vigo International Journal of Applied Linguistics. 9: 133-159. Zakaria, M. K. 2016. Metadiscourse in the Academic Writing of Malaysian and Arab Pre-university Students at the IIUM. Unpublished Master Thesis. International Islamic University Malaysia, Kuala Lumpur, Malaysia. RHETORICAL STYLE ACROSS CULTURES: AN ANALYSIS OF METADISCOURSE MARKERS 37 Appendix Frequency of Use of Interactional MD Features in Thai Students’ Thesis Discussion (Individual Results) MD frequency occurrence Hedges Boosters Attitude Engagement markers Self-mention Subject no. markers Item % (s) Item % (s) Item % (s) Item % (s) Item % (s) 1 18 36.73 13 26.53 7 14.28 9 18.36 2 4.08 2 33 53.22 24 38.70 3 4.83 2 3.22 0 0 3 46 43.80 33 31.42 10 9.52 16 15.23 0 0 4 11 28.20 13 33.33 9 23.07 6 15.38 0 0 5 16 31.37 15 29.41 12 23.52 8 15.68 0 0 6 86 46.73 44 23.91 29 15.76 25 13.58 0 0 7 23 32.39 23 32.39 14 19.71 11 15.49 0 0 8 2 9.52 12 57.14 2 9.52 5 23.80 0 0 9 33 41.77 31 39.24 7 8.86 7 8.86 1 1.26 10 9 15.51 20 34.48 23 39.65 5 8.62 1 1.72 11 50 48.54 32 31.06 16 15.53 5 4.85 0 0 12 58 61.05 15 15.78 9 9.47 13 13.68 0 0 13 2 12.50 6 37.50 4 25.00 4 25.00 0 0 14 5 25.00 9 45.00 3 15.00 3 15.00 0 0 392 41.13 290 30.43 148 15.52 119 12.48 4 0.42 Total