Intercultural and interdisciplinary variation in the use of epistemic lexical verbs in linguistics and economics research articles

Intercultural and interdisciplinary variation in the use of epistemic lexical verbs in linguistics and economics research articles Olga Dontcheva-Navratilova (Brno) ABSTRACT This paper explores rhetorical variation in academic discourse, focusing on the choice and use of epistemic lexical verbs in linguistics and economics research articles written in English by Anglophone and Czech scholars. Drawing on Hyland’s (1998a) taxonomy of epistemic lexical verbs, the contrastive analysis combines quantitative and qualitative methods to consider how rhetorical variation is affected by both the culture of the discipline and the culture of the writer. The investigation is carried out on aspecialised corpus comprising 48 research articles (12 per discipline and cultural background) published in international and national (Czech) academic journals. Apart from establishing the frequency of occurrence of judgement and evidential epistemic lexical verbs, the analysis considers the immediate co-text of the target items and the distribution of different types of epistemic lexical verbs across the rhetorical sections of research articles. The results of the investigation indicate that while the lower frequency of use of epistemic lexical verbs in research articles by Czech writers is due to intercultural variation, the preferences towards the use of specific types of epistemic lexical verbs, the clusters they form, and their distribution across the rhetorical sections of research articles seem to reflect both cultural and disciplinary considerations. These findings suggest that culture and discipline seem to govern different aspects of rhetorical choices in academic discourse. KEYWORDS epistemic lexical verbs, persuasion, intercultural variation, disciplinary variation, research articles 1. INTRODUCTION The dominant position of English as the lingua franca of globalized academia has profoundly affected the way academic knowledge and discourse are construed and disseminated. While facilitating knowledge exchange and scholarly networking, the predominance of English has forced non-Anglophone researchers to adhere at least to some extent to the dominant Anglophone academic writing conventions when they strive to publish their work in an international context. Thus the English-medium academic discourse of non-Anglophone researchers is marked by ‘glocal’, hybridized rhetorical practices (Peréz-Llantada 2013) merging features of the Anglophone norms and the native language and culture of the writers. Previous research in the field of intercultural rhetoric (e.g. Bennett 2011; Clyne 1987; Dontcheva-Navratilova 2014; 2016; Duzsak 1994; Lorés-Sanz 2011; Mauranen 1993; Yakhontova 2006) has provided convincing evidence for the development of this process suggesting that we are witnessing the emergence of “alternative academic written Englishes” (Mauranen, Pérez-Llantada and Swales 2010: 647), i.e. variants of academic English used as academic lingua franca by non-Anglophone authors from aspecific linguacultural OLGA DONTCHEVA-NAVRATILOVA 155 background. It is therefore essential to explore these emerging academic Englishes in order to raise the awareness of expert and novice authors of the existence of intercultural variation in academic writing conventions and thus provide them with abetter chance of having their work accepted for publication in an international context. Yet variation in academic discourse is not only intercultural but also interdisciplinary, as different disciplines have their own specific ways of shaping arguments and constructing knowledge through discourse (e.g. Hyland 2002, 2004, 2008). Disciplinary variation in academic persuasion is associated with the use of rhetorical strategies and language choices established in the discursive practice of aspecific disciplinary culture (Swales 1990) which are instrumental in enhancing writer credibility through the construal of the writer’s “identity of aperson with authority” (Ivanič 1998: 88) and contribute to content reliability through expressing different degrees of commitment to claims and relating new ideas to shared knowledge (cf.Hyland 2008; Livnat 2012). As an important rhetorical means allowing authors to modify the degree of certainty and commitment they opt for when negotiating their views and claims with the readers, the expression of epistemic modality has been shown to vary significantly across cultural and disciplinary contexts (cf. Dahl 2004; Fløttum et al 2006; Hyland 1998a; Pérez-Llantada 2010; Vold 2006). Despite the growing body of research into intercultural and interdisciplinary variation in the use of epistemic modality markers, very few studies have adopted adoubly contrastive approach to investigate whether culture or discipline is the most important variable in the expression of epistemic modality in academic discourse. This study attempts to fill this gap by exploring intercultural and interdisciplinary variation in the expression of epistemic modality for modelling persuasion in English-medium linguistics and economics research articles by Anglophone (judging by their names and affiliations) and Czech authors. The investigation focuses on rhetorical variation in the choice and use of epistemic lexical verbs in aspecialised corpus of research articles written in English by Anglophone and Czech scholars representing two soft academic disciplines— linguistics and economics. The aim of the investigation is to explore whether the culture of the discipline or the culture of the writer is amore decisive factor affecting rhetorical choices in English-medium academic discourse. 2. EPISTEMIC MODALITY IN ACADEMIC DISCOURSE: EPISTEMIC LEXICAL VERBS Epistemic modality may be described as ‘propositional’ modality, as it is “concerned with the truth-value or factual status of the proposition” (Palmer 2001: 24). Thus “any utterance in which the speaker explicitly qualifies his commitment to the truth of the proposition […] is an epistemically modal, or modalized utterance” (Lyons 1977: 797). Palmer (2001) differentiates two types of propositional modality— epistemic (speculative, deductive and assumptive) and evidential (reported and sensory); however, most linguists (e.g. Bybee and Fleishman 1995; Coates 1983; Huddleston and Pullum 2002) subsume evidential modality into the epistemic category. While Ihave adopted 156 LINGUISTICA PRAGENSIA 2/2018 the latter approach to the classification of modal meanings, in agreement with Hyland (1998a) and Peréz-Llantada (2010), Palmer’s sub-categories of epistemic and evidential modality are used as abasis for the taxonomy of epistemic lexical verbs used in this investigation. In academic discourse, epistemic modality has asemantic function associated with indicating the source of knowledge, modifying the status of knowledge conveyed, and differentiating facts accepted by the disciplinary community from claims, and apragmatic function related to engaging with readers in negotiation of meaning and expressing humility and politeness (cf. Hyland 1998a; Myers 1989; Peréz-Llantada 2010; Salager-Meyer 1994; Vold 2006; Warchał 2015; Yang et al. 2014). Intercultural variation in epistemic modality stems from differences in rhetorical conventions across different academic cultures, while disciplinary variation reflects differences in the subject matter, argumentation strategies and rhetorical organisation of research articles. These differences are reflected in preferences towards the use of specific realizations of epistemic modality as well as in the pragmatic and rhetorical functions conveyed by epistemic markers. Epistemic modality is realised by awide spectrum of linguistic means comprising modal verbs (e.g. may, can, must), epistemic lexical verbs (e.g. believe, infer, seem), epistemic adverbials (e.g. certainly, perhaps. probably), epistemic adjectives (certain, doubtful, likely) and epistemic nouns (e.g. assumption, probability, view). This investigation focuses on epistemic lexical verbs as one of the most common devices for the expression of subjective evaluation in academic discourse used to hedge an author’s commitment or assertiveness. It adopts Hyland’s (1998a) taxonomy of epistemic lexical verbs broadly based on Palmer’s (1986) typology of modality in English, which comprises the following categories of epistemic lexical verbs: Judgement verbs— conveying the appraisal of the speaker of the factive status of events — speculative verbs— express an opinion qualifying the mode of knowing through various degrees of commitment to the truth of the proposition (e.g. believe, suggest, think) — deductive verbs— express adeduction or conclusion based on inferential or theoretical reasoning (e.g. infer, deduce, conclude) Evidential verbs— indicating the source of knowledge — quotative— justify opinions and claims on the basis of ‘hearsayʼ/reported evidence mainly from the research literature (e.g. report, claim, note) — sensory— justify opinions and claims by referring to sensory evidence (e.g. appear, seem, view) The specific way in which academic authors combine epistemic judgement and evidential lexical verbs to modulate their commitment to propositions is expected to reflect disciplinary and cultural considerations. OLGA DONTCHEVA-NAVRATILOVA 157 3. CORPUS AND METHOD This investigation has been carried out on the academic subcorpus of The Corpus of English and Czech Specialised Discourses (CECSD), which comprises 48 single-authored English-medium linguistics and economics research articles published in six linguistics and six economics journals in the period 2006–2016 (four articles per journal). The corpus was subdivided into two subcorpora compiled along the same criteria to ensure their comparability: alinguistics subcorpus including 24 linguistics articles (12 by Anglophone authors and 12 by Czech authors) and an economics subcorpus including 24 economics articles (12 by Anglophone authors and 12 by Czech authors). The research articles by Anglophone scholars were published in high impact international journals, i.e. for the linguistics subcorpus: Applied Linguistics (Oxford journals), Journal of Pragmatics (Elsevier), and Discourse &Communication (SAGE); for the economics subcorpus: Journal of International Economics (Elsevier), Economic Systems (Elsevier), Journal of Accounting and Economics (Elsevier), while the research articles by Czech scholars were published in national journals indexed by SCOPUS and ERIH Plus, i.e. for the linguistics subcorpus: Linguistica Pragensia (Charles University, Prague), Brno Studies in English (Masaryk University, Brno) and Discourse and Interaction (Masaryk University, Brno); for the economics subcorpus: Economics and Management (Technical University of Liberec), Prague Economic Papers (University of Economics, Prague), Central European Review of Economic Issues (Technical University of Ostrava). Obviously, these journals provide different contexts of publication and target different audiences; nevertheless, it is considered that the research articles by Anglophone and Czech authors included in the corpus may be seen as reflecting the conventions of social interaction in the respective disciplinary discourse communities. Table 1 provides an overview of the size and composition of the corpus. Discipline Cultural background No of articles No of words Linguistics subcorpus (LING) Anglophone (ENG) 12 77 000 Czech (CZENG) 12 57 000 Economics subcorpus (ECON) Anglophone (ENG) 12 93 000 Czech (CZENG) 12 62 000 Total 48 289 000 Table 1.Composition of the corpus The research articles in the corpus were coded into rhetorical sections following Swales’s (1990, 2004) IMRD framework for the analysis of rhetorical moves in experimental research articles. There were several problems with the coding of rhetorical sections. While the Introduction section is clearly identifiable in all texts, several articles do not comprise an explicitly labelled Methods section, and the majority of the articles display amerged Results and Discussion section followed by afinal Conclusion or Implications section. For the purposes of this study, the sections labelled ‘Results’ or ‘Results and discussion’ were coded as Results section, and the sections labelled 158 LINGUISTICA PRAGENSIA 2/2018 ‘Conclusion’, ‘Implications’ or any other label broadly corresponding to the moves of the Discussion section (Ruyjing and Allison 2003) were coded as Discussion sections. This approach differs from that adopted by Hyland (1998a), who coded separately the Results section, the Discussion section and the merged Results and Discussion section, which may explain some differences in the reported quantitative data. The contrastive analysis combines quantitative and qualitative methods and draws on Hyland’s (1998a) taxonomy of epistemic lexical verbs to study the frequency of occurrence, patterning and distribution of judgement (speculative and deductive) and evidential (quotative and sensory) epistemic lexical verbs across the rhetorical sections of linguistics and economics research articles. Based on the results of previous results and on the corpus data (some initially targeted verbs like presume, deduce and speculate showed no occurrence in the corpus and therefore were excluded), the following 34 epistemic lexical verbs were subjected to analysis: — speculative judgement verbs— argue, believe, consider, doubt, expect, imply, indicate, propose, suggest, suspect, think — deductive judgement verbs— assume, calculate, conclude, deduce, demonstrate, estimate, infer, presume, reason, suppose — evidential quotative verbs— argue, believe, claim, indicate, maintain, note, propose, report, show, suggest — evidential sensory verbs— appear, observe, notice, seem, view The frequency-based investigation was carried out using the SketchEngine corpus tool to identify the rate of occurrence, patterning and distribution of the target epistemic lexical verbs in the corpus. In agreement with the common procedure in contrastive corpus-based research, the raw frequencies were normalised to frequencies per 10,000 words to neutralise the difference in word-count between the Anglophone and the Czech sub-corpora and to allow for comparison with data reported in previous research. After this, the functions of the target items were checked manually, which served as abasis for acontrastive qualitative analysis of the recurring clusters comprising epistemic lexical verbs focusing on their pragmatic functions and their role in enhancing persuasion in academic discourse. 4. RESULTS AND DISCUSSION In consonance with previous research indicating discipline as amajor variable affecting the expression of epistemic modality (cf. Dahl 2004; Hyland 1998b; Vázquez and Giner 2008), the disciplinary results show substantial divergences (Table 2). The rate of epistemic lexical verbs in my corpus is higher than the rate of 39.9 for biochemical research articles reported in Hyland (1998a) and the rate of 39.03 for legal research reported in Vass (2017), which seems to reflect the argumentative character of linguistics and economics. Although the role of argumentation in the knowledge construction process is prominent in both disciplines, it is economists that tend to adhere to amore fixed text pattern, signal more explicitly what discourse act is be- OLGA DONTCHEVA-NAVRATILOVA 159 ing performed and modulate their level of commitment to the proposition (cf. Dahl 2004; Fløttum et al. 2006). As aresult, research articles in the field of economics display ahigher rate of evidential lexical verbs than those in the field of linguistics both in the Anglophone and Czech subcorpora. In addition, while both Anglophone and Czech linguists seem to prefer evidence to judgement (59.8/60.5% vs 40.2/39.5%), economics texts in both subcorpora display ahigher frequency of epistemic judgement verbs (54.3/51% vs 45.7/49%). This difference is mostly due to the higher frequency of sensory evidence in linguistics research articles (about 25% of all evidential verbs in the linguistics subcorpus and less than 15% in the economics subcorpus). Linguists seem to exploit the potential of the perception verb seem to not only code “direct, observation-based source of information” (Fetzer 2014: 341) but also to “encode probability and express more or less certainty” (Aijmer 2009: 65) when modelling their degree of commitment to their statements and claims. Subcorpus / ELVs Linguistics Economics ENG CZENG ENG CZENG n. rate %n. rate %n. rate %n. rate % Judgement verbs 21.4 40.2 14.2 39.5 36.1 54.3 28.2 51.0 Evidential verbs 31.9 59.8 21.7 60.5 30.5 45.7 27.1 49.0 Total 53.4 100 35.7 100 66.6 100 55.3 100 Table 2.Frequency of epistemic lexical verbs in the sub-corpora (per 10,000 words) As to the culture variable, the results of the contrastive analysis summarised in Table 2 show that Anglophone scholars use epistemic lexical verbs considerably more frequently than Czech writers both in the linguistics (53.4 vs 35.7) and economics fields (66.6 vs 55.3). There is little divergence in the ratio of evidential and judgement verbs in the Anglophone and Czech sections of the linguistics and economics subcorpora; however, the difference in their rate of occurrence is more prominent in the linguistics subcorpus. This supports the findings of previous research (DontchevaNavratilova 2016a) which indicated that Czech linguists generally tend to use fewer hedges and boosters than their Anglophone counterparts. These quantitative differences may be interpreted as indicating intercultural variation reflecting the lower degree of interactiveness typical of Czech academic discourse (Čmejrková and Daneš 1997) and the primarily national context of publication of the research articles, which does not force Czech authors to adhere more consistently to the Anglophone rhetorical conventions marked by ahigher degree of reader-awareness and atendency to lead the reader towards an intended interpretation of the text (cf. Mauranen 1993 on the differences between Anglophone and Finnish economics discourse). However, as Peréz-Llantada’s (2013) analysis of social sciences discourse suggests, when published in the same international impact factor journals, research articles by non-native writers tend to approximate the dominant Anglophone conventions, as in her 160 LINGUISTICA PRAGENSIA 2/2018 corpus the difference in the rate of epistemic lexical verbs in articles by Anglophone authors was only slightly higher (59.8 vs 52.4) than that in English-medium texts by Spanish scholars. Acloser look at the most frequent judgement and evidential verbs used in the corpus (Table 3) reveals many similarities across disciplines and cultures. The existing disciplinary differences in the choice of specific epistemic lexical verbs seem to reflect the more interpretative character of linguistics as opposed to the more quantifiable findings in economics. Thus, while linguists tend to argue more explicitly through the speculative verbs suggest, consider and argue (cf. Fløttum et al. 2006: 238), economists prefer to rely on deductions conveyed by the verbs indicate, estimate, calculate and imply. The speculative verb suggest is highly frequent in all sub-corpora (apart from CZENG ECON), but argue is prominent only in the Anglophone linguistics articles, probably because it conveys astronger feeling of authorial presence and Czech academic discourse tends towards backgrounding authorial presence. The deductive verbs estimate and calculate occur only in the economics subcorpus; the high frequency of calculate in articles by Czech authors may be explained by “language leakage” (Jenkins 2015), i.e. the influence of the L1, as there is averb with asimilar form in Czech (kalkulovat). There is less disciplinary variation in the expression of evidentiality, where show, seem and appear are invariably among the most frequent epistemic verbs. The higher frequency of show in the economics texts stems from its potential to refer to previous research and to tables, graphs and equations summarizing research results. While rarely used as an evidential quotative verb in the CZENG subcorpora, suggest is often used by Anglophone linguists and economists to refer to previous research, thus opening adialogue with the readers in which differing viewpoints are recognised and tolerated. Subcorpus / ELVs Applied Linguistics Economics ENG CZENG ENG CZENG Judgement verbs argue (6.2) suggest (4.2) assume (2.8) demonstrate (2.2) indicate (1.5) suggest (2.9) demonstrate (2.5) consider (2.3) expect (1.4) imply (1.4) suggest (9.8) indicate (7.1) imply (4.0) estimate (3.9) predict (3.22) calculate (8.2) estimate (5.5) indicate (2.2) assume (1.8) consider (1.8) Evidential verbs show (7.8) seem (7.8) argue (6.2) appear (2.8) suggest (2.8) show (8.1) seem (4.9) view (2.5) appear (1.4) claim (1.2) show (13.2) appear (6.2) report (3.3) seem (3.0) suggest (2.4) show (13.2) seem (4.5) conclude (2.1) observe (1.3) appear (1.1) Table 3.Most frequent epistemic lexical verbs in the sub-corpora (per 10,000 words) As the results of acontrastive analysis of the distribution of epistemic lexical verbs across the rhetorical sections of research articles summarized in Figure 1 show, the texts in the whole corpus seem to follow asimilar general tendency: epistemic lexical verbs peak in the Results section, where the authors present and interpret the results OLGA DONTCHEVA-NAVRATILOVA 161 of their research and provide evidence for their claims, the lowest rate of epistemic lexical verbs is found in the Methods section dealing with the research procedure, while their frequency in Introductions and Discussions lies between that of the Results and Methods sections. This tendency was also evidenced by Hyland (1998a) and PerézLlantada (2010) in their biomedical discourse corpora; however, since the Results and Discussion sections in biomedical articles tend to be clearly separated, it is Discussions that scored the highest rate of epistemic lexical verbs. Despite these similarities, the results of the investigation indicate the existence of disciplinary and cultural variation. Thus, economics articles display ahigher rate of epistemic lexical verbs in all rhetorical sections apart from Discussions, where their score is lower than that in linguistics articles. Although Czech authors generally use fewer epistemic markers than their Anglophone counterparts, in Methods sections the frequency of epistemic lexical verbs is higher in the CZENG subcorpus. Figure 1.Distribution of epistemic lexical verbs across research articles sections An analysis of the interplay of judgement and evidence in the rhetorical sections of research articles (Table 4) indicates further aspects of variation in the ways scholars strive to achieve persuasion across disciplinary and cultural contexts. The range of judgement frequencies is more significant in the ENG subcorpus (0.6 in Method to 11.4 in Results for linguistics, 1.5 in Discussion to 23.5 in Results for economics) than in the CZENG subcorpus (0.5 in Method to 10.0 in Results for linguistics, 2.3 in Discussion to 14.2 in Results for economics) and it exceeds the range of evidence frequencies in the ENG (0.7 in Method to 19.3 in Results for linguistics, 1.9 in Discussion to 16.6 in Results for economics) and the CZENG (2.1 to 10.0 in Results for linguistics, 2.3 in Discussion to 13.4 in Results for economics) subcorpora. The narrower ranges across sections in the CZENG texts seem to suggest that the rhetorical choices of Czech authors are to acertain extent restricted by the use of English as an academic lingua franca and may be 162 LINGUISTICA PRAGENSIA 2/2018 affected by the input provided in academic writing courses, which typically do not take into account disciplinary variation. In Introduction sections Anglophone authors tend to keep abalance between judgement and evidential verbs, although in the linguistics subcorpus there is aslight preponderance of evidential and, in the economics subcorpus, judgement verbs. In the CZENG subcorpus evidential verbs clearly outnumber judgement verbs, which is primarily due to the considerably lower occurrence of speculative verbs modulating commitment to views and claims. As to disciplinary variation, linguists rely primarily on speculative and quotative verbs to establish aterritory, evaluate previous research and occupy aniche, while economists construct arguments by combining speculative and deductive judgements with quotative and sensory evidence. The Methods section is marked by considerable disciplinary and cultural variation. In the linguistics subcorpus, the rate of epistemic judgement verbs is similarly low in research articles by Anglophone and Czech linguists; there is, however, aconsiderable difference in the occurrence of evidential verbs, which is much higher in the CZENG LING than in the ENG LING subcorpus. The results for the economics subcorpus indicate that Anglophone authors show apreference for evidential, mostly quotative verbs and Czech writers tend to rely on deductive judgements. Corpus Applied Linguistics Economics ENG CZENG ENG CZENG Introduction Judgement verbs 6.36 2.6 8.3 4.8 Evidential verbs 7.14 5.4 7.7 7.7 Method Judgement verbs 0.6 0.5 2.4 6.6 Evidential verbs 0.7 4.9 5.4 3.7 Results Judgement verbs 11.4 6.8 23.5 14.2 Evidential verbs 19.3 10.0 16.6 13.4 Discussion Judgement verbs 3.0 3.5 1.5 2.6 Evidential verbs 4.5 2.1 1.9 2.3 Table 4.Distribution of epistemic judgement and evidential verbs across research articles sections Similarly to Introductions, the Results section displays disciplinary variation concerning apreference towards evidentiality in linguistics and towards judgement in economics. Within the epistemic judgement category, linguists tend to provide more speculative judgements to modulate their claims, while economists rely on deductions. Despite the existing variation in frequency, evidence is primarily quotative in all subcorpora. The markedly lower rate of occurrence of evidential quotative verbs in the CZENG subcorpus is in consonance with the findings of previous research (Dontcheva-Navratilova 2016b) indicating that Czech authors rarely use the optional Reference to previous research (comparison) step of the Commenting on results move in the Results section. This also seems to account to alarge extent for the existing intercultural difference in Discussion in the CZENG LING subcorpus, where the occurrence of both quotative and sensory evidential verbs is lower than in the