An intercultural analysis of the use of hedging by Chinese and Anglophone academic English writers

1 Introduction

Recent discussions of English as the international language of scientific communication indicate that academic writers with different first languages (L1s), including writers in English as a foreign language (EFL) or English as a second language (ESL) (hereinafter referred to as L2), are expected to develop competencies in intercultural rhetoric in order for their research findings to be accepted by the Anglo-sphere academic community (e. g., Connor 2008; Martín and Pérez 2014; Moreno 2013; Mu et al. 2015; Mur-Duenas 2011). This is because if they wish to publish their research in English-medium journals, the standard for making judgements on the use of English is usually that of how competent they are in using the language for conveying their ideas or arguments (Gea-Valor et al. 2014; Moreno 2010, 2013). Neither reviewers nor editors of academic journals typically show sympathy with L2 English writers and make decisions in their favour because of the fact that the authors are L2 English writers. Specifically, it is essential that L2 academic writers improve their L2 pragmatic competence and make rhetorical and stylistic accommodations (Connor 2004, 2008, 2011; Moreno 2013) to satisfy the academic English discourse norms and the requirements of English-medium journals for facilitating their readers’ understanding of their intended meanings accordingly.

Hedging strategies, among other things, are one of the vital indicators of L2 pragmatic competence (Hyland 1996; Skelton 1988a), which merit our attention. The use of hedge as a linguistic term traces back at least to Lakoff’s (1973) work, in which hedge refers to “words whose job is to make things fuzzier or less fuzzy” (p. 195). According to Lakoff, hedging is a means for distinguishing the more marginal members of the conceptual categories of language from the central items. Since Lakoff’s early work, the connotation of the concept of hedge itself became more abundant and different views took shape from perspectives such as “politeness” (Brown and Levinson 1987) and “linguistic vagueness” (Channell 1994). Given the fact that “scientific texts are not only content-oriented and informative but also seek to convince and influence their audience” (Martin-Martin 2008: 135), hedges are frequently used in academic discourse as sociopragmatic rhetorical strategies (Hyland 1996, 2016; Martin-Martin 2008). When academic writers employ hedges, they aim to achieve a variety of important pragmatic purposes, for example, modulating their commitment to a proposition, opening up a dialogic space, acknowledging subjectivity, or mitigating face-threatening criticisms, among others. In short, hedging is part of a writer’s pragmatic competence, the lack of which would result in his/her pragmatic failure in written communication.

Pragmatic competence, as used in our study, refers to the writer’s ability to use language appropriately in dynamic contexts to produce an academic text which can meet the rhetorical norms and expectations of a discourse community. According to Chen (2009), pragmatic competence consists of: 1) pragmalinguistic competence, 2) sociopragmatic competence, 3) pragmacognitive competence, and 4) discoursal competence. Given the focus of our study as well as space constraints, we concentrate on the first two dimensions of pragmatic competence. Pragmalinguistic competence, in our view, means that pragmatically competent writers must have at their disposal necessary linguistic resources for choice in academic writing. Sociopragmatically competent writers should also be sensitive to the culturally-based rhetorical conventions in academic writing. Thus L2 academic writers need to attach great importance to hedging strategies, which are particularly useful for promoting their L2 pragmalinguistic and sociopragmatic competence so as to achieve their communication goals interculturally (Moreno 2013).

Pragmatic competence is also represented in the way that different sections of research articles (RAs) embody different rhetorical functions with various linguistic resources (Abdollahzadeh 2011; Salager-Meyer 1994). Studies have begun to examine these functions of text units, as reported in Moreno (2013). Scholars have found that the conclusion section of a RA boasts important rhetorical functions, where, typically, the author’s persuasive text is intended to be interpretative rather than factual statements (Abdollahzadeh 2011; Swales 1990, 2004; Swales and Feak 2004). The conclusion section is usually where authors are provided with an opportunity to demonstrate the significance of the findings in the study (e. g., theoretical and/or practical implications) (Chen 2012). It often involves what the research results might imply: Do they confirm or refute a theory, or modify existing theories, or clarify a controversial issue, or develop a new model, or enrich our understanding of a phenomenon? The practical implications are also included in this section if the results can be applied to solve certain practical problems. Briefly, the conclusion section, as the essence and destination of RAs, is an overall judgment and evaluation of research results, functioning as a finishing touch for the whole text (Qi and Yang 1999: 109). Thus, authors employ hedging strategies to achieve their intended purposes because appropriate uses of hedging strategies could make authors better adjust their relationships with their own propositions and arguments, and help construct certain power relations between them and readers as well as reviewers (Mu et al. 2015). Therefore, if the authors hedge properly, expressing their claims and arguments tentatively or in a less assertive manner, it is more likely that their conclusions would be accepted by a specific academic discourse community. Otherwise, their concluding remarks might sound blunt or even impolite as to incur readers’, editors’, as well as peer reviewers’ doubt, oppositions, and/or dissatisfaction. As a result, authors would not be able to realize their intercultural communication goals effectively.

Abdollahzadeh (2011) conducted a comparative study of interpersonal metadiscourse in the conclusion section of applied linguistics RAs written in English by Anglo-American and Iranian academic writers. He reported that native writers of English and L2 writers with Arabic as their L1 are not different in terms of using hedges. Hedging the arguments, for these two groups of writers, serves as a strategic means for obtaining academic community acceptance and solidarity with readers. In contrast, Yagiz and Demir (2014) found some significant differences in the use of hedging strategies in published articles between native writers of English and L2 writers with Turkish as their L1. These differences are not so much in the quantity as in the manner in which hedges were used (see also Mu et al. 2015). They concluded that “in many cases L2 writers do not employ hedging structures in tune with the rules and norms of Anglo-sphere academic discourse community” (p. 267).

What about the use of hedging devices by L2 English writers with Chinese as their L1 in contrast to those by native English-speaking writers? Can L2 English writers communicate interculturally in an effective manner? Unfortunately, only a few sporadic studies are documented about L2 English writers with Chinese as their L1 (Bunton 1999; Cheng and Steffensen 1996; Hyland 2005). For instance, Feng and Zhou (2007), Hu and Cao (2011), and Wang (2008) focused on the use of hedging in the abstracts of RAs, and Jiang and Tao (2007) on that in the discussion section of medical RAs. A very recent study by Mu et al. (2015) investigated the use of hedging devices in the whole body of applied linguistics RAs. Understandably, these studies have helped us how to understand RAs. Nevertheless, up to now we have not found any research on how the L2 English writers with Chinese as their L1 use hedging strategies in the conclusion section of RAs. Given the significance of the conclusion section of RAs for advancing the authors’ arguments, our study was conducted to fill the research gap. Specifically, we explored if there would be similarities and/or differences between Chinese academics and their Anglophone counterparts in the use of hedges in the conclusion section of applied linguistics RAs written in English. We concentrated on three research questions.

1. What are the similarities and differences (if any) in the frequency of use of various categories of hedges between Chinese and Anglophone academic writers?

2. What are the similarities and differences (if any) in the pattern in the choice of various categories of hedges between Chinese and Anglophone academic writers?

3. What are the similarities and differences (if any) in the linguistic expressions of various categories of hedges between Chinese and Anglophone academic writers?

It is our hope that our research findings will enable us to offer some implications for developing L2 writers’ academic writing skills and enhancing their L2 pragmatic competence.

2 Taxonomies of hedges

There are various taxonomies in the existing literature and each has its own merits and weaknesses. On the basis of Lakoff’s (1973) study, Prince et al. (1982), through empirical research on oral medical discourse, classify hedges according to their functions. They divide hedges into two categories: Approximators and shields. Approximators refer to hedges that affect the proposition content, whereas shields refer to hedges that implicate that the speaker is not fully committed to the truth of the proposition. Shields fall into two subcategories: plausibility shields and attribution shields. Plausibility shields implicate that the speaker is asserting a belief acquired via plausible reasoning (e. g., I think, seem, probably), whereas attribution shields simply attribute the belief in question to someone other than the speaker (e. g., according to Dr. Smith). Their classification criteria have been questioned by some scholars. As Skelton (1988b) points out, Prince et al.’s classification seems possible only in abstracts, because in the process of language use the illocutionary force of shields may penetrate through longer stretches of discourse, thus transforming approximators into shields. Although his criticism needs further discussion, Skelton suggests that Prince et al.’s classification is at least debatable (Varttala 1999). Clearly, the social purpose of the medical conversations is to make an accurate diagnosis, so the classification of hedges according to function is inevitably affected by the nature of the corpus. For instance, attribution shields are a means of conferring status to reported speech (e. g., within the hierarchy of “doctors, nurses, and patients”). As Prince et al. (1982) put it in their discussion of attribution shields, the speaker’s own degree of responsibility for the truth of a proposition is “only indirectly inferable” (p. 89). Skelton (1988b) also notes that it appears difficult to determine whether a reporting verb expresses a comment (i. e., hedge) or just a report. Admittedly, it is sometimes argued that Hyland’s work (1996, 1998) would be the “locus classicus” on hedging in scientific RAs, providing the definitional and other baselines of hedges for further study. Regrettably, attribution shields in Prince et al.’s (1982) terminology were not excluded in Hyland’s (1998: 124) work. Such a drawback in Hyland’s work led us to a decision against using Hyland’s taxonomy as our conceptual framework despite its popularity in the field of academic writing.

In addition to the purely functional classification, there exist other taxonomies of hedges. Salager-Meyer (1994: 154–155), by the contextual analysis of written medical English discourse, establishes the formal-functional classification, in which hedges are divided into five categories (summarized as follows):

1. shields: all modal verbs expressing possibility (might), semi-auxiliaries (seem, appear), probability adverbs (probably), and their derivative adjectives, epistemic verbs (suggest, speculate);

2. approximators: (approximately, somewhat, often, occasionally);

3. expressions of the authors’ personal doubt and direct involvement (I believe, to our knowledge);

4. emotionally-charged intensifiers (extremely difficult, absolutely interesting, of particular importance, surprisingly);

5. compound hedges (It may suggest that …, It would seem likely that …).

Also, her approach appears to have some questionable areas. As we can see, there are apparent overlaps between the categories. For example, concerning the third category of hedges, it seems that only expressions including a direct mention of the author can be used to express his/her personal doubt and direct involvement. However, studies (e. g., Lachowicz 1981) have shown that the author is not necessarily referred to directly so as to make his/her doubt or involvement known to us all. Thus, many expressions in Category (1) “shields” (e. g., appear, suggest and probably) can also pertain to Category (3) in certain contexts (Varttala 1999).

Based on a full analysis of various definitions and classifications of hedges, Crompton (1997) proposes his own definition and taxonomy of hedges. With reference to Lyons’s definition of epistemic modality (1977: 797), Crompton defines hedges in academic discourse as “an item of language which a speaker uses to explicitly qualify his/her lack of commitment to the truth of a proposition he/she utters” (1997: 281). His taxonomy of hedges includes six categories (1997: 284), and they were adapted for use in our study (see Table 1).

As discussed above, Prince et al. (1982) categorize hedges in the light of their functions, and Salager-Meyer (1994) builds the formal-functional classification of hedges. In contrast, Crompton’s (1997) taxonomy is primarily based on the form/structure of hedges (e. g., epistemic copulas/modals, adjectives/adverbs expressing probability) in academic discourse (see Table 1). While there are a variety of linguistic resources serving as hedging devices (e. g., Markkanen and Schröder 1988; Varttala 1999), Crompton’s taxonomy is limited to a choice of “items of language”, which are most likely to be seen as hedges by the academic discourse community. Of course, this narrow selection excludes some non-lexical hedging elements, e. g., tense and questions. It is argued, however, that lexical items are most easily associated with hedging in the eyes of the readers of academic texts (Varttala 1999). Also, Hyland (1994) notes that, when Anglophone writers employ hedges to indicate tentativeness, discretion and doubt, the selection falls largely upon lexical hedging elements. Further, since Crompton’s taxonomy is mainly concerned with the form/structure of hedges, his approach is immune to the influences of the nature of the corpus in Prince et al.’s (1982) classification and avoids overlaps over Salager-Meyer’s (1994) classification. We, therefore, decided to adopt Crompton’s taxonomy as the conceptual framework of this study.

3 Methodology

4 Results and discussion

5 Conclusion

We took a corpus linguistics approach to investigate the use of hedges in the conclusion section of applied linguistics RAs written in English by authors of two distinct groups. We then attempted to triangulate the findings with qualitative interview data. We found that the Anglophone academic English writers, as a whole, used more hedges than their Chinese counterparts. Nevertheless, both groups of writers demonstrated the same tendency in the frequency of use of various categories of hedges except for only one subcategory, viz., H5.1. We also found that the two groups of writers displayed a similar pattern in the choice of various categories of hedges. Both groups saw epistemic modals and non-factive verb phrase structures as very important hedging devices. The two groups differed somewhat in the linguistic expressions of some categories of hedges. In the first category of hedges, Chinese L2 English writers used tend more often than Anglophone counterparts. In the second category, Chinese writers underused should as an epistemic modal. In the fifth category, Chinese writers rarely used the “I/we + non-factive verb” structure.

To sum up, similarities in the two groups of writers’ use of hedging strategies imply that the Chinese academics, to a large extent, have developed L2 English pragmatic competence and made rhetorical and stylistic accommodations to academic English discourse norms so as to communicate interculturally and publish their RAs in English-medium journals. Nonetheless, there are still some differences. Admittedly, we would like to contend that academic English discourse norms or conventions are not static but dynamic and changing (Connor 2011; Gillaerts and Velde 2010), which is why we built our corpus within a time frame (i. e., 2009–2011). Causes of the differences mentioned above might be quite complex, but different social cultures, rhetorical traditions, and levels of pragmatic competence of these academic English writers are probably vital factors that contributed to such findings. It can be argued that non-native English-speaking writers of academic English ought to have written in their own style so that their own authorial voice could have been better projected (Zhang 2013). Nonetheless, for purposes of presenting their ideas clearly to their readers, who are not particularly familiar with the discourse of argument or rhetorical style, they had to adapt to the “common core” (Canagarajah 2015) of Anglophone academic discourse norms and be considerate of their readers’ experiences.

Based on the findings we may conclude that the present study, in a sense, has enriched our understanding of L2 pragmatic acquisition, especially as regards using written academic English for international publishing. For various reasons (e. g., limited research design, small sample size, or limited data), academic writers often have difficulty in providing more precise and definitive information. Thus, hedging the conclusions appropriately in L2 academic writing could mitigate the force of the claims, leave room for negotiation with the audience, reduce the risk of opposition, and minimize face-threatening acts involved in argument making; doing so would enable L2 academic writer(s) to make claims moderately and more politely for convincing the academic English discourse community more forcefully on reasonable grounds. In our view, these are significant manifestations of L2 pragmatic competence.

We anticipate that the findings of this study would help to improve ESL/EFL learners’ and users’ pragmatic awareness and ability so that they will be enabled to make use of hedging devices more felicitously. Furthermore, we think that our findings will not only deepen our understanding of academic discourse, but also shed some light on L2 academic writing, especially in relation to writing academic English textbooks and enhancing academic English instruction.

Frankly, our study is exploratory in nature and a statement on its limitations is in order. Due to limited time and constraint of resources we analysed only 60 RAs that spread across three years in two applied linguistics journals, AL and CJAL. The findings, therefore, might be skewed by the small size of data to some extent. Obviously, further work needs to be done in this area of research, particularly when a larger sample size is possible. We believe that the findings from a larger sample size will be more illuminating of the phenomenon in question.