Reflexive metadiscourse in Chinese and English sociology research article introductions and discussions

Reflexive metadiscourse in Chinese and English sociology research article introductions and discussions

Journal of Pragmatics 159 (2020) 47e59 Contents lists available at ScienceDirect Journal of Pragmatics journal homepage: www.elsevier.com/locate/pra...

543KB Sizes 1 Downloads 62 Views

Journal of Pragmatics 159 (2020) 47e59

Contents lists available at ScienceDirect

Journal of Pragmatics journal homepage: www.elsevier.com/locate/pragma

Reflexive metadiscourse in Chinese and English sociology research article introductions and discussions Zhijun Li a, Jinfen Xu b, * a b

College of Foreign Languages, Huaqiao University, China School of Foreign Languages, Huazhong University of Science and Technology, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 23 February 2019 Received in revised form 22 January 2020 Accepted 3 February 2020

To extend the relatively neglected line of cross-linguistic research on metadiscourse in research articles (hereafter RAs) in Chinese and English languages and cultures, this article, adopting the reflexive approach, investigated the use of metadiscourse in Chinese and English RA introductions and discussions. Based on manual discourse analysis of 60 Chinese and English RAs in sociology, it is found that more impersonal than personal metadiscourse was deployed and the distribution of the impersonal subcategories displays similarity in both Chinese and English introductions and discussions. However, English sociologists tend to include more metadiscourse as a whole, and the respective personal and impersonal metadiscourse, particularly personal metadiscourse and in the discussion section in their RAs, than their Chinese colleagues. Furthermore, the gap between impersonal and personal metadiscourse use was notably more salient in Chinese than in English in both sections. Linguistic features, sociocultural factors and rhetoric functions may be responsible for these differences. © 2020 Elsevier B.V. All rights reserved.

Keywords: Reflexive metadiscourse Research articles Reflexive approach Chinese English Sociology

1. Introduction € Metadiscourse is “text about the evolving text, or the writer's explicit commentary on her own ongoing discourse” (Adel, 2006:20). Based on a view of writing as a social and communicative engagement between writers and readers, metadiscourse has been conceptualized as an essential pragmatic construct which helps project writers into their work to signal their communicative intentions and contributes to readers' proper understandings of both the text and their attitudes as well as commitments to its content and the audience (Hyland, 1998b: 437). Metadiscourse as such is likely to reflect “cultural norms, values, and belief systems prevailing in discourse communities which constitute social contexts of texts'’ (Golebiowski and Liddicoat, 2002: 59). The above expectation has been supported by a large body of research on academic genres, both spoken and written, that has investigated the use of metadiscourse in various languages and disciplines. Among this research, a few analyzed meta€ discourse in spoken academic discourse (e.g., Mauranen, 2001; Adel, 2010). The majority of such research explored metadiscourse in written academic genres. Some research on written genres examined the use of metadiscourse from a crossdisciplinary perspective and found abundant evidence of disciplinary influences. On one hand, differences in the use of metadiscourse were identified across the broad soft (the humanities and social sciences) and hard (the sciences) disciplines (e.g., Harwood, 2005b). On the other hand, more studies reported discrepancies in metadiscourse use across various specific

* Corresponding author. School of Foreign Languages, Huazhong University of Science and Technology, Wuhan 430074, Hubei Province, PR China. E-mail addresses: [email protected] (Z. Li), [email protected] (J. Xu). https://doi.org/10.1016/j.pragma.2020.02.003 0378-2166/© 2020 Elsevier B.V. All rights reserved.

48

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

disciplines such as mechanical engineering, electrical engineering, marketing, philosophy, sociology, applied linguistics, n, 2010), applied linguistics physics, and microbiology (Hyland, 2007), applied linguistics and business management (Milla and psychology (Cao and Hu, 2014), and economics, medicine and linguistics (Salas, 2015). Meanwhile, a huge group of contrastive research has investigated the use of metadiscourse across languages/cultures, particularly across English and other Indo-European languages including Finnish (e.g., Mauranen, 1993a), Italian (e.g., Molino, rez-Llantada, 2010; Mur-Duen ~ as, 2011), Ukrainian and Russian (e.g., Yakhontova, 2006), Norwegian 2010), Spanish (e.g., Pe and French (e.g., Dahl, 2004), German (Sanderson, 2008), German, French, Russian and Bulgarian (Vassileva, 2001), and Persian (e.g., Zarei and Mansoori, 2011). All of these studies have found language-/ culture-specific features of metadiscursive strategies. Then what about the use of metadiscourse in Chinese RAs in contrast to English ones? Do Chinese scholars use metadiscourse in the same way as English scholars or differently? Unfortunately, only a few sporadic studies have been documented about Chinese RAs. For instance, Hu and Cao (2011), Yang (2013), and Chen and Zhang (2016) focused on hedges (and boosters) in Chinese and English RA abstracts or across Chinese-authored English and Chinese and English-authored English RAs, and Loi and Lim (2013) on metadiscourse in Chinese and English educational psychology RA introductions. Mu et al. (2015) recently compared the use of metadiscourse in applied linguistics RAs. While these studies have helped us to understand how metadiscourse is deployed in Chinese and English RAs, a couple of issues merit further research. First, most of these studies (Hu and Cao, 2011; Yang, 2013; Chen and Zhang, 2016) examined only a subset of metadiscourse and very few systematically investigated metadiscourse use in their totality across Chinese and English RAs. Second, these studies rarely delved into the use of metadiscourse in specific sections of RAs with the exception of Loi and Lim (2013) who focused on the introduction section. Other sections such as the discussion section where metadiscourse may play an essential role to help fulfill its function of interpreting research findings and justifying the writer's argumentations (Swales, 1990) receive little attention. In addition, past cross-linguistic/cultural research on metadiscursive features centered predominantly on such disciplines as linguistics (e.g., Molino, 2010; Hu and Cao, 2011; Mu et al., 2015), broad domains of the humanities, social sciences and hard sciences (e.g., Zarei and Mansoori, 2011; Yang, 2013) and educational psychology (e.g., Loi and Lim, 2013). Studies focusing on sociology are scarce. To our knowledge, Is¸ık-Tas¸ (2018) explored how authorial identity is represented through first person pronouns in Turkish and English sociology RAs published in national and international journals and identified obvious effect of the publication context on the discoursal choices of academic writers. In addition, Bruce (2010, 2016) investigated the metadiscourse devices in English sociology university essays and Hyland (2007, 2008) reported functions of subsets of metadiscourse (code glosses, stance and audience engagement markers respectively) in English sociology RAs in their crossdisciplinary studies. Nevertheless, so far, we have found no research on metadiscourse use across the Chinese and English language/culture in the field of sociology. Although metadiscourse in RAs has been researched in numerous previous studies, in today's multilingual contexts, crosscultural studies into the use of metadiscourse deserves more attention since writing is “very much shaped by the educational system in a writer's native culture” and thus writing on academic research is definitely “influenced by the variation in the writing cultures that carry it.” (Mauranen, 1993a:4) Therefore, it remains important to investigate how metadiscourse e a crucial device for the construction of academic writing – is employed by writers from different national cultures even though the same language is involved, or alternatively, by writers using different languages in different cultural settings, especially languages and cultures other than the Indo-European ones (that has been adequately examined as reviewed above), among which Chinese, a dominant Sino-Tibetan language and culture, is particularly worth more attention. Furthermore, given the discipline-specific characteristics of metadiscourse employment as evidenced by the aforementioned cross-disciplinary research, sociology might present a different picture in metadiscourse use from other disciplines. In view of these, we contrastively analyze RAs written and published within the discipline of sociology in two different language and cultural contexts – the US context, which is open to an international readership, and the Chinese national context – with a view to investigating cross-linguistic/cultural differences in the use of metadiscourse. Our analysis focus on two rhetorically forceful RA sections – introductions and discussions in that they are where the writer persona and the imagined reader, both of which are essential components of metadiscourse, are most likely to be € present, explicitly or implicitly (Adel, 2006:69). We address the following two questions:

(1) To what extent is metadiscourse use similar and divergent in the international English context and the Chinese national context within a particular disciplinary community? (2) How can these similarities and divergences in metadiscourse use in the same disciplinary community but different languages/cultural contexts be explained?

It is our hope that this study will contribute to cross-cultural research in English for Academic Purposes (EAP) and foreign/ second language writing.

2. Theoretical approach Concerning how to define, delimit and categorize metadiscourse, two different approaches have emerged in the past € decades – the interactive approach and the reflexive approach (e.g., Flowerdew, 2015:19e20; Adel, 2017:54). The interactive

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

49

approach is broad and inclusive, viewing metadiscourse primarily as a form of textual interaction and emphasizing the textual and interpersonal functions in Hallidayan terms. The reflexive approach is narrow and restricted, seeing metadiscourse essentially as a form of language reflexivity and focusing on the writer's commentary on the on-going text (e.g., Mauranen, € 1993b; Adel, 2006). € Based on Jakobson’s (1998) functional model of language, the reflexive approach (Adel, 2006) emphasizes the metalinguistic, the expressive and the directive function, which respectively corresponds to the three main components of € metadiscourse: the discourse or text itself, the writer/speaker persona and the real/imagined audience (Adel, 2006:18). Reflexivity is the essence of this approach. Metadiscourse “displays an awareness of the current text or its language use per se € and of the current writer and reader qua writer and reader” (Adel, 2006:20) and is characterized by four properties – being fuzzy, functional, multifunctional and context-dependent. Accordingly, the reflexive approach introduces four features for identifying metadiscourse: “explicitness”, “world of discourse”, “current discourse”, “writer qua writer” and “reader qua € reader” (Adel, 2006:27e29). “Explicitness” refers to the requirement that the reference to the world of discourse be overtly stated. “World of discourse” focuses on the ongoing discourse rather than on other “worldly” or “real world” activities or phenomena that are external to the text. “Current discourse/text” (Mauranen, 1993b) stresses how texts refer to themselves instead of other texts (references to other texts are considered intertextual in this model). “Writer qua writer” and “reader qua reader”, which only apply to personal metadiscourse, mean that references to writers and/or readers are made to the producer and/or receiver of the current discourse in their roles as discourse participants (e.g., as I discussed earlier), as opposed to an experiencer in the real world (e.g., I will teach lessons next Monday). The interactive approach has received wide scholastic attention, with Ken Hyland being the representative, whose framework (e.g., Hyland, 2005) has been applied in a wealth of research on metadiscourse in academic genres, particularly in the prestigious genre of RAs (e.g., Hyland, 2007; Loi and Lim, 2013; Cao and Hu, 2014; Mu et al., 2015). By contrast, attention to rez-Llantada’s, 2010; Salas, 2015; the reflexive approach is relatively limited in that merely a few studies (Toumi, 2009; Pe Zhang, 2016) have adopted it to explore metadiscourse features in RAs. Furthermore, the limited quatitative studies taking the reflexive approach showed that “The amount of metadiscourse material is very limited compared to the amount of rez-Llantada, 2010:60) and found merely around one metadiscursive marker per 100 propositional, expository material” (Pe words (Salas, 2015), much lower than the occurrences widely reported in research taking the interactive approach, e.g., ~ as’ (2011) study. However, this roughly one metadiscourse occurrence per 14 words in Hyland’s (1998a) and Mur-Duen conflicting finding is not sufficiently convincing as evidence provided by quantitative studies taking the reflexive approach is rare. Therefore, this study aims not only to enrich research on metadiscourse use in RAs taking the reflexive approach but also to attest the above conflicting finding about the occurrence of metadiscourse taking distinct approaches. We decide to employ the reflexive approach for two considerations. Firstly, it is theoretically more rigorous in that it puts forward clear and precise criteria to demarcate the fuzzy concept of metadiscourse, delimiting its scope and discerning it from other bordering or neighboring phenomena (e.g., stance, intertexuality), and thereby depicts a more accurate picture of the metadiscourse phenomenon than other broader frameworks. Secondly, it avoids the main objections to the concept of metadiscourse – that linguistics has no true ‘meta-’ level, for example, linguistics has no separate formal system to describe the object of study – (e.g., Sinclair, 2005) by highlighting the reflexivity of metadiscourse, through including the participants in the discourse. 3. Analytical framework Considering the nature of metadiscourse characterized as fuzzy, (multi-)functional and thus context-dependent (Hyland, € 2005; Adel, 2006), we constructed the categorization of reflexive metadiscourse through careful observation and cautious manual analysis and annotation of the data. It is noteworthy that there is so far no taxonomy of reflexive metadiscourse about Chinese. The taxonomy for Chinese was therefore constructed through cautious manual analysis and annotation of the present data in light of previous taxonomy of English metadiscourse. Metadiscourse unit was chosen as the unit of analysis in this study. A metadiscourse unit is an utterance or a chunk of words which express(es) a metadiscursive function. For instance, noun phrases like the last section (of this article) and other utterances giving comment on the current text or the writer's discourse such as put simply and in sum perform specific metediscursive functions and are counted as separate metadiscursive units. However, there has been conflicting views on the delimitation of “a chunk of words”. Some scholars (e.g., Crismore and Farnsworth, 1990) chose to count fairly large chunks € whereas others (e.g., Adel, 2006) chose to count relatively small ones as one occurrence of metadiscourse. Take the sentence I will bring up this topic in Section 3. as an example. It would be counted as one single occurrence of metadiscourse, or one metadiscourse unit, according to Crismore and Farnsworth’s (1990) framework since it serves one general metadiscursive € function – announcement of the topic for discussion in the text. By contrast, following Adel’s (2006) framework, it could be counted as two linguistic-functional units: unit (a) I will bring up this topic, a subject þ VP part which is writer-oriented and functions as a discourse label, and unit (b) in section 3, a prepositional phrase (PP) that is text-oriented and functions as reference to the following text. Although there might be good arguments for considering larger chunks of words as metadiscourse units, we consider the latter micro-level analysis allows for more fine-tuned distinctions to be drawn and more accurate comparison to be made across texts, which is particularly crucial in the present study, in which Chinese and English writings are compared. Therefore, we adopted the micro-level delimitation of linguistic-functional utterance or chunk of words as the unit of analysis for metadiscourse in this study.

50

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

Table 1 Functional categories of personal metadiscourse.

Meta-text

Discourse function

English Examples

Defining defines academic terms and concepts Saying foregrounds that something is communicated Introducing Topic proclaims what the text is to be about, facilitating readers' processing of the subsequent text Focusing announces informational focus and narrows it down. Concluding a topic

... what I call ...

Reminding points backwards to textual materials that has been said before Arguing stresses the discourse act being performed by using performative verbs Adding explicitly states that new information is being added to existing one(s) Exemplifying introduces an example Writer-reader Anticipating R's Reaction predicts readers' Interaction reaction to what is said Clarifying specifies matters in order to avoid misinterpretation by readers Aligning perspectives presupposes the reader's agreement with the writer Appealing to the reader conveys the writer's attitude with the aim of entreating or correcting the reader.

But can we say that …? (In the following section,) I provide a brief overview of ...

Chinese Examples

我们不能(笼统地)说… [We cannot say… in (a general way).] ...这是我们尝试(在此)回答的一个问题。[This is a question we attempts to answer (here).]

I focused here on... I conclude from this that... As I demonstrated (above)... I highlight that ...

I also ...

我们得出下述结论… [We draw the following conclusions…] 正如我们(前面)所指出的…[As we pointed out above…] 我们认为可能的解释是…[We think the possible explanation might be …] 我们也注意到… [We also notice that …]

I draw on the case of ... to illustrate ... I acknowledge some limitations... I do not argue (here) that ... We could assume that...

我们可以看出… [We can see that…]

We should be alert that...

...因此我们可能需要更加谨慎。[…Therefore, we need to be more alert.]

Under the reflexive framework, metadiscourse is classified into “personal” and “impersonal” in accordance with the realizations of metadiscourse. Personal metadiscourse directly refers to the writer and/or reader as participants in the current € discourse (Adel, 2006:19) by pronominal forms, i.e., the first and second person pronouns I, my, me, we, our, us and you, your and their Chinese counterparts 我(们)(的) [I (me)(my)(we)(us)(our)]1,2,你(们)(的) [you (you)(your)] and self-referring nouns the author/ writer as well as their Chinese equivalents 作/笔者. Besides, generic (pro)nouns like one and 人们 (people) may also constitute metadiscoursal resources that refer to the current writer/reader. However, no instance of the second-person pronoun as reflexive meradiscourse marker of the imagined reader was identified in our data. Metadiscursive selfreferring nouns the author/writer and 作/笔者 were also rare, only 5 cases in Chinese and none in English. Occurrence of Generic (pro)nouns as metadiscourse was even negligible. Therefore, the overwhelming occurrences of personal metadiscourse were units of first-person pronouns but no metadiscursive singular first-person pronoun occurred in the Chinese subdata. Personal metadiscourse consists of two categories: ‘Metatext’ and ‘Writer-reader Interaction’. Metatext spells out writers' textual organization and language use, while Writer-reader Interaction primarily focuses on writerereader relations by € addressing readers and engaging them in a mock dialogue. All of Adel's further subcategories under these two categories (see Table 1) were identified according to their discourse functions in our study, except for a Meta-text subcategory – Contextualizing, and two Writer-reader Interaction subcategories – Imagining Scenarios and Hypothesizing About the Reader. However, instances of some subcategories were absent either from the Chinese or English subdata. € Impersonal metadiscourse, however, implicitly refers to the current discourse participant (Adel, 2006:14) by means of various impersonal constructions (e.g., this article). It displays the text structure or refers to the text itself lexically and is divided into four functional categories: References to the Code/Text, Phorics, Code Glosses and Discourse Labels, discourse functions and examples of which are shown in Table 2. What is noteworthy is that a large fraction of discourse verbs’ (under the subcategory of Discourse Labels) finite forms were captured in the examination of personal metadiscourse. All instances of these verbs having personal pronouns or self-referring nouns as subjects were therefore disregarded and only their impersonal usage (i.e., having impersonal entities as subjects, e.g., this article aims to …), non-finite verb forms (e.g., to conclude, additionally) and nouns (e.g., conclusion, example) were considered in the investigation into impersonal metadiscourse. The discourse functions of impersonal Discourse Labels subcategories are the same as the corresponding personal Meta-text subcategories (as shown in Table 1) with the difference being impersonal vs. personal realization forms. Hence, the discourse functions of impersonal subcategories are not specified in Table 2.

1

All examples in Tables 1 and 2 are from the present data. Plural personal pronouns in Chinese are formed by adding the character “们” (which is like a suffix in English indicating “more than one”) to singular personal ones and possessive pronouns are formed by adding “的” (which is close to the meaning of “of”) to subject ones. Oblique forms are the same as subject forms. 2

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

51

Table 2 Functional categories of impersonal metadiscourse. Category

Discourse function

Refer. to Code/Text refers to the words/expressions used in the text / the text itself at various levels Phorics Previews announce what is to happen in the text Reviews remind the reader of previous chunks in the text Enumerators order specific parts of the discourse in relation to each other Textual deixis draws attention to the current point or specific chunks in the text Introducing & closing proclaim the beginning /ending of topics Code Glosses interprets elements or comment on ways of responding to elements in the text Discourse Saying /defining Labels Introducing topics

English Examples

Chinese Examples

…this article … … Fig. 1…

本文… [this article …] 图1…[Fig. 1…]

… the following … … below … …(mentioned) above … (discussed) earlier First … Second … …here… …now…

以下 … [the following …] 如下 … […below] 上述… [… above (stated)] 前述… [… earlier(described)] 第一, …[First, …] 第二, …[Second, …]

Finally, …

最后, … [finally, …]

i.e., … Specifically, … …be referred to as … be defined … The central goal of this article is … This article aims to …

即… [i.e., …] 总体而言 … [in general…] (本文)将…称为… [ (This article) refer to… as…] …是本文要回答的问题. […are the questions to be answered in this article.] 本文尝试… [This article attempts to …] 总之 …[in sum …] 本文有两个基本结论…[Two conclusions can be drawn in this article …] 如… [for example, …] 诸如 … [such as …] 本文强调…[This article emphasized that …], 可能的解释是…[The possible explanation is …] 此外, … [ In addition, …] 另外, … [Additionally …]

Concluding

To conclude … In sum, …

Exemplifying

e.g., … such as … This article highlights…, One possible explanation is … Additionally, … In addition, …

Arguing Adding

4. Methodology 4.1. Data We made every effort to construct parallel subcorpora for the present study according to the principles of Tertium of Comparationis proposed by Connor and Moreno (2005), such as genre, subject matter and writers' expertise, among other things. First, we selected peer-reviewed leading sociology journals in the two languages from two influential citation indices – Chinese Social Sciences Citation Index (CSSCI) and Social Sciences Citation Index (SSCI). The final chosen journals are: Sociological Studies and Society published in mainland China and American Sociological Review and American Journal of Sociology published in the U.S. All these journals are widely read by scholars in their respective communities, and represent the best of sociological research in China and internationally. Second, in terms of genre and subject matter, all selected RAs are original single-authored ones on empirical study in applied sociology. Single-authored articles were chosen to ensure all plural form realizations of self-mentions are metadiscursive since previous research (e.g., Mu et al., 2015) has found that the plural forms of metadiscursive first-person pronouns are more prevalent in Chinese than in English. The choice of articles on empirical studies is mainly out of the consideration that this type of articles usually takes “a highly codified, institutionalized form” (i.e., the introductionemethoderesultsediscussion structure) (Bazerman, 1988:259) which helps to ensure the comparability of the specific sections – an essential dimension of the present study. Third, all the RAs were taken from the publication period of 2014e2018. Fourth, 30 RAs were selected for both subcorpora. Description of the data is shown in Table 3.

Table 3 Description of the data. Chinese

Average text length by words Average number of sentences Total number of Chinese / English words

English

Introduction

Discussion

Introduction

Discussion

611.3 19.6 18,338

1076.8 33.7 32,304

923.7 40.1 27,711

1702.3 72.4 51,068

Note: Here every introduction or discussion section is treated as a single text when statistics are made.

52

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

4.2. Procedures We followed the procedures below to conduct this research: Step 1 We selected single-authored original articles whose author names are English (both given names and surnames were searched online and even on their personal website(s) if possible to make sure they were English and RAs whose authors’ last names did not look English, e.g., Lee, were definitely removed from our sample) and Chinese and author affiliations are tertiary institutions in the U.S. and mainland China respectively by going through the “table of contents” and “author information” of each issue of the journals. Step 2 We sifted the above selected articles and retained those on empirical study by skimming the abstracts, and the full text in the case of difficult judgment solely made on abstracts. Step 3 We decided on those articles containing the independent (from the literature review) introduction section (hereafter “introductions”) and “discussion and (&/ /) conclusions” or “discussion” section (hereafter “discussions”) as the final samples. Step 4 The pdf or html version of all articles was downloaded and transformed to text documents for the convenience of software concordancing. Then the introduction and discussion sections were extracted from each article to constitute the subdata of both languages. Subsequently, the Chinese subdata were segmented into words using the software Yacsi 0.96. Step 5 The texts were manually annotated through careful reading of the data. The annotation consists of two parts. First, metadiscourse units were identified and classified into personal or impersonal based on whether their realization structure contains pronominal forms or nouns referring to the current author and reader of the text (see Section 3) or not in accordance with the four criteria mentioned in Section 2. Then the identified units were further categorized according to the discourse functions they serve by following the analytical framework in Section 3. To ensure the reliability of annotation, half of the data were first annotated by the authors and the inter-coder agreement rate was 0.89. Based on a discussion about the between-coder discrepancies, the coding scheme was revised and the analytical framework (see Section 3) was constructed. Then the first author annotated the rest data twice with a one-month interval in between. This process produced 38 discrepancies which were resolved through careful discussion between the authors. Step 6 Raw frequencies of metadiscourse units in the subdata of both languages were counted. Independent samples t-test was conducted to determine any significant differences in the frequency of metadiscourse across the subdata. The significance level was established at <0.05. 5. Results and discussion A total of 1648 metadiscourse units were identified, of which 716 were Chinese and 932 were English. Around 141 and 118 occurrences of reflexive metadiscourse per 10,000 words appeared in the Chinese and English subdata respectively, rez-Llantada’s (2010) and Salas' (2015) findings of a very low occurrence of metadiscourse in RAs under the consistent with Pe reflexive framework but in sharp contrast with Mu et al.’s (2015) finding of 240.3 occurrences per 10, 000 Chinese characters and 647.8 occurrences per 10, 000 English words under the interactive framework. This variation in the occurrences of metadiscourse is readily understandable considering the more rigorous delimitation of metadiscourse under the reflexive approach. For instance, most of the evidentials, transitions, hedges and boosters which occurred the most frequently in Mu et al.’s (2015) study were not tallied as metadiscourse in our study. The proportion of personal and impersonal metadiscourse is shown in Fig. 1. We can see that in both languages, impersonal metadiscourse was employed far more frequently than personal metadiscourse. This proportional discrepancy is even more marked in Chinese than in English, indicating that Chinese writers tend to use personal metadiscourse very cautiously.

Fig. 1. Metadiscourse in Chinese and English subdata.

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

53

5.1. Similarities in metadiscourse use across languages and sections and influencing factors The t-test results on the frequency of metadiscourse between the subdata are displayed in Table 4 (for introductions) and Table 5 (for discussions). It can be seen from the mean value that in both languages, impersonal metadiscourse remarkably outnumbered personal one in both sections. This is in line with Zhang’s (2016) finding that metadiscourse in academic writing has the prominent function of presenting the text. In both sections, the distribution of the four subcategories of impersonal metadiscourse displays similarity in both languages, i.e., the subcategory of Discourse Labels and Phorics were predominant, together accounting for around 70% of impersonal metadiscourse or approximately 60% of all metadiscourse, followed by Code Glosses and Reference to Text/code. The above finding about the more frequent occurrence of impersonal metadisocourse compared to personal one in both languages shows that in this particular genre of academic writing, writers prefer to communicate with readers in an € impersonal rather than personal way and this is consistent with Adel's (2006) and Zhang's (2016) findings. This predominance of impersonal metadiscourse, can be accounted for by the nature of RAs, the chief role of metadiscourse as “guiding the reading process by indicating discourse organization and clarifying propositional connections and meanings” (Hyland, 2005:92) and their association with the establishment of coherence and logic in academic rhetoric (Mauranen, 1993b). Mauranen (1993a:4) pointed out that the universal aspects of academic writing tend to be constrained by genre, whereas the more changeable aspects are shaped by rhetoric and further assumed that some rhetorical means are similar in various writing cultures but their preferred uses and frequencies may differ. As one informational genre with empirical evidence, abstract concepts and formal logic, academic genre, particularly RAs, need more metadiscourse than other genres to be more reader-friendly, helping readers read and comprehend more easily (Zhang, 2016). The relative distribution of impersonal and personal metadiscourse seems to be one of those shared aspects of the RA genre in different languages and cultures (here in Table 4 Metadiscourse use in Chinese and English introductions.

Personal

Subcategory

Group

Mean

SD

t

df

Sig.

Meta-text

Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English

0.17 1.3 0.57 0.83 0.73 2.13 1.47 1.07 2.73 2.63 3.03 3.2 1.47 1.13 8.7 8.03 9.43 10.17

0.461 1.685 0.935 1.464 1.143 2.315 1.655 1.258 3.248 2.593 2.684 2.041 1.525 1.456 6.545 5.102 7.001 6.701

3.554

33.322

0.001

0.841

58

0.404

2.970

42.336

0.005

1.054

58

0.296

0.132

58

0.896

0.271

58

0.788

0.866

58

0.390

0.440

58

0.662

0.414

58

0.680

Writer-reader Interaction Subtotal Impersonal

Reference to text/code Phorics Discourse Labels Code Glosses Subtotal Total

Table 5 Metadiscourse use in Chinese and English discussions.

Personal

Subcategory

Group

Mean

SD

t

df

Sig.

Meta-text

Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English Chinese English

0.27 1.57 0.6 1.67 0.87 3.23 2.13 1.53 4.4 5.27 4.7 8 2.33 2.87 13.57 17.67 14.43 20.9

0.521 2.096 1.276 2.845 1.252 3.655 1.925 1.852 3.349 3.205 2.351 4.291 2.057 2.209 6.050 6.718 6.084 8.104

3.297

32.569

0.002

1.874

58

0.066

3.355

35.715

0.002

1.230

58

0.224

1.024

58

0.310

3.694

44.972

0.001

0.968

58

0.337

2.484

58

0.016

3.495

53.807

0.001

Writer-reader Interaction Subtotal Impersonal

Reference to text/code Phorics Discourse Labels Code Glosses Subtotal Total

54

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

English and Chinese ones) as well as in different disciplines as evidenced by Salas (2015). Additionally, the frequent occurrence of Discourse Labels and Phorics, especially their subcategories such as Exemplifyings and Enumerators serve the navigating and connecting functions to guide readers through the text and meet the requirement of explicit and logic expression of ideas entailed by the informational, interpretive and persuasive nature of the RA genre so that they are more obviously favored, regardless of the language/culture setting or analytical framework. 5.2. Divergences in metadiscourse use across languages and sections and influencing factors In both sections, the use of metadiscourse as a whole was more frequent in English than in Chinese, but this difference was only statistically significant in the discussion section. This heavier reliance on metadiscourse in English than in Chinese in sociology aligns with the results of Mu et al’s (2015) finding concerning the discipline of applied linguistics, implying that there appears to be no discipline-specific variation in terms of the rhetorical construction of knowledge by means of metadiscourse in totality. Also, the English subdata preponderate in the use of both personal and impersonal metadiscourse and this preponderance is significant in the personal category, particularly in the Meta-text subcategory, in both sections and in the impersonal category in discussions as well. However, the gap between impersonal and personal metadiscourse use was notably more salient in Chinese than in English as indicated by their mean values, suggesting that Chinese writers show a preference for commenting on their text/language over interacting with their readers by means of reflexive metadiscourse compared with their English colleagues, thereby contributing to an overall impression of a more objective and detached style in Chinese than in English. The above discrepancy in the frequency of metadiscourse in English and Chinese RAs seems to indicate that in both languages/cultures somewhat different roles were assigned to the writer, the text and the reader in academic communication. The more affluent employment of metadiscourse in English appears to mark a rhetorical style of English writers which makes the reader's role easy to ensure successful communication. This is analogous to Hinds' (1987) dichotomy between “writerresponsible” and “reader-responsible” writing styles in various cultures. In a writer-responsible culture (e.g., English), it is the writer who primarily takes the responsibility for effective communication, whereas in a reader-responsible culture (e.g., Japanese), the responsibility for effective communication mainly lies with the reader (Hinds' (1987):143). English writers' more frequent use of metadiscursive devices assists in a more explicit organization of the text, a clearer orientation of and interaction with readers. This rhetorical style may correlate with the English language's trait of being hypotactic (Lian, 2010:73). English emphasizes “form over meaning” (Tse, 2010:355) and is inflective. As a result, discourse relations in English tend to rely on explicit coherence. To achieve this, writers take on the responsibility to make clearly organized statements, establish transparent relationship between ideas for readers by deploying a wealth of Metadiscourse, a major tool for being a more explicit writer, and thus seems to more directly guide readers' interpretation of writers' intention. By contrast, Chinese writers' relatively infrequent employment of metadiscourse indicates that they make limited effort for clear organization of the text, explicit guidance of and intimate interaction with readers, suggesting a characteristic implicitness of Chinese rhetorical strategy which demands considerable interpretative effort from readers. In view of this, it appears that Chinese writing falls under the “reader-responsible” style. This is consistent with the linguistic characteristics of Chinese being markedly paratactic (Lian, 2010:73). Chinese emphasizes “meaning over form” (Tse, 2010:355) and words, sentences, etc. are mainly juxtaposed by implicit coherence rather than explicit cohesive ties. Furthermore, distinct from Indo-European languages, lexical linkage in Chinese is semanticity – rather than inflection/declension-indexed. Consequently, discourse relations in Chinese are not so explicitly marked out. When writing in Chinese, Chinese writers may assume their readers understand their texts not so much on the basis of explicit discourse marking as implicit semantics. Hence, they do not necessarily need to employ a wealth of metadiscourse. This may also help explain the absence or a very modest presence of some subcategories of personal and impersonal metadiscourse in both sections in Chinese. Additionally, the above Chinese and English rhetorical styles might reflect distinct assumptions of shared knowledge in communication within the academic community. Chinese writers' weak intention of orientating readers' interpretation process seems to indicate that considerable shared knowledge is assumed between writers and readers. This assumption is legitimate when target readers are comparatively homogeneous. In the Chinese cultural context, such an assumption sounds reasonable, since the entire community is local, Chinese-dominant, relatively isolated (e.g., by the language for publication) and homogeneous in most important social respects (e.g., education) compared with other societies around the world. Consequently, it is not strange for writing conventions to be relatively implicit. In the broad Anglo-American English academic context which is open to the whole world and culturally more heterogeneous (Mauranen, 1993a), relatively less shared knowledge can be reasonably assumed between writers and readers. It is thus imperative to follow more explicit writing conventions which leave finite room for readers’ possible misinterpretation of what is intended. The text length may also contribute to the above cross-linguistic discrepancy in the use of metadiscourse. With regard to the average text length measured by words and by the number of sentences per text, English introductions and discussions in € the present study are much longer than Chinese ones respectively (see Table 1). As Adel (2006:102) pointed out, “The longer a text, the more opportunities there are for the reader to get lost in the text, so the greater the need for explicit signposting”. Paltridge and Starfield (2007:89) also attributed thesis writers' greater use of metadiscourse to “much lengthier” text characteristic of the thesis genre. Therefore, the much greater length of English texts as compared to Chinese ones probably entails the necessity for English writers to guide readers through the text more explicitly and interact with readers frequently via more metadiscourse.

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

55

The cross-linguistic difference in personal metadiscourse in introductions can be mainly attributed to the subcategory of Introducing Topic (t ¼ -2.384, p ¼ 0.023) and Arguing (t ¼ 3.067, p ¼ 0.004) whereas in discussions this difference widely lies in Introducing Topic (t ¼ -2.262, p ¼ 0.031), Focusing (t ¼ -2.262, p ¼ 0.031), Reminding (t ¼ -2.297, p ¼ 0.028), Anticipating Readers’ Reaction (t ¼ -3.968, p ¼ 0.000), with Chinese using far fewer of them than English. Among the nine subcategories of Meta-text metadiscourse, only three (Introducing Topic, Defining and Saying, and Arguing) were identified in Chinese introductions and four (i.e., Saying, Reminding, Concluding, and Arguing) in Chinese discussions, both with very sparse cases, but six in English introductions and seven in English discussions with relatively more occurrences. This indicates that Chinese writers tend to use more limited subcategories of Meta-text in both sections. Furthermore, almost all Meta-text instances in English introductions and discussions were in the form of singular first-person-pronoun units whereas most in Chinese introductions were of third-person-noun units and most in Chinese discussions of plural first-person-pronoun units. Almost all of the Writer-reader Interaction metadiscourse was in the form of plural first-person-pronoun units in both Chinese and English introductions and in Chinese discussions but nearly one-third of the Writer-reader Interaction metadiscourse was represented by singular first-person-pronoun units and the rest by plural first-person-pronoun units in English discussions. Such variations in the use of personal metadiscourse suggests that Chinese writers are much less visible in the text than their English colleagues. Although all RAs were single-authored, instead of using singular first-person-pronoun units to present themselves as explicitly visible individuals, Chinese writers, as (1) and (2) exemplify, used plural first-personpronoun units or more often inanimate forms which convey lower levels of authorial intervention and personal voice than singular first-person-pronouns (Harwood, 2005a) to play down ownership and responsibility for their statements. Furthermore, such personal usage is very limited in that it only appeared in one third of the Chinese subdata, indicating that just a small number of authors chose to visibly present themselves. Additionally, Chinese writers prefer to hide themselves even when introducing topics, as evidenced by the fact that in introductions, the personal subcategory of Introducing Topic was used significantly less frequently in Chinese than in English, whereas the impersonal one was used vice versa (as shown later). All these show that Chinese sociologists tend to adopt the role of solitary writers who focus on the propositional content of the text and downplay their role as active participants in the text. This finding is contradictory to Mu et al’s (2015) finding that Chinese applied linguists tend to use more self-mentions than English applied linguists, suggesting crossdisciplinary variation in authorial visibility across Chinese and English academic communities. In contrast, English writers' frequent signaling of their personal presence in the text, along with their explicit and careful guidance of readers realized by the wealthy impersonal metadiscourse, creates the impression that they act as guides and invite readers to take a tour of the text together with them, playing the role of marketing writers.

(1) 这 是 我们 尝试 在此 回答 的 一个 问题。(C_INTRO203) [This is a question we are attempting to answer here.] (2) 本文 最 关注 的 问题 是 城乡 教育 机会 不平等 的 变化 趋势。 (C-DISC03) [The issue that this article is most concerned with is the trend of education opportunity inequality for urban and rural children.] (3) The article proceeds as follows. First, I develop …. Second, I present … Finally, I demonstrate …. (E-INTRO02)

We hold here that traditional academic conventions and some socio-cultural factors may be responsible for the above phenomenon. The Chinese National Criteria of Rules for Abstract Writing published in 1986 and the submission requirements of many Chinese journals suggest that academic writing should be objective and some even stipulate that first-person pronouns be avoided to give room for the third-person references. Also, writing, especially academic writing, has been a neglected skill in college education in China in the past decades and a majority of teachers do not know how to teach academic writing (Zhang, 2017: 208). They may still follow some outdated teaching materials and encourage novice writers to formally remove themselves from their texts. These academic conventions may subliminally affect Chinese scholars’ choice of linguistic strategies, accounting for their reluctance to resort to personal means so that an impression of objectivity can be strengthened by subordinating their own voice to that of their results. Chinese writers' avoidance of personal voice may also be derived from the Chinese collective culture where the authority of the group or organization an individual is affiliated to is stressed over that of the individual. Chinese writers may feel uncomfortable with the individualistic authority, subjectivity, and personal commitment that personal metadiscourse implies since they seem to clash with Chinese beliefs and practices that value more collectivist forms of self-representation as indicated in previous studies on Chinese students' English writing (e.g., Ouyang and Tang, 2006). Scollon (1994) also noted that the use of first-person pronouns was largely unacceptable in Asian cultures due to its connection with individual rather than collective identity. Additionally, the Confucius' teaching principles of Ren (humanity and love) and Li (propriety) (Lu, 2000) may have also influenced the rhetoric of Chinese writing. According to these principles, the presence of others (audience/readers) must be accepted (Loi and Sweetnam Evans, 2010). To this end, mutual communication is critical in the

3 This code indicates the source of the example sentence in front of the parenthesis. The first letter represents the language, “C” for Chinese and “E” for English. The following letters refer to the RA section with “INTRO” for introductions and “DISC” for discussions. The last two digits are the sequence number of the text in the subdata. Therefore, “C_INTRO20” indicates the twentieth text in the subdata of Chinese RA introductions.

56

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

rhetorical construction of a text to develop interpersonal relationships. Such relationships can be set up when the reader infers the writer's intentions via cooperative interpretation of the text. To leave room for the reader to interpret the text in accordance with their prior knowledge and cognition, the text is expected to be reader-responsible to some extent and a low degree of the writer's visibility in the text is thus anticipated. In contrast, English writers exhibit greater writer visibility by explicitly presenting their writer persona in singular firstperson-pronoun units (as illustrated in (3)) in around two thirds of the English subdata. Authors’ personal visibility in their texts helps to establish a credible scholarly identity and underline what they have to say (Ivanic, 1998). For this reason, as mentioned above, English writers use meta-text metadiscourse mainly to state arguments, introduce topics or remind readers of what they have mentioned in the previous text. The use of personal metadiscourse in this way helps promote an impression of confidence and authority and clearly link authors with their main contribution, which is congruent with the individualistic Anglo-American culture where personal originality and authority are valued. If we compare the mean values of impersonal subcategories as displayed in Tables 4 and 5, we can see that in both languages all impersonal categories were much more frequently employed in discussions than in introductions. The t-test results provide further detailed depiction of these subcategories. Specifically, in introductions, although no conspicuous cross-linguistic discrepancies were identified in the use of all impersonal categories, significant cross-linguistic differences occurred in two subcategories of Discourse Labels: Exemplifying (t ¼ -2.611, p ¼ 0.012) and Introducing Topic (t ¼ 3.703, p ¼ 0.001). In other words, English tend to use noticeably more impersonal Exemplifyings than Chinese but Chinese are inclined to use more impersonal Introducing Topics than English. Meanwhile, it is noteworthy that the cross-linguistic difference in the counts of impersonal Introducing Topic is in stark contrast with that in the corresponding personal subcategory as aforementioned. In discussions, cross-linguistic differences were remarkably present in the frequency of Discourse Labels (t ¼ -3.694, p ¼ 0.001). Specifically, Exemplifyings (t ¼ -6.288, p ¼ 0.000) display even more pronounced cross-linguistic divergency than in introductions. In addition, although no significant cross-linguistic difference was identified in the use of Phorics, its subcategories Previews (t ¼ 3.101, p ¼ 0.004) and Reviews (t ¼ 2.299, p ¼ 0.025) appeared conspicuously more frequently in Chinese than in English while Textual Deixis (t ¼ -4.089, p ¼ 0.000) were more favored in English than in Chinese. The above findings not only show a discernible divergence in impersonal metadiscourse across introductions and discussions but also the atypical occurrence of some subcategories of impersonal metadiscourse, i.e., Previews, Reviews and Textual Deixis, in that the first two are the only subcategories which were preferred by Chinese while the last one exclusively by English in discussions. This atypical distribution of Previews and Reviews in the subdata is in conflict with the significantly less frequent appearance of endophoric markers (which almost correspond to the combination of Previews and Reviews in our research) in Chinese than in English in applied linguistics as reported by Mu et al. (2015), indicating another disciplinary difference in metadiscourse use across Chinese and English cultures. Such differences in impersonal metadiscourse between the two sections may be correlated to the rhetorical functions of introductions and discussions. Introductions generally justify the current study by critically reviewing previous literature and showing how it is to extend previous research or fill gaps in previous literature (e.g., Swales, 1990; Paltridge and Starfield, 2007). Discussions mainly function to respond to the questions posed in the introduction and to comment on the results by interpreting, explaining, and comparing them with previous work (Yang and Allison, 2003). It is thus a floor for authors to situate and/or claim their research findings, support and justify their claims by the relevant literature from a scientific perspective (Kwan and Chan, 2014). Accordingly, writers have the opportunity of running more words to interpret their findings and justify their argumentations (Swales, 1990) – the overwhelmingly greater length of discussions than introductions in both languages of the present data is good support of this. Therefore, compared with introductions, discussions are more persuasive in that a greater investment of the writers’ persona takes place here (Milagros del Saz Rubio, 2011) and “results and interpretations need to be presented in ways that readers are likely to find persuasive” (Hyland, 2005:176). Hence it is not surprising that in both languages remarkably more impersonal metadiscourse were employed to fulfill the interpretive and argumentative functions of discussions in the present study. On the other hand, English writers' outstanding advantage in impersonal metadiscourse, particularly in the subcategory of Discourse Labels, in discussions, may have some bearing on the special readership and high degree of publishing competitiveness of the English publications. Writers who publish in these international publications have to address a presumably broader and more heterogeneous readership who may not have much shared knowledge with them. Furthermore, the English international context may be viewed as a rather ‘urban’ (Becher and Trowler, 2001) disciplinary community which is close knit in terms of research topics and problems addressed and thus writers aiming to publish here face fierce competition. To cope with this, English writers favored the use of impersonal metadiscourse, especially discourse labels, in discussions, to interpret and justify their findings, arguments and claims to a greater extent than Chinese in order to persuade their readers. Contrarily, the Chinese national ~ as, 2011) disciplinary community which is loose knit in terms of context may be interpreted as a relatively ‘rural’ (Mur-Duen problems and issues addressed and has homogenous readership. Chinese writers do not confront so much competition as international English peers do. They therefore may not feel pressured to explicitly interpret and justify their results and arguments via the application of impersonal metadiscourse to the same extent as their English colleagues. As for the interestingly atypical occurrence of subcategories of Previews, Reviews and Textual Deixis, we consider it may be attributed to the specific functions of these subcategories along with the writing patterns in the two contexts. These three subcategories all fall under Phorics or Crismore and Farnsworth’s (1990) category of “commentary”. Drawing on anaphoric and cataphoric references, Preview and Reviews explicitly orientate readers' attention to a point which has already been mentioned (see (4)) and which will be discussed later (see (5)) respectively, thereby are strongly retrospective and

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

57

prospective. The significantly greater use of Previews and Reviews in Chinese renders argumentation less linear than in English. Previously presented hypothesis, findings, arguments or other ideas are more frequently resumed and new or related ideas or arguments (mainly conclusions or elaboration of arguments or claims) are more frequently introduced in Chinese than in English. This, on one hand, aligns with the interpretive and argumentative characteristics of the discussion section, and on the other hand supports the “spiral” hypothesis that in Chinese writing a topic isn't unfolded in a straightforward way but often diverges before finally converges back to the central point (Jia, 2000:391). The fewer presence of these preview and review metadiscourse markers in English, however, probably lends credence to the preferred linear pattern of English writing which usually unfolds in a way of staying close to the main subject, with the main statement and supporting statements following a hierarchical arrangement by resorting to other subcategories (e.g., Exemplifying, Textual Deixis, as illustrated in (6)) more frequently than in Chinese. Textual Deixis (e.g., here) refer to “the current text, a point in the current text, or the € situation of writing” (Adel, 2006:104). By referring to selected previous parts of the current text, although sometimes the referent is ambiguous as illustrated in (6), Textual Deixis shorten the distance and establish an essential connection between the writer and the writing situation as well as the reader. English writers' exclusive adoption of this metadiscourse reinforces the interaction among the writer, the text and the reader, contributing to the writer-responsible writing convention and interpretive and persuasive function of the discussion section, while the absence of this subcategory of metadiscourse in the Chinese subdata demonstrates that Chinese texts are greatly disconnected from the reader and the writing situation, further supporting the finding that Chinese are less explicit in displaying the organization of their texts.

(4) 上述 变化 趋势 证实 了 笔者 在 本文 开头 时 提出 的 猜测 …。(C-DISC03) [The above changes verify the speculation that the author raised at the beginning of this article ...] (5) 通过 对 儿童 校外 教育 消费 数据 的 分析, 本 研究 得出 如下 结论…。(C_DISC18) [Through the analysis of the data on children's out-of-school education consumption, this study draws the following conclusions….] (6) … A particular concern here is the possibility that respondents may downplay heterogeneity d for instance, if a white respondent reports that his or her partner is white when the partner would describe himself or herself as Hispanic. (E_DISC03)

6. Conclusion and implications The above reported research sets up taxonomy of reflexive metadiscourse for the academic genre of RAs in Chinese and English and provides further evidence for cross-linguistic/cultural comparison of metadiscourse use across Chinese and English RAs. Specifically, two similarities in metadiscourse use were identified across Chinese and English sociology RAs. First, more impersonal than personal metadiscourse was deployed across the two languages for the purpose of explicit and logic idea conveying. Second, the distribution of the impersonal subcategories displays similarity in both languages, with Discourse Labels and Phorics being predominant, followed by Code Glosses and Reference to Text/code. Despite these similarities, significant cross-linguistic/cultural differences were also observed. English sociologists tend to include more Metadiscourse as a whole, the respective personal and impersonal metadiscourse, particularly personal metadiscourse and in the discussion section in their RAs, than their Chinese colleagues. Additionally, the gap between impersonal and personal metadiscourse use was notably more salient in Chinese than in English in both sections These differences indicate that within the disciplinary community of sociology English and Chinese scholars respond to their readers’ diverging expectations and needs in distinct ways, coordinating their writing with the prevalent conventions in each context. Meanwhile, these results further support rez-Llantada, 2010; Mur-Duen ~ as, 2011), previous cross-linguistic/cultural studies (e.g.; Mauranen, 1993a; Molino, 2010; Pe which show that the extent of metadiscourse use is conditioned by the language and broad socio-cultural context where texts are written and published. This research also indicates the same favor of metadiscourse use in totality by English scholars compared with Chinese ones but different preferences of some specific metadiscourse subcategories by Chinese and English across disciplines like sociology and applied linguistics. Given the findings reported above, this study has implications for EFL/ESL writing and EAP teaching. Our findings indicate that the divergent rhetorical construction by means of metadiscourse across Chinese and English RAs may arise from some linguistic, sociocultural and rhetorical factors. Therefore, EFL/ESL scholars in China and similar contexts (especially those in sociology) should take caution when trying to get their research published in international English journals and addressing an international readership. They may need to, at least partially, adjust their writing conventions, especially rhetorical and discursive practices, according to what English academic writing should be like in those contexts since that is the norm western or more specifically Anglo-American traditions expect of academic writers (Mu et al., 2015), so as to facilitate the publication of their research. Founded on this, future intercultural/language studies could explore whether or to what extent non-compliance with the rhetorical/discursive norms of the international community may hinder EFL/ESL scholars from publishing in specific disciplinary fields. In terms of EAP teaching, academic writing teachers should teach novice EFL/ESL writers the varied rhetorical construction of the different sociology (and other related disciplines) RA sections by means of metadiscursive devices across Chinese (or other associated languages) and English along with the causes underlying these variations. Such a cross-cultural/linguistic pedagogy could help enhance novice writers’ cross-cultural (meta)discursive

58

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

awareness and better their understanding of conventions and disciplinary cultures in the international English academic community so that they can make informed choices about (meta)discursive strategies for successful and effective English writing in the future. Despite the above claims, we need to admit the limitations of our study. Our corpus was restricted to 60 RAs from the single discipline of sociology. Therefore, the results cannot be extrapolated to the whole academic culture or to other disciplinary fields and should be taken cautiously. Future research based on a larger corpus will help verify our findings. Acknowledgements Sincere gratitude goes to the editor and anonymous reviewers of the Journal of Pragmatics as well as Prof. Hou, Guojin and Huang, Xiaoping from Huaqiao University, China for their valuable advice about the revision of this article. This work was supported by the Social Science Research Foundation of Fujian Province, China (Project No. FJ2016B263), The Innovative Pragma-rhetoric Team Program of Huaqiao University (Project No. 2018007) and Apple-Reading Seminar of Huaqiao University (Project No. 18YJG07). References € Adel, A., 2006. Metadiscourse in L1 and L2 English. John Benjamins, Amsterdam. € Adel, A., 2010. “Just to give you kind of a map of where we are going”: a taxonomy of metadiscourse in spoken and written academic English. Nord. J. Engl. Stud. 9 (2), 69e97. € Adel, A., 2017. Remember that your reader cannot read your mind: problem/solution-oriented metadiscourse in teacher feedback on student writing. Engl. Specif. Purp. 45, 54e68. Bazerman, C., 1988. Shaping Written Knowledge: the Genre and Activity of the Experimental Article in Science. University of Wisconsin Press, Madison. Becher, T., Trowler, P., 2001. Academic Tribes and Territories: Intellectual Inquiry and the Cultures of Disciplines. Societies for Research in Higher Education/ Open University Press, Milton Keynes. Bruce, I., 2010. Textual and discoursal resources used in the essay genre in sociology and English. J. Engl. Acad. Purp. 9, 153e166. Bruce, I., 2016. Constructing critical stance in university essays in English literature and sociology. Engl. Specif. Purp. 42, 13e25. Cao, F., Hu, G., 2014. Interactive metadiscourse in research articles: a comparative study of paradigmatic and disciplinary influences. J. Pragmat. 66, 15e31. Chen, C., Zhang, L.J., 2016. An intercultural analysis of the use of hedging by Chinese and Anglophone academic English writers. Appl. Ling. Rev. 8 (1), 1e34. Connor, U., Moreno, A.I., 2005. Tertium comparationis: a vital component in contrastive rhetoric research. In: Bruthiaux, P., Atkinson, D., Eggington, W., Grabe, W., Ramanathan, V. (Eds.), Directions in Applied Linguistics: Essays in Honor of Robert B. Kaplan. Multilingual Matters, pp. 153e164. Clevedon, UK. Crismore, A., Farnsworth, R., 1990. Metadiscourse in popular and professional science discourse. In: Nash, W. (Ed.), The Writing Scholar: Studies in Academic Discourse. Sage Publications, Newbury Park/London/New Dehli. Dahl, T., 2004. Textual metadiscourse in research articles: a marker of national culture or of academic discipline? J. Pragmat. 36, 1807e1825. Flowerdew, J., 2015. Revisiting metadiscourse: conceptual and methodological issues concerning signalling nouns. Iberica 29, 15e34. Golebiowski, Z., Liddicoat, A., 2002. The interaction of discipline and culture in academic writing. Aust. Rev. Appl. Ling. 25, 59e71. Harwood, N., 2005a. ‘Nowhere has anyone attempted... In this article I aim to do just that’: a corpus-based study of self-promotional I and we in academic writing across four disciplines. J. Pragmat. 37, 1207e1231. Harwood, N., 2005b. ‘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, 343e375. Hinds, J., 1987. Reader versus writer responsibility: a new typology. In: Connor, U., Kaplan, R.B. (Eds.), Writing across Languages: Analyses of L2 Texts. Addison-Wesley, Reading, pp. 141e152. Hu, G.W., Cao, F., 2011. Hedging and boosting in abstracts of applied linguistics articles: a comparative study of English- and Chinese-medium journals. J. Pragmat. 11, 2795e2809. Hyland, K., 1998a. Hedging in Scientific Research Articles. Benjamins, Amsterdam. Hyland, K., 1998b. Persuasion and context: the pragmatics of academic metadiscourse. J. Pragmat. 30, 437e455. Hyland, K., 2005. Metadiscourse: Exploring Interaction in Writing. Continuum, London. Hyland, K., 2007. Applying a gloss: exemplifying and reformulating in academic discourse. Applied Linguistics 28, 266e285. Hyland, K., 2008. Disciplinary voices: interactions in research writing. Engl. Text Constr. 1 (1), 5e22. Is¸ık-Tas¸, E.E., 2018. Authorial identity in Turkish language and English language research articles in sociology: the role of publication context in academic writers' discourse choices. Engl. Specif. Purp. 49, 26e38. Ivani c, R., 1998. Writing and Identity: the Discoursal Construction of Identity in Academic Writing. John Benjamins, Amsterdam. Jakobson, R., 1998. On language: roman Jakobson. In: Waugh, L.R., Monville-Burston, M. (Eds.). Harvard University Press, Cambridge, MA and London. Jia, Y., 2000. Intercultural Communication. Shanghai Foreign Language Education Press, Shanghai. Kwan, B.S.C., Chan, H., 2014. An investigation of source use in the results and the closing sections of empirical articles in information systems: in search of a functional-semantic citation typology for pedagogical purposes. J. Engl. Acad. Purp. 14, 29e47. Lian, S.N., 2010. Contrastive Studies of English and Chinese. Higher Education Press, Beijing. Loi, C.K., Lim, J.M.H., 2013. Metadiscourse in English and Chinese research article introductions. Discourse Stud. 15, 129e146. Loi, C.K., Sweetnam Evans, M., 2010. Cultural differences in the organization of research article introductions from the field of educational psychology: English and Chinese. J. Pragmat. 42 (10), 2814e2825. Lu, X., 2000. The influence of classical Chinese rhetoric on contemporary Chinese political communication and social relations. In: Heisey, D.R. (Ed.), Chinese Perspective in Rhetoric and Communication. Ablex Publishing Corporation, Stamford, Connecticut, pp. 3e23. Mauranen, A., 1993a. Contrastive ESP rhetoric: metatext in Finnish-English economics texts. Engl. Specif. Purp. 12, 3e22. Mauranen, A., 1993b. Cultural Differences in Academic Rhetoric: A Textlinguistic Study. Peter Lang, Frankfurt. Mauranen, A., 2001. Reflexive academic talk: observations from MICASE. In: Simpson, R., Swales, J. (Eds.), Corpus Linguistics in North America: Selections from the 1999 Symposium. University of Michigan Press, Ann Arbor, MI, pp. 165e178. Milagros del Saz Rubio, M., 2011. A pragmatic approach to the macro-structure and metadiscoursal features of research article introductions in the field of agricultural sciences. Engl. Specif. Purp. 30 (4), 258e271. n, E.L., 2010. ‘Extending this claim, we propose...’: the writer's presence in research articles from different disciplines. Iberica 20, 35e56. Milla Molino, A., 2010. Personal and impersonal authorial references: a contrastive study of English and Italian Linguistics research articles. J. Engl. Acad. Purp. 2, 86e101. Mu, C., Zhang, L.J., Ehrich, J., Hong, H., 2015. The use of metadiscourse for knowledge construction in Chinese and English research articles. J. Engl. Acad. Purp. 20, 135e148.

Z. Li, J. Xu / Journal of Pragmatics 159 (2020) 47e59

59

~ as, P., 2011. An intercultural analysis of metadiscourse features in research articles written in English and in Spanish. J. Pragmat. 43, 3068e3079. Mur-Duen Ouyang, H.H., Tang, S.Y., 2006. The authorial identity in English argumentative essays written by Chinese college students. J. PLA Univ. Foreign Lang. 29 (2), 49e53. Paltridge, B., Starfield, S., 2007. Thesis and Dissertation Writing in a Second Language: a Handbook for Supervisors. Routledge, London. rez-Llantada, C., 2010. The discourse functions of metadiscourse in published academic writing: issues of culture and language. Nord. J. Engl. Stud. 9 (2), Pe 41e68. Salas, M.D., 2015. Reflexive metadiscourse in research articles in Spanish: variation across three disciplines (linguistics, economics and medicine). J. Pragmat. 77, 20e40. Sanderson, T., 2008. Interaction, identity and culture in academic writing: the case of German, British and American academics in the humanities. In: € Adel, A., Reppen, R. (Eds.), Corpra and Discourse: the Challenges of Different Settings. John Benjamins, Philadelphia, pp. 57e92. Scollon, R., 1994. As a matter of fact: the changing ideology of authorship and responsibility in discourse. World Englishes 13, 34e46. Sinclair, J., 2005. Language as a string of beads: discourse and the M-word. In: Tognini-Bonelli, E., Camiciotti, G.D.L. (Eds.), Strategies in Academic Discourse. John Benjamins, Amsterdam, pp. 163e168. Swales, J., 1990. Genre Analysis: English in Academic and Research Settings. Cambridge University Press, Cambridge. Toumi, N., 2009. A model for the investigation of reflexive metadiscourse in research articles. Lang. Stud. Work. Pap. Univ. Read. 1, 64e73. Tse, Y.K., 2010. Parataxis and hypotaxis in the Chinese language. Int. J. Arts Sci. 3 (16), 351e359. Vassileva, I., 2001. Commitment and detachment in English and Bulgarian academic writing. Engl. Specif. Purp. 20, 83e102. Yakhontova, T., 2006. Cultural and disciplinary variation in academic discourse: the issue of influencing factors. J. Engl. Acad. Purp. 2, 153e167. Yang, R., Allison, D., 2003. Research articles in applied linguistics: moving from results to conclusions. Engl. Specif. Purp. 22, 365e387. Yang, Y., 2013. Exploring linguistic and cultural variations in the use of hedges in English and Chinese scientific discourse. J. Pragmat. 50, 23e36. Zarei, G.R., Mansoori, S., 2011. A contrastive study on metadiscourse elements used in humanities vs. non-humanities across Persian and English. Engl. Lang. Teach. 4 (1), 42e50. Zhang, L.J., 2017. Reflections on the pedagogical imports of western practices for professionalizing ESL/EFL writing and writing-teacher education. Aust. Rev. Appl. Ling. 39 (3), 203e232. Zhang, M., 2016. A multidimensional analysis of metadiscourse markers across written registers. Discourse Stud. 18 (2), 204e222. Zhijun Li is an associate professor and postgraduate supervisor at the College of Foreign Languages, Huaqiao University, China. She teaches English for Academic Communication at the graduate level and College Comprehensive English at the undergraduate level. She is mostly interested in academic discourse analysis and corpus linguistics. Her articles on corpus-based academic discourse analysis, cognitive linguistics and second language writing have appeared in both international journals and Chinese key academic journals. Jinfen Xu is a professor and Ph. D supervisor of the School of Foreign Languages, Huazhong University of Science and Technology, one of the top ten universities in China. Her main research interests are foreign language teacher development, EAP writing, foreign language education, second language acquisition, and learner autonomy. Her articles (more than 130 articles in total) on teacher development, second language writing, EFL classroom research, and foreign language learner autonomy have appeared in System (2015, 2018), Language Policy (2016), TESOL Quarterly (2017), Language Awareness (2019) and in many key academic journals (CSSCI) in China.