Modal hedging verbs in English as a Lingua Franca ( ELF ) Business Management research articles

When publishing the results of their research, scholars need to convince their readers of the validity of their claims, adjusting their writing to the prevailing discursive and rhetorical conventions. Hedges play a crucial role in persuading readers of such validity. Previous cross-cultural research in English for Academic Purposes (EAP) has revealed significant differences in the use of these interpersonal features in research articles (RAs) in different languages and publication contexts. In this paper, hedging modal verbs will be contrastively analysed in a corpus of RAs written by English as a Lingua Franca (ELF) users and by native English scholars affiliated to Anglophone institutions in the field of Business Management. The frequency of use of hedging modal verbs, their main functional uses, and their particular phraseological realizations will be examined. The results will help us gain an insight into how new knowledge claims are negotiated in international English-medium publications in this discipline as well as into the particular shaping of ELF written academic communication.


Introduction
The great spread of the use of English to disseminate research results, especially as a consequence of policies being enforced promoting English-medium publications (see Bennett 2014a; Kuteeva & Mauranen 2014), has triggered a great deal of research into specific academic genres.This research has frequently focused on the genre par excellence in the academia, the research article, and an intercultural rhetoric perspective (Connor 2004) has been frequently taken.Cross-cultural studies within EAP (English for Academic Purposes) have shed light into the use of particular rhetorical and discursive features in RAs published in different languages and different contexts of publication (e.g.German & English (Kreutz & Harres 1997), Bulgarian (Vassileva 1997(Vassileva , 2001)), French and Norwegian (Fløttum et al. 2006;Vold 2006), Italian (Giannoni 2005;Diani 2008;Molino 2010), Hungarian (Árvay & Tankó 2004), Portuguese (Hirano 2009), Chinese (Loi 2010;Loi and Evans 2010), Lithuanian (Usonienė & Šinkūnienė 2014), French and Spanish (Salager- Meyer et al. 2003) and Spanish (Burgess 2002;Moreno 2004;Sheldon 2009;Mur-Dueñas 2010, 2011;Vázquez-Orta 2010;Williams 2011)).Some analyses have also been undertaken of their use by non-native English scholars when publishing their research in international English-medium publications (e.g.Burgess 2002;Shaw 2003;Martínez 2005;Lorés-Sanz 2011;Murillo 2012;Mur-Dueñas 2012, 2014).However, this research has tended to emphasise the differences between the use made of English by ENL scholars and scholars from different linguacultural backgrounds, and has assumed that the Anglophone conventions prevailing in international publication contexts should be met -or at least should be considered -to reach successful publication outcomes.From an English as a Lingua Franca (ELF) perspective (e.g.Jenkins et al. 2011;Seidlhofer 2011;Dewey 2007;Mauranen 2012) scholars publishing their research outcomes in international English-medium publications are to be considered legitimate users of the language who contribute to shaping its uses and form.This use of ELF in academic writing could also allow for discursive heterogeneity in international publications, not necessarily promoting 'core' or 'centring' publication practices and favouring the participation of scholars in the (semi)periphery (Bennett 2014b).
Given the current use of ELF in academic communication, especially in certain disciplinary domains, it seems necessary to look at how it is being shaped by all its users not just English Native Language (ENL) researchers.Although descriptions of ELF in academic settings have been carried out in oral encounters (e.g.Björkman 2009Björkman , 2013;;Metsä-Ketelä 2010;Mauranen et al. 2010;Mauranen 2014), descriptions of ELF which focus on written academic genres are scarcer (e.g.Carey 2013; Mur-Dueñas 2015; Lorés-Sanz 2016) but particularly necessary (Mauranen 2012).In this context, this paper seeks to analyse the particular use made of an interpersonal feature, hedging modal verbs, in ELF RAs.These RAs are considered instances of ELF in that they are published by scholars from varied linguacultural backgrounds in international journals in which authors representing multiple similects 1 constitute a high percentage of contributors.Results on the frequency of use, functional uses and specific phraseological realizations 1 "Similects are lects of English spoken by those who share a particular first language but do not form a language community based on the use of ELF as a shared language."(Mauranen 2014, 242) of the hedging modal verbs may, might, could, can and would in ELF RAs will be compared to results stemming from the analysis of ENL RAs, published in international journals by scholars based at Anglophone institutions.The analysis will focus on RAs in the field of Business Management, which can be considered a 'highly internationalised' discipline (Petersen & Shaw 2002), in which networks are frequently established among scholars from different linguacultural backgrounds who use English for international communication.

Hedging modal verbs in academic writing
Hedges are included in an RA to express meaning tentatively, leaving space open for readers' possible different views or interpretations.In this sense, they contribute to avoiding threatening peer readers and to expressing modesty towards them (Myers 1989).In addition, it has been claimed that the use of hedges can respond to other pragmatic functions; their inclusion in academic texts may respond to a need to convince readers of their arguments so that they become accepted, taking into account the communities' norms and values (Hyland 1998).Hedging has come to be seen as a key characteristic of academic discourse, due to the indeterminate nature of knowledge produced and distributed in the academia.As Mauranen (1997, 115) states, "academic discourse is a world <...> of uncertainties, indirectness, and non-finality -in brief, a world where it is natural to cultivate hedges".
Hedging is considered a problematic concept and there does not seem to be full agreement on what counts as a hedge.One of the problems that hinders homogeneity and consensus in what is to be considered a hedge is "the multiplicity of forms hedges may take" (Varttala 2001, 24).Hedges can be seen to constitute an open category, to which new realisations can be added.Another problem is that not all instances of a linguistic form can be considered hedges.They are context-dependent: "no linguistic items are inherently hedgy but can acquire this quality depending on the communicative context or the co-text" (Markkanen & Schröder 1997, 4).
In this study hedging refers to lexico-grammatical choices in the text which help writers withhold full commitment from a proposition, and which contribute to modulating the expression of (un)certainty, commitment and (im)precision in view of the readership the text is intended for.This can be the result of not having enough empirical evidence or assurance to claim that a proposition is absolutely true or certain, or the result of a desire not to express that certainty (Hyland 1998), even if the writer counts on support for it.Their use thus allows for alternative voices and viewpoints, opening up negotiation of meaning and space for possible opposing views, which may be expected in a given disciplinary community.
Modal verbs are considered the most prototypical realisation of hedging.Previous analyses of hedging in written academic discourse include modal verbs as a first and main category (e.g.Salager-Meyer 1994;Crompton 1997;Hyland 1998Hyland , 2005;;Varttala 2001).The study here presented, thus, focuses on the study of epistemic modality understood as "the speaker's assumptions or assessment of possibilities and, in most cases, it indicates the speaker's confidence (or lack of confidence) in the truth of the proposition expressed" (Coates 1983, 18).It draws on the expression of 'certainty' or 'possibility' meanings (Palmer 1990) by means of modal verbs.The expression of possibility and of the degree of commitment on the part of the authors is especially relevant in academic writing, as discussed above, in which disciplinary, language and culture-driven socio-pragmatic conventions need to be followed.Hedging, in general, and hedging modal verbs, in particular, have been shown by previous intercultural EAP research to be used differently across languages and contexts of publication (e.g.Bulgarian (Vassileva 1997); German (Kreutz & Harres 1997); French and Norwegian (Vold 2006); Spanish (Vázquez-Orta 2010); Lithuanian (Usonienė & Šinkūnienė 2014)).It is, therefore, interesting to look into the use of this interpersonal feature made by scholars from different linguacultural contexts when publishing their research results in international English-medium publications and explore the extent to which their use of ELF resembles or differs from the use made of the same interpersonally-driven feature in ENL RAs.In particular, may, might, could, can, and would will be analysed, as they contribute to expressing the writer's lack of full confidence in the propositions they accompany.
Whereas the epistemic value of may, might and could is unquestionable, there is more controversy regarding the extent to which can expresses epistemic modality.As stated by Biber et al. (1999, 491), "[c]an in academic prose commonly marks both ability and logical possibility"; it "is especially ambiguous in academic prose, since it can often be interpreted as marking either ability or logical possibility" (1999,492).Although first approaches to modality did not consider can as an epistemic marker, Coates (1995) points at the development of its epistemic meaning in declaratives.In academic discourse it has been shown that can may express epistemic possibility meanings (Biber et al. 1999;Varttala 2001).Accordingly, it is here believed that can may be included in the text to reduce assertiveness and commitment; as such, merger cases, in which the two readings are possible and both meanings -ability and possibility -can be processed (Coates 1983(Coates , 1995)), were considered in this study.Finally, would as a "hypothetical marker" (Coates 1983) or as a marker of tentative prediction has also been included in the analysis of hedging modal verbs in ELF RAs.

Methods
The analysis is based on a corpus of 48 RAs published between 2002 and 2010.This corpus is part of a bigger compilation of texts, SERAC (Spanish-English Research Article Corpus), carried out by the research group InterLAE at the Universidad de Zaragoza (Spain).The ELFBM (English as a Lingua Franca Business Management) sub-corpus comprises 24 RAs published in 3 journals (British Management Journal, Journal of Management Studies and Research Policy) in which a great percentage of publications are authored by non-native English scholars representing different similects.As such, these sites of publication can be considered platforms of ELF communication in as much as "contact between hybrids" (Mauranen 2014, 229) is fostered since authors that represent a different hybrid address other peers representing different hybrids.The selection of RAs from different similect groups was carried out proportionally and at the same time trying to include the maximum number of similect groups, after a careful analysis of the varied linguacultural backgrounds of the scholars authoring all published RAs in the different issues throughout the years 2006-2010.Such analysis rendered the following representation and proportion of similect groups: The hedging modal verbs may, might, could, can and would were searched for in the two sub-corpora using Wordsmith Tools 4.0 (Scott 1996).All concordances obtained were carefully looked into and those tokens which did not contribute to expressing the authors' lack of full commitment were dismissed.As a result, tokens of non-epistemic uses of the modal verbs were left out (Examples 1-5), and also those not encoding the author's viewpoint towards the certainty of the propositions (Examples 6-9), that is, those tokens included in quotations or as part of other integral and non-integral citations and impersonal citations (Lorés 2006).
(1) While this may limit the scope of our study, it provides a clear frame of reference within which we can conduct a solid empirical analysis, and <…> (ELFBM12) (2) As most of the projects we looked at were very effective, the influence of the fuzzy front end on effectiveness could not be explored.(ELFBM19) (3) However, in small and large organizations, there is a tendency to focus on younger, less experienced people; people who are acquainted with the latest new knowledge, but who may not be well connected yet.(ELFBM18) (4) There was the threat that the workers might go on strike, but the laying off workers, particularly in 1908, indicates that the factory workers' resource dependence based influence was no more than moderate.(ELFBM16) (5) At the time of the data collection, the CIWs worked with one of the independent CIOs; after the merger, they would have to work with one and the same merged CIO.(ELFBM8) (6) Verworn et al. (2008: 3) stated that "the more a risk or uncertainty can be reduced during the front end of this process, the lower the deviations from front end specifications during the subsequent project execution phases and hence the greater the success in product development."(ELFBM19) (7) Zaheer, Schomaker and Genc (2003) argue that there are two ways in which the equality principle can be operationalized.(ELFBM6) (8) The focal firm cannot be certain of its partners' cooperative orientations, as such orientations may be deeply rooted (Barney, 1991) and may not be readily disclosed (Parkhe, 1993).(ELFBM9) (9) Prior research also suggests that POS may mediate the relationship between HR practices and individual HR outcomes.(ELFBM14) Examples 6 to 9 may be considered to be "attributed" hedges, and following Crompton (1997), for hedges to be seen as performing an interpersonal function, they should necessarily express the authors' stance.It would be difficult to determine whether by reporting other tentative voices, authors could be ascribing to or just transferring that modalisation.As a result, in this particular study they were not considered in the overall counts or in the analysis.
A data-driven approach was taken in the analysis; as such, no categories or particular uses and functions were imposed on the data, but rather the context of the concordances was studied in detail to try and discern particular preferences or recurrences, determining similarities and differences in terms of functional uses and also in terms of preferred lexico-grammatical patterning in the two corpora.The results show a similar frequency of use of hedging modal verbs in the two subcorpora, which may point to some degree of homogenization of published written outcomes, at least within the same discipline.Even though the texts in the ELFBM subcorpus are authored by scholars from varied linguacultural backgrounds, the frequency of use of these particular interpersonal features in their RAs is rather similar to that in ENL RAs; specific details on their particular functions and preferred lexico-grammatical patternings will be provided in the following sections.

Analysis of hedging modal verbs in ELFBM and ENGBM RAs
Given that the RAs in the two sub-corpora are addressed to an international, heterogeneous readership, the findings support the idea that the size of the audience is a determining factor in the choice of rhetorical and persuasive features in written academic discourse.Previous research has pointed out differences in the frequency of use of hedges in general in academic texts in English and in other languages (e.g.Vassileva 19997;Kreutz & Harres 1997;Vold 2006;Vázquez-Orta 2010;Usonienė & Šinkūnienė 2014).This may lead us to suggest that when scholars write in English for an international readership, their use of hedges, as far as hedging modal verbs are concerned at least, is somehow adjusted to the new context, meeting the prevailing rhetorical conventions and the readers' likely expectations.
Not only is the overall frequency of use of these interpersonal features showing the authors' stance similar in both sub-corpora but also the extent to which each hedging modal verb is used in the two corpora.May is by far the most common hedging modal verb, which is in agreement with previous studies based on RAs (e.g. Hyland 1998;Varttala 2001;Vold 2006), followed by can and would in the two sub-corpora.Could and might, on the other hand, are the least commonly used; nevertheless, their frequency of use is reversed in the ELFBM and ENGBM RAs, could being used more frequently than might in the ELFBM texts, and might being used more frequently than could in the ENGBM ones.
When the particular use of each hedging modal verb is looked at in detail, similarities and differences arise between both sub-corpora in the type of constructions in which they are used and in the combinations or clusters they form.This contrastive analysis unveils preferred lexico-grammatical realizations of the expression of tentativeness entailing the use of particular hedging modal verbs in one and the other sub-corpus.It will be argued that some such realizations which are only used, or used to a greater extent, in the ELFBM sub-corpus may be considered ELF innovations (Dewey 2007), creative expressions (Seidlhofer 2011) or emerging patterns (Jenkins et al. 2011).

May
In the ELFBM corpus may is commonly used in the statement of findings as part of a that-clause complementing a main clause which includes a personal reference or a reference to the particular study undertaken, as in the following examples: (10) We suspect that the size effect may partly reflect the tendency that larger subsidiaries can finance their R&D activities more easily.(ELFBM22) (11) We maintain that these internal factors dealing with personality traits may be considered as the basic psychological underpinnings of human capital.(ELFBM21) (12) This study indicates that planning may fulfill different purposes throughout the new product development process.(ELFBM19) May is also found in such constructions in the ENGBM sub-corpus, but to a lower extent.
Whereas may seems to be more commonly used to provide tentative findings in the ELFBM RAs, it seems to more commonly soften implications stemming from findings in the ENGBM sub-corpus, as the subjects of clauses containing may more frequently refer to domain-specific entities, entailing the expression of topic-oriented evaluation rather than research-oriented evaluation (Thetela 1997), as in Examples 13 and 14.The author seems to assess the possibility that these implications arise (it is possible that increasing the frequency of favor exchange provides… or it is possible that accessing foreign knowledge provides…).
(13) Thus, increasing the frequency of favor exchange may provide a means for employees to overcome the potential trade-offs between enhancing their reputations (that is, social status) and increasing their ability to perform assigned tasks (that is, productivity) that result from adopting more or less balanced patterns of favor exchange.(ENGBM4) ( 14) Accessing foreign knowledge may provide these firms with technological knowhow that could be used in both home and foreign markets.(ENGBM20) This hedging modal verb is also used to suggest possible venues for future research.These uses are found to a similar extent in both sub-corpora (7 tokens in the ELFBM corpus vs 11 in the ENGBM corpus).
(15) Hence, future research may want to build on a similar approach and further develop the measures employed.(ELFBM12) (16) On the other end of the timing spectrum, future studies may be conducted to determine how these acquisitions perform over time.(ENGBM9) May is also commonly used following attended and unattended this (Swales 2005;Wulff et al. 2012) or these -that is, as a free-standing pronoun (Examples 17 and 18) or as a determiner attending a head noun phrase (Examples 19 and 20).May preceded by unattended this or these is more commonly used in the ELFBM corpus (27 tokens) than in the ENGBM corpus (5 tokens), whereas the frequency use of may preceded by attended this or these is similar in both sub-corpora (15 vs. 17 tokens) 2 .The range of nouns attended by this is rather wide in both sub-corpora: choice, process, increase, network, result(s), finding, observation, association, selectivity, variety of viewpoints, practices, principles, attitudes in the ELFBM texts and paradox, fact, information, route, strategy, finding, conclusion, claims, differences, effects, organizations, enterprises, benefits, arguments, claims in the ENGBM texts.
(17) To summarize, communication is generally seen as an important factor in merger processes, but relatively little is known about the way in which communication affects post-merger identification.This may have to do with the nature of the available research on organizational identification.(ELFBM8) (18) Neither of the two measures used to assess market opportunity -number of competitors and degree of market saturation -was found to be significant in either 2 It may be interesting to explore whether there are differences between the two subcorpora in the overall extent of use of attended and unattended this, irrespective of whether it is followed by a hedging modal verb, which may explain this difference in its specific modalised use.
of the regressions.This may suggest that although untapped market opportunity may account for some of the success of early entrants <…>.(ENGBM19) (19) Furthermore, the correlation matrix reveals that employees having managerial positions had a more positive view of the employee -organization relationship than did the tellers and the advisors.This observation may also lend some support to a less 'holistic' perspective of HRM <…>.(ELFBM14) (20) By imitating the observable aspects of successful competitors (Mauri and Michaels, 1998), organizations can reduce the uncertainty associated with developing specialized internal competencies (Alchian, 1950).These arguments may seem incompatible with RBV, which suggests that <…>.(ENGBM21) There are some constructions and clusters which are used to a similar extent in both sub-corpora.However, some phraseological realisations entailing the hedging modal verb may seem to be distinctive of the EFLBM corpus, especially, as regards its combination with discourse verbs in personal and impersonal it-clauses.Although first person subject pronouns followed by may are scarce in both sub-corpora (9 tokens in ELFBM and 2 tokens in ENGBM), it is worth pointing out that tokens of We may followed by a main discourse verb (e.g.we may also postulate, we may argue, we may also infer, we may say that) are only found in the ELFBM sub-corpus.Similarly, may used as part of the anticipatory it-pattern containing a discourse verb seems to be characteristic of the ELFBM corpus: (25) However, it may be argued that by increasing his/her stake in the firm, the manager may overwhelm other minority shareholders <…>.(ELFBM2) (26) As a result, it may be suggested that more controlling modes of HRM will be more effective when an organization is characterized by low quality EOR.(ELFBM14) These can be considered framing devices with specific semantic content (epistemic modality in this case) emphasizing the interpersonal dimension of RAs, which Bennett (2014a) reports to be characteristic of Portuguese academic written discourse.These may be also used in other (Romance) languages and authors belonging to different similects may make greater use of them when drafting their RAs in ELF.

Can
Unlike what happened with may, can is commonly used in both sub-corpora, but especially in the ENGBM one (10 vs. 19 tokens), in the statement of findings as part of a that-clause complementing a main clause which includes a personal reference or a reference to the particular study undertaken: (27) We argue that acquisition outcomes can be best explained by a combination of these two perspectives.(ENGBM9) (28) Contrary to conventional wisdom, however, we find that the positive and negative feedback effects can have a mutually strengthening relationship.(ELFBM20) Can is frequently used in impersonal passive constructions with discourse verbs (e.g.characterize, clarify, conceive, consider, define, describe, express, regard, perceive, see, or view) to tentatively express the authors' understanding of particular concepts, theories or models.Whereas such expressions are found in the two sub-corpora, they are far more commonly included in the ELFBM sub-corpus than in the ENGBM one (52 vs 10 tokens).
(29) Board leadership structure and managerial ownership can be considered as substitutive corporate governance mechanisms.(ELFBM2) (30) After all, a merger can be seen as a threat to one's own group identity, involving uncertainties about the extent to which this current group identity will survive.(ELFBM8) Just as in the case of may, can is also used after the first person plural pronoun we in the ELFBM sub-corpus to a greater extent than in the ENGBM one (11 tokens vs. 3 tokens), especially with discourse verbs such as discuss, think, hypothesize, or distinguish.
Examples of the anticipatory it-pattern in which a discourse verb and the hedging modal verb can are combined were found in the ELFBM sub-corpus; just one such construction was found in the ENGBM sub-corpus.
(31) Our regressions analysis find that those firms which considered patents as an important means to protect IPRs of their innovation outcomes tend to be more active in the university -industry cooperation.Then, it can be said that technology holding companies promoted by the Korean government and some universities might be a good idea as one of the main functions of such companies is to facilitate licensing of college-held patents.(ELFBM17) The expression of a cause-effect relationship by means of the verbal expression lead to was used in combination with may to a similar extent in both sub-corpora, as shown above.It is also used in combination with the second most common hedging modal verb, can, but this combination is more frequent in the ENGBM than in the ELFBM sub-corpus (9 vs 2 tokens).
(32) Close supervision of SBP employees can lead to alienation (the implicit message being that employees cannot be trusted despite their skills).(ENGBM14) Another common combination in which the hedging modal verb can is found in the ELFBM corpus is after attended (9 tokens) and unattended this or these (6 tokens).No such combinations can be found in the ENGBM corpus, although examples were found in which This was followed by may (see section 4.1).This indicates that this pattern is not exclusive of the ELFBM RAs analysed but it seems to be preferred in these texts than in the ENGBM RAs.
(33) It starts from R&D effort, which can lead to some R&D outcomes captured in either patents and other IPR forms or tacit knowledge.These new knowledge [sic.]can then be commercially and practically utilized to increase sales or productivity.(ELFBM17) (34) Third, although information asymmetry cannot be entirely eliminated, more concern should be directed at raising information availability in the Egyptian stock market.This can be attained by tightening disclosure and transparency rules and requirements, as well as establishing corporate governance rating agencies.(ELFBM2)

Would
Whereas may was used to a similar extent in the statement of possible avenues for further research in the ELFBM and ENGBM texts, would is only found with this purpose in the ENGBM RAs.It is used in combination with verbs such as enhance, extend or advance.
(35) Future studies that analyze more extensively who a firm's main competitors are would further enhance our understanding of some of the issues considered in this paper.(ENGBM20) (36) Such comparison and/or longitudinal research would advance the present research.
Just as in the case of may, would is also used in combination with an adverb and an adjective, but such uses are found to a greater extent in the ENGBM corpus than in the ELFBM one (6 vs. 1 tokens).
(37) Generating insights into the performance implications of these decisions would be particularly valuable.(ENGBM15) In both sub-corpora would is frequently used in combination with the verb expect.
Nevertheless, unlike the previous hedging modal verbs, it is more commonly used in active constructions (we would expect) in the ENGBM sub-corpus than in the ELFBM sub-corpus (9 vs. 3 tokens).Its inclusion in passive constructions (would be expected), however, is only found in the latter (7 tokens).A few tokens are also found of the impersonal construction one would expect (4 in the ELFBM sub-corpus and 2 in the ENGBM one).
Clusters are found containing would and other lexical hedges also expressing tentativeness, in particular, main verbs, semi-auxiliary verbs and adverbs, to a rather similar extent in both sub-corpora and to a higher extent than other hedging modal verbs.That is the case of would expect as referred to above, and also of other main verbs such as suggest, tend to, indicate or imply, the semi-auxiliary verbs seem or appear, and adverbs such as probably and likely, as illustrated in the examples below.
(38) To our knowledge, this is the first study of complexity absorption in hospitals to include dynamism as a contingency variable, and yet the assessment of internal complexity as a response to the environment would seem to suggest inclusion of firm-level perceptions.(ENGBM10) (39) Future research that seeks to clarify the nature of the effect of controllability on the discrimination of threat and opportunity would appear to be a fruitful avenue for the theoretical refinement of threat bias.(ENGBM17) (40) Being able to do so would likely be considered as requiring openness to change and to innovation <…>.(ELFBM7) This hedging modal verb is also used in combination with lead to to express tentative cause-effect relationships in the two subcorpora (5 tokens in the ENGBM sub-corpus and 2 tokens in the ELFBM one), although to a lower extent than the previous hedging modal verbs.Also, would is used in combination with unattended this, especially in the ELFBM sub-corpus (8 vs. 5 tokens) (41) Accordingly, this study indicates that increasing managers' top-down knowledge inflows within an organization, without enabling bottom-up or horizontal inflows, is likely to increase managers' exploitation activities without increasing their exploration activities.This would lead to exploitation driving out exploration.

(ELFBM15)
The anticipatory-it subpattern containing an attitudinal adjective is frequently modalised by would.Whereas this pattern is also included in the ENGBM sub-corpus, it is not modalised to the same extent as in the ELFBM texts and the choice of certain evaluative adjectives as part of them is rather characteristic (see Mur-Dueñas 2015).
(42) It would be interesting to complement the results of our work by studying how the impact of psychological variables on performance is moderated by the business model of an organization.(ELFBM5) (43) It would be intriguing to observe whether the evolutionary process suggested by multilevel selection theory (McAndrew, 2002)  Also, as in the case of other hedging modal verbs, could is used in anticipatory-it patterns with discourse verbs, especially in the ELFBM sub-corpus (e.g. it could be said that, it could be argued that, it could be hypothesized that) (1 vs. 6 tokens).
Finally, it is interesting to note that no tokens of could fulfilling a hedging function were found in either of the two sub-corpora with personal plural subject pronouns.As indicated in the methods section, those examples not expressing tentativeness, but rather ability were dismissed (e.g.we could not find, we could not address, we could choose, etc.).

Might
As shown in Just as in the case of the previous hedging modal verbs, might is found to be used in combination with attended and unattended this or these in the ELFBM sub-corpus (4 tokens).No such tokens have been found in the ENGBM sub-corpus.
(48) This might be explained by the specific merger context of this study: <…>.(ELFBM8)

Summary of results
Overall, similarities in the overall frequency of use of hedging modal verbs were found.However, when looking closely into the specific functional uses and preferred lexicogrammatical patterns, some differences arise between the ELF and the ENL RAs.These are summarized in Table 3. Especially relevant are those features that are only found or that are found to a considerably higher extent in the ELFBM texts, and which have been highlighted in bold in Table 3, since, as indicated above, they can be seen as innovations (Dewey 2007), creative expressions (Seidlhofer 2011) or emerging patterns (Jenkins et al. 2011).For instance, all hedging modal verbs have been found to be used in combination with attended and unattended this or these in the EFLBM sub-corpus, whereas only a few of them are used in this combination in the ENGBM one, and to a lower extent.As a result, this may be considered characteristic of ELFBM RAs, at least in this discipline.
Whereas in the ENGBM sub-corpus may, would, could and might have been found to be used when suggesting possible avenues for further research, in the ELFBM subcorpus only may and could, which seem to be distinctively preferred over the rest, have been found.This finding regarding the preference for a hedging modal verb over the others may point at some sort of possible grammatical restriction or concentration in ELF regarding the expression of particular meanings or communicative intentions, a phenomenon which should be further studied.In addition, uses of first person plural we in combination with hedging modal verbs followed by discourse verbs, as well as parallel impersonal it-clauses and passive sentences containing a hedging modal verb seem also to be characteristic of ELF RAs.A final distinguishing pattern that arises from the comparative study refers to their use, especially that of would, in the anticipatory itpattern with an attitudinal adjective.

Concluding remarks
The comparative analysis of the use of hedging modal verbs in RAs published in ELF sites of publications and in RAs published in Anglophone journals has revealed similarities in their overall frequency of use and in the preferences of some of them over others, which points at similarities in the expression of the degree of commitment towards their statements by Anglophone and ELF authors.This is interesting given the wealth of previous cross-cultural research pointing at significant differences in the inclusion of hedges in English RAs and in RAs in other languages (see section 1. Introduction).On the basis of these similarities, therefore, it could be suggested that scholars publishing international RAs in English-medium publications do not transfer the diverging rhetorical conventions in their L1s and national contexts of publication regarding this particular interpersonal feature, but rather adjust them to the new context of publication, in which the conventions of Anglocentric journals seem to prevail.Nevertheless, at the functional and lexico-grammatical level, some differences have been traced in the particular combinations in which hedging modal verbs have been found in one and the other corpus.
Overall, then, the findings in this study point at some degree of discursive homogeneity (Mauranen et al. 2010) in international English-medium publications, which may be expected within a single discipline.Nevertheless, some particular uses of the hedging modal verbs under study seem to be prevalent in ELF written scholarly communication, at least in the field of Business Management, which may be pointing at some degree of hybridity, evolution or dynamism of the English language as used internationally for academic purposes.This conclusion needs to be taken cautiously, since the study is based on a small, though representative and balanced, corpus.The study of these and other interpersonal features (especially those which have been shown by previous cross-cultural research to be language or culture-bound) needs to be replicated in bigger corpora.Also, further research should determine whether preferred and particular linguistic and discursive choices are restricted to disciplines or whether they cut across different disciplines, making them characteristic of written academic ELF.
The aim of the paper was to describe discursive choices regarding the use of hedging modal verbs in ELF written publications, given the scarcity of studies with such a focus and also given their necessity (Mauranen 2012).Since the English language in the academia is currently in the hands of its global users, it is they who will shape it to suit their communicative needs and in accordance to their discursive preferences, which may partly be a reflection of their L1 particular rhetoric and common stylistic choices.The extent to which ELF becomes a reality, and the extent to which discursive hybridity and different ways of expression are entrenched in global scholarly written communication, will very much depend on gatekeepers' (editors and reviewers') acceptance of varied discursive and rhetorical choices.This progressive acceptance is, indeed, desirable, especially in particular disciplines (e.g.social sciences and humanities), as homogenized monolingual academic practices may carry with them undesired effects on research, especially on the (semi)periphery (Bennett 2014a).

Table 1 .
The Netherlands (3 RAs), Turkey (1 RA).The ENGBM (English Business Management) sub-corpus comprises 24 RAs published in 3 different journals (Academy of Management Journal, Journal of Management and Strategic Management Journal) in which a very high percentage of publications are authored by ENL scholars affiliated to Anglophone institutions.Description of the corpus

Table 2 .
Table2below shows the frequency of use of each hedging modal verb in the two subcorpora: i) the ELFBM, consisting of RAs written by authors from different linguacultural backgrounds in journals in which ELF users are predominant, and ii) the ENGBM one, consisting of RAs written by ENL authors in journals in which a high percentage of articles are published by Anglo-American scholars.Frequency of use of hedging modal verbs in the ELF and ENG sub-corpora That is the case of may be + adverb + adjective (Examples 21 and 22) -9 tokens in the ELFBM sub-corpus and 6 in the ENGBM sub-corpus -and may lead to (Examples 23 and 24) -7 tokens are found in the ELFBM sub-corpus and 4 in the ENGBM sub-corpus.
(ENGBM8)ombined approach may be particularly fruitful when studying networks that include direct competitors, and as such may have more or less cooperative partners.(ELFBM9)(22)Insidershavegreateraccess to subjective information regarding top-management performance and may be especially useful when information processing at the board level is critical.(ENGBM18)(23)Cautionisneeded in this interpretation because current patents may lead to an increase in sales or productivity in the next period.(ELFBM7)(24)Withinmanagement groups, the difficulties associated with moderate levels of heterogeneity may lead to negative performance outcomes for an organization.(ENGBM8) (ELFBM21)y and would, which are not found in the ELFBM sub-corpus to tentatively express avenues for future research, could is very frequently used for that purpose in combination with future studies, future research and verbs such as examine, explore, verify, extend, etc. Could with such purposes is also found in the ENGBM sub-corpus, but to a lower extent (15 vs. 5 tokens).This study indicates that planning may fulfill different purposes throughout the new product development process.Future research could explore this proposition in more detail.<…>Thepresentstudy only considered two factors which represent the project execution phase.Future research could verify whether there are further factors which should be taken into consideration.(ELFBM19)As in the case of other hedging modal verbs, could is commonly used preceded by unattended, and also attended this or these in the EFLBM sub-corpus (8 tokens).Only one example of could used in this combination has been found in the ENGBM subcorpus.Both the subjective cognitive style and predisposition to changes are at the basis of human capital, though they have represented for a long time an underresearched area in the relevant literature.These internal factors could make the difference in the adoption of innovations in firms, while <…>.(ELFBM21)

Table 2 ,
might is more frequently used in the ENGBM sub-corpus than in the ELFBM sub-corpus.It is frequently employed to express areas of future research (11 tokens), as in Example 47.Only two such uses of might have been found in the ELFBM sub-corpus.

Passive sentences with discourse verbs can can Anticipatory it-pattern with an attitudinal adjective wouldTable 3 .
Summary of findings: functional uses and lexico-grammatical patterning including hedging modal verbs