Establishing Rapport with the Reader: Engagement Markers in the Discussion Section of a Research Article

Texto integral

Introduction

Publishing in high-ranked journals has become an essential requirement. This fact leads to a ubiquitous demand for the academia to write well-organised research articles (RAs) as it becomes pivotal in ensuring their acceptance for publication [1; 2]. Traditional organisation of RAs according to the IMRAD (Introduction, Methods, Results and Discussion) structure helps the scholars worldwide to present the results of their research in a consistent and persuasive manner. The IMRAD model proved to be applicable across disciplines [3; 4]. Following this structure makes it easier for the reader to understand the logic of the author’s presentation of the research results. What is more, the IMRAD structure allows the researchers to organise the move and step concepts [5; 6], which have had wide-reaching implications¹ [7–9].

As it was stated by numerous scholars [10–12], a Discussion section appears to be the most important and the most demanding part of any RA. Considering its role in constructing and strengthening the principles of argument under consideration, and, therefore, presenting the importance of research results, the Discussion section helps the researchers to state their contributions to the field studied [13–15]. The Discussion section focuses on findings supported by arguments in previous studies [2; 11; 12], which requires the author of a new manuscript to have the skill to convince the reader of the significance of their own arguments, based on the results obtained during the research. Metadiscourse markers, when used skilfully, can significantly enhance the persuasiveness of argument presentation. The key success factors in this process are awareness of the conventions of the subject area and the authorʼs ability to use metadiscourse markers naturally. To ensure an effective Discussion section, the authors should realise its communicative function through its rhetorical structure based on using proper metadiscourse markers.

In other words, not only the scientific knowledge but also skilful use of rhetorical and language choices make academic texts persuasive by appealing to their ability to have a resonance for the shared beliefs, expectations and generally accepted rules of a specific discourse community [16]. The researchers are investigating the rhetorical patterns of the RA sections and trying to identify a list of words or expressions and to characterise the different rhetorical moves being part of the core organisation of the different sections of the RA [17]. The use of metadiscourse markers enhances the reader’s involvement in the author’s rhetoric making texts more persuasive [18–20]. Writing the Discussion section in RAs requires presenting, assessing and interpreting the results obtained and also justifying the worth of the scholarly contribution of the paper to scientific knowledge of the field [16].

Attention to both rhetorical actions and the means of accomplishing them structurally and linguistically should be paid. An appropriate relationship with the reader appears to be crucial for any research writing. The ability to create a text that establishes both solidarity and disciplinary affiliation, maintains the writer’s reputation in the community and helps prevent objections to his arguments [21]. The RA Discussion section appears to be quite difficult to compile as it is highly argumentative and interactive in nature [16; 22; 23]. To acknowledge the presence of the reader in the text, the writers can directly address the readers, focus their attention on some evidence or even treat them as discourse participants by using engagement markers [18].

The way metadiscourse features are involved in the text is constrained by discipline-specific conventions² [19; 24]. Dontcheva-Navratilova has found that RAs in the field of linguistic studies use significantly more engagement markers, especially proximity markers, while RAs in economics are likely to use more positioning features [20]. This can be explained by the more interpretative character of linguistics which always engages the readers into the dialogue and shares the author’s beliefs, values and perceptions. By contrast, economics employs methods of statistics and mathematics associated with the vast use of directives for positioning readers by instructing and focusing on the aspects relevant to the progress of the argument [20].

Although a great deal of research proves that culture can influence interpersonal phenomena³ [24], little attention has been paid to the use of the language of engagement in RAs written by authors employing English as a lingua franca (ELF) for their research. This paper illustrates how engagement markers contribute to the persuasiveness of the academic discourse in the RA Discussion sections in the field of English teaching and medicine written by Anglophone authors and authors that use ELF for communicating research.

Research questions:

1. Is there disciplinary variation between language teaching and medicine RAs in the use of engagement markers to highlight like-mindedness and ensure interaction with the reader?
2. What are the reasons for the existing differences in the usage of engagement markers in language teaching and medicine RAs?
3. To what extent is global academic communication in English able to reflect the conventions of the global academic English language? In other words, how significant is the difference in the use of metadiscourse markers for reader engagement in texts produced by Anglophone native English speakers and non-native English speakers who present the results of scientific research in academic English as a lingua franca?

 Literature Review

English as an Academic Lingua Franca. English, being an international language, has recently acquired not only the status of language for science and technology, but also that of the world language of academia [25–27]. As a result, scholars worldwide disseminate and make visible the results of their studies publishing research articles in English. Thus, English is considered a global lingua-franca of academia and is spoken mostly by non-native speakers. This fact has generated a lot of controversy around the quality of the English language and has led to a myriad of native speakers having been recruited to help non-native scientists to overcome the problems with the language and proofread the papers prior to publication. However, this practice evoked a number of questions, and the main assumption concerned the fact that if the majority of scholars were not native speakers of the English language, it would be more consistent to consider clarity and efficiency in communication from the point of view of non-natives rather than from Anglophones⁴ [25; 27; 28].

English as a lingua franca designates this new language variety that does not absolutely coincide with English as a national language and cultural vehicle [28]. This English for reference and communication purposes has become a variety of its kind, and those who speak it are treated as language users capable of creating their own standards of acceptability, comprehensibility, and correction⁵ [29]. Leyi [30] introduced the notion of ‘World Englishes’ (WE) – local varieties of English that overlapped functional dimensions of English used in specific contexts, for example, academia or business⁶  [31; 32]. WE and ELF are similar in their positive and accepting attitude towards using English as a ‘working language’ by non-native speakers who use it to achieve their goals in research.

This use of English in a variety of specialised communicative contexts has made this language more dynamic and flexible and much less standardised, with a greater focus on effective communication than grammatical accuracy [27]. Due to the necessity to publish in international journals and to participate in research collaborations, non-anglophone academia have to be able to read and write research articles in English. Therefore, non-native speakers as well as the native ones need to be aware of the rhetorical organisation and engagement markers traditionally used in their fields of scientific interest [25]. Faber [28] believes that discourse structure, syntax, vocabulary, pragmatic content, or a combination of all of these elements may cause difficulties for ELF users and lead to the tendency to apply the rules of their mother tongue to the English language.

Language of Engagement in Academic Research. Academic writing is considered as an interaction or a dialogue between the reader and the researcher. Scholars are in search of ways of a credible representation of the results of their work. To make their argument more convincing and to reveal their personality, the authors claim solidarity with readers, evaluate their investigations and admit alternative points of view [33; 34]. A lot of research has brought to light the ways in which various disciplines study distinct patterns of academic rhetoric [18; 20; 35]. Some research on metadiscourse, stance and engagement and interaction and persuasion patterns [18; 19] has revealed that writers in the humanities and social sciences are more involved in their dialogue with the readers than the authors of papers in the science and engineering fields [20; 33]. It is necessary to present results and interpretations in persuasive ways; therefore, researchers should draw on these to express their positions, represent themselves, and engage their audiences. Engagement is treated as a dimension where the writers recognize their audience, connect to them, pull them along with their attention, admit the uncertainties, involve them into their discourse and guide them to their interpretations using reader pronouns, questions, directives, shared knowledge and personal asides [18]. These features make it possible for the writer to achieve various goals: reader mentions and personal asides deliver solidarity and create the atmosphere of community membership, while questions, directives and appeals to shared knowledge let the researcher conduct a dialogue with the readers taking into account their positions, beliefs, admitting their views and let the researcher anticipate criticism thus construing a sense of sharedness and agreement [18; 20; 21]. Moreover, the researcher seeks to guide the reader through their text and micro-manages interpretations, anticipating possible objections and providing coverage of key features.

Hyland & Jiang’s model of engagement [21] describes five types of engagement markers, each type being organised into two large categories reflecting the interaction coordination between the reader and the writer. The first type of such rhetorical interaction is proximity, which is aimed at highlighting like-mindedness of the reader and the writer within the same discipline thus making them equal participants. The second type describes the way the writer interacts with the reader through guiding them to the desired interpretation and anticipating any possible objections and differences in understanding views and claims. This category is referred to as positioning.

The category of proximity is implemented through the following engagement features: the inclusion of the reader in the discourse is realised with the help of first-person inclusive pronouns and possessives (we, us our, one) and second-person pronouns and possessives (you, your), thus uniting the producer and the recipient of the discourse through sharing values, views and positions. Moreover, the development of a relationship between the writer and the reader is supported by the contribution of subjective comments of the generous author thus representing his/her personal asides.

Positioning is realised by appeals to shared knowledge while seeking agreement with the reader through awareness of prior topical knowledge, the context and the similarity of views and practices. Hyland & Jiang [21] subdivided appeals to shared knowledge into logical reasoning (e.g., of course, obviously), routine conditions (e.g., normally, routinely), and familiarity with tradition (traditionally, commonly). In addition, the writer and the reader are involved in a dialogue via questions aimed at capturing the reader’s interest and gradually attracting his/her attention to the core aspects in the writer's argument. What is more, the reader can be addressed directly by imperatives (suppose, note), obligation modals (e.g., must, should) and predicative adjectives (e.g., it is important to). These directives refer to a different piece of text guiding the readers towards the expected interpretation of the argument.

Materials and Methods

Corpus. In order to study the ways in which reader engagement is implemented in Discussion sections of RAs, we built a corpus. The corpus comprises Discussion sections of 68 English-medium articles, in which Discussion is presented as a separate section, which is not merged with Results or any other parts of the RA. The size of the resulting corpus is 87,404 words. This corpus was subdivided into four contrastive sub-corpora according to two criteria: disciplinary fields and linguacultural background of the authors. To ensure comparability, balance and representativeness of the sub-corpora⁷, they were built according to the same principles: sources of RAs, number of RAs in each corpus, RA structure, authors’ background and origin.

All RAs were published between 2019 and 2022 and come from international peer-reviewed English-medium journals indexed in Scopus 2^nd quartile and having similar indicators, such as average citations per document and source normalised impact per paper. The journals are published by Wiley-Blackwell, SAGE, Cambridge University Press, Oxford University Press, American Physiological Society, and Springer Nature. This similarity in the status of publications can ensure consistency of linguistic data, which will be obtained.

According to the discipline, the RAs are subdivided into two fields, foreign language teaching (FLT) and medicine (MED), each represented by 34 RAs from four different titles. FLT RAs were selected from recent issues of Language Learning, Language Teaching Research, Studies in Second Language Acquisition, TESOL Quarterly; MED RAs come from International Immunology, Journal of Applied Physiology, Journal of Diabetes, European Journal of Clinical Microbiology and Infectious Diseases. All RAs are empirical papers which were published between 2019 and 2022 by teams of authors as well as single authors (Table 1).

 

Table  1.  Sources

Source title	CiteScore	% Cited	SNIP	SJR	Impact Factor
FLT
Language Learning	6.4	87	2.941	2.882	4.667
Language Teaching Research	6.0	94	1.869	1.663	3.899
Studies in Second Language Acquisition	5.7	83	2.186	2.207	4.313
TESOL Quarterly	3.9	72	2.005	1.737	3.692
MED
International Immunology	6.8	90	1.256	1.860	4.823
Journal of Applied Physiology	5.6	82	1.280	1.253	3.531
Journal of Diabetes	5.5	79	0.978	0.949	4.006
European Journal of Clinical Microbiology and Infectious Diseases	5.4	80	1.174	1.154	3.267

_{Source: Hereinafter in this article all tables were made by the authors}

 

The second criterion for creating sub-corpora is linguistic peculiarities of the RAs. The first group of papers was written by Anglophone authors (NS), while the second one demonstrates the use of ELF for communicating research (MULT). Referring authors to the Anglophone or Non-Anglophone group is a challenge in the highly globalised academic world. Understanding these limitations, in our paper we used a combination of criteria. We referred authors to the Anglophone group if the authors’ affiliation was with an institution located in an English-speaking country, if the authors hold degrees from English-speaking countries, especially the first degree (based on the information available on their web-sites or CV), if the authors have an English-looking name, and if they publish only in English. The authors whose education, background, name and publication record are different we referred to the second Non-Anglophone group. This classification will enable us to compare and contrast the linguistic means Anglophone native speakers and representatives of other linguocultural communities choose to engage their readers in the Discussion sections of their papers. The NS sub-corpus represents the use of English by British, American, Australian and Indian researchers. Non-Anglophone authors include speakers of a variety of languages, including French, Spanish, Italian, Chinese, Japanese, Korean, German, Danish, Norwegian, Finnish, Greek, Russian, Polish, Turkish, Hebrew, Persian, who come from 19 different countries (Belgium, Chile, China, Denmark, Finland, Germany, Greece, Iran, Israel, Italy, Japan, Korea, Norway, Poland, Spain, Switzerland, Turkey, Saudi Arabia, Russia). This range can ensure representativeness of the sample. To mitigate the possible effect of the journal requirements and editing process on the RA text, we selected the same number of NS and MULT RAs from every journal. This way, if a NS sub-corpus includes five NS RAs, a corresponding MULT sub-corpus will also include five RA from the same journal.

Instruments and Procedure. The selected groups of RAs were processed using corpus building and text analysis tool Corpus sketchengine. As a result, we obtained four sub-corpora: FLT MULT, FLT NS, MED MULT, MED NS, the details and composition of which are summarised in Table 2 (Appendix 1).

 

Table  2.  Composition of the corpus

Sub-corpora	NS		MULT		Total
	RAs	Words	RAs	Words	RAs	Words
RAs FLT	17	22,103	17	21,925	34	44,028
RAs MED	17	23,144	17	20,232	34	43,376
Total	34	45,247	34	42,157	68	87,404

 

We used the compiled corpus to assess the presence and the frequency of each marker of engagement and to compare them across sub-corpora. In order to obtain reliable results, we had to combine automated results returned by the software and manual processing of concordances. Starting with the search for individual words (e.g., obviously, could, our), patterns (adj to V), punctuation marks (dashes and parentheses), we collected lists of concordances, which were then analysed to remove instances irrelevant to engagement. Each occurrence was studied within a broader context to identify its discourse features and confirm relevance to engagement. These instances include engagement markers in quotations, use of ‘we’ referring to researchers exclusive of the reader, non-addressee modals (e.g., pronunciation instruction should help make learners comfortably intelligible), comments in parentheses and inside dashes that do not serve as asides. To identify imperatives, following the procedure described by Hyland and Jiang [21], we created a special search rule describing the position of the verb in the base form in the text. As a result, frequencies of all engagement markers were calculated as raw numbers for each sub-corpus, and then normalised per 10,000 words in order to neutralise differences in the word count of the sub-corpora.

We used Pearson’s chi-squared test (χ²) to evaluate how likely it is that any observed difference between the sets arose by chance.

Analytical Framework. Following Hyland & Jiang’s model of engagement [21], quantitative analysis was used to determine the frequency of engagement markers (reader mentions, asides, shared knowledge, questions, directives) in discipline specific corpora composed of RAs Discussion sections written by the authors of different linguacultural backgrounds. Statistical methods enabled us to identify statistically significant differences in the use of engagement markers across medical and foreign language teaching RAs.

Results

The conveyed quantitative cross-disciplinary analysis predictably revealed that engagement features are significantly more apparent in Anglophones’ papers in comparison with non-Aglophones’ ones (Appendix 2). Moreover, even in papers by Anglophones, the frequency of engagement markers is higher in RAs on foreign language teaching than in medical RAs (χ² 26.994, p-value <0.01) (Table 3). The results can be attributed to the inherently argumentative and interpretive nature of knowledge in the humanities, which leads researchers to engage more intensely with readers and foster an atmosphere of solidarity, anticipating criticism and acknowledging alternative views. In FLT sub-corpora, the most significant engagement features are directives addressing the readers directly, instructing and guiding them to the desired interpretation of the argument [36]. They are appreciably more frequent in the FLT sub-corpus than in the MED sub-corpus (FLT: 24.5; MED: 7.6).

 

Table  3.  Frequency of engagement categories in the two disciplinary sub-corpora

Engagement markers	MED		FLT
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader-mentions	49	11.3	52	11.8
Asides	3	0.7	13	3.0
Shared knowledge	61	14.1	58	13.2
Questions	1	0.2	6	1.4
Directives	33	7.6	108	24.5
Total	147	33.9	237	53.9

 

Comparing the reader mentions and appeals to shared knowledge which include the reader into the discussion of the results and appeal to him/her seeking his agreement within apparently naturalised boundaries of disciplinary understandings [18], revealed that both sub-corpora (FLT and MED) are practically the same for the Anglophones and multilinational teams of writers. It goes against the statement of Hyland [18] that explicit appeals to collective understandings prevail in the soft papers whereas the writers of scientific papers expect their readers to have profound knowledge and signal their appeals less explicitly. At the same time, discussion sections of FLT RAs contain more personal asides and questions due to the more communicative and interpretative character of the humanities (3.0 and 1.4 respectively).

As it can be seen from Table 4, there is no significant difference in the use of engagement markers in the medical articles written either by native speakers or by multinational teams (χ² 3.002, p-value <0.01). It highlights that the declarative character of the medical discourse limits the engagement of the reader. However, some researchers [2] point out that not all studies in the field of medicine are aimed exclusively at medical professionals, but also hold broad interest for potential patients. This can also result in a wide range of metadiscourse markers in the Discussion section of such papers.

 

Table  4.  Frequency of engagement categories in the Medical sub-corpora

Engagement markers	MED NS		MED MULT
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader-mentions	25	10.8	24	11.9
Asides	3	1.3	0	0
Shared knowledge	28	12.1	33	16.3
Questions	1	0.4	0	0
Directives	18	7.8	15	7.4
Total	75	32.4	72	35.6

 

However, the analysis of Discussion section of foreign language teaching RAs (Table 5) revealed that Anglophone researchers use engagement language more frequently than those who use ELF for communication (χ² 18.367, p-value <0.01). The significant difference is observed in the use of the directives and personal asides. Native speakers outnumber multilingual authors and involve engagement categories more frequently. Appeals to shared knowledge are exploited more extensively by the Anglophones as the Anglophone academic culture places a stronger emphasis on building rapport with readers, creating a sense of connection, and expressing personal opinions and perspectives. Interestingly, Dontcheva-Navratilova found a less frequent use of appeals to shared knowledge by the native speakers [20]. As for reader-mentions, intercultural variation is not observed in the both sub-corpora.

 

Table  5.  Frequency of engagement categories in the Foreign Language Teaching sub-corpora

Engagement markers	FLT NS		FLT MULT
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader-mentions	25	11.3	27	12.3
Asides	10	4.5	3	1.4
Shared knowledge	35	15.8	23	10.5
Questions	4	1.8	2	0.9
Directives	64	29.0	44	20.1
Total	138	62.4	99	45.2

 

Table 6 highlights that intercultural variation in the NS and MULT sub-corpora is statistically significant (χ² 18.526, p-value <0.01). The Anglophones use engagement markers more frequently in both FLT and MED RAs. This can be explained by the fact that native speakers are generally more familiar with the conventions and nuances of that language, including the use of engagement markers. They have grown up speaking and interacting in the language, which gives them a natural advantage in employing these markers effectively [37].

 

Table  6.  Frequency of use of engagement categories in the NS and MULT sub-corpora

Engagement markers	NS		MULT
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader-mentions	50	11.1	51	12.1
Asides	13	2.9	3	0.7
Shared knowledge	63	13.9	56	13.3
Questions	5	1.1	2	0.5
Directives	82	18.1	59	14.0
Total	213	47.1	171	40.6

 

As Tables 7 and 8 indicate, there is a statistical difference in the involvement of engagement categories across disciplines in both sub-corpora (χ² 16.245, p-value <0.01 in NS and χ² 13.539, p-value <0.01 in MULT). This is consistent with our first finding (Table 3) that the FLT sub-corpus outnumbers the use of engagement features in both language sub-corpora.

 

Table  7.  Frequency of engagement categories in the Anglophone sub-corpora

Engagement markers	MED NS		FLT NS
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader mentions	25	10.8	25	11.3
Asides	3	1.3	10	4.5
Shared knowledge	28	12.1	35	15.8
Questions	1	0.4	4	1.8
Directives	18	7.8	64	29.0
Total	75	32.4	138	62.4

 

Table  8.  Frequency of use of engagement categories in the ELF sub-corpora

Engagement markers	MED MULT		FLT MULT
	Raw number	Per 10,000 words	Raw number	Per 10,000 words
Reader mentions	24	11.9	27	12.3
Asides	0	0	3	1.4
Shared knowledge	33	16.3	23	10.5
Questions	0	0	2	0.9
Directives	15	7.4	44	20.1
Total	72	35.6	99	45.2

 

Categories and Examples of Engagement Markers. Further sub-sections will provide specific results regarding the types of engagement markers, highlighting both the similarities and differences observed in the sub-corpora.

Reader Mentions. The rhetorical interaction between the writer and the disciplinary readership is realised through the use of personal pronouns and possessives. This inclusion helps the authors to reduce the distance between them and the reader, addressing the reader directly. The FLT and MED sub-corpora show almost the same rate of using first person forms as the markers of the reader’s involvement (Table 9 and Table 10).

 

 

Table  9.  Reader mention features across the sub-corpora (raw number)

Reader mention features	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
we/our/us	22	19	22	25	44	44	41	47
one/one’s	5	6	2	0	7	6	11	2

 

Table  10.  Reader mention features across the sub-corpora (normalised per 10,000 words)

Reader mention features	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
we/our/us	10.0	8.6	10.9	10.8	10.4	9.7	9.3	10.8
one/one’s	2.3	2.7	1.0	0	1.7	1.3	2.5	0.5

 

It can be treated either as the act of politeness or the close bond between the writer and the reader being the members of the same disciplinary community and revealing agreement and the feeling of solidarity. Example 1 positions the reader as an experienced colleague who shares the same knowledge about the case under discussion and is competent enough to be engaged in the argument. In Example 2 the reader is considered as a peer capable of forecasting the possible clinical effects of the treatment.

1) We can conclude that there is a greater risk of IPD due to resistant serotypes after being treated with long-duration macrolides, considering the individual and population factors involved (MED MULT).

2) This study further refines our understanding of both clinical effects of TI and the appropriate dosing relationships of TI compared with commonly used prandial insulin analogues (MED NS).

As it can be seen from Table 10, multilingual sub-corpus reveals very slight differences in comparison with native speakers, though the difference in disciplines is more evident, especially in case of using indefinite reference devices (one/one’s). This indefinite pronoun occurs more frequently in the FLT sub-corpus (2.5.vs 0.5). Example 3 treats the reader as a researcher in general. At the same time, the MED sub-corpus shows very low incidence of the indefinite pronoun. The low occurrence of this device can be attributed to the nature of medical articles. Medical articles maintain an objective tone, focusing on factual information. Thus, the authors strive to present information in an unbiased manner and avoid uncertainty [38].

3) Although intuitively one might suppose that singing a song should be more beneficial for L2 pronunciation development than merely listening to it, this generalization might not hold with beginner-level learners of a language or with participants who are not used to singing (FLT MULT).

Personal asides. Personal asides allow the writers to address readers directly by temporarily interrupting the argument to suggest a statement on what has been said. Hyland & Jiang [21] claim that there has been a decline in their use in academic discourse. It is not surprising that the occurrence of this engagement feature in our corpus is also rare (MED 0.7 vs FLT 3.0; NS 2.9 vs MULT 0.7), although Anglophones use this marker more, especially in FLT. Example 4 and Example 5 present the author’s comments that interrupt the argumentation and appeal to the reader. These additional comments imply that the reader has the same expertise and is able to understand the details of the study.

4) The obvious strategy to ensure a presence in the closing stages of a race would be for the athlete to run close to their CS , but only if successful in generating a “gap on the field”, (so that other runners are not able to draft in their slipstream), whereas others are running below their CS and “wasting” time (MED NS).

5) When this same group subsequently took the L1-to-L2 test, the advantage of semantic relatedness (again, in the absence of form similarity) only occurred when the direction of translation at study had been from L1 to L2 (FLT NS).

Appeals to Shared Knowledge. The feeling of sharedness with the reader is created through the use of positioning features, one of them being appeal to shared knowledge. Although Lafuente-Millán believes that appeals to shared knowledge can be identified only by the disciplinary members and are difficult to be identified by the scholars outside the field, we managed to find the examples in the sub-corpora [24]. Following Hyland [39], we treated appeals to shared knowledge as the markers that do not construct new knowledge but try to persuade the readership of already existing claims in academia.

As in the case of reader mentions, the use of appeals to shared knowledge in the MED and FLT sup-corpora differs insignificantly (14.1 vs 13.2 respectively). In the MED sub-corpus, the multilingual authors outnumber the Anglophones in their use of appeals to shared knowledge (MED NS 12.1 vs MED MULT 16.3), while in the FLT sub-corpus native speakers tend to employ more positioning engagement markers of this type. The rhetorical potential of the both sub-corpora allows the author to position the reader as an expert within the discipline and share the common methods and notions. It is more interesting that the multinational teams of authors show a higher rate of using appeals to shared knowledge than the native speakers in medical papers. It can happen due to the desire of the authors that use ELF for communicating research to establish a connection with an audience of medical experts. Appeals to shared knowledge in this case allow them to speak a common language with readers and to strengthen the bond between the author and the audience. Moreover, appeals to shared knowledge can help avoid misunderstandings and ambiguities in medical papers [19], make the text clearer and more understandable for readers, especially if they have different language and cultural backgrounds.

Scholars [21] differentiate between three subcategories of appeals of shared knowledge: referring to logical reasoning (e.g., obviously, of course); to routine condition (e.g., normally, regularly); and to familiarity with tradition (e.g., common, traditionally). Tables 11 and 12 show the distribution of these engagement markers across sub-corpora. Logical reasoning markers show the highest frequency in the FLT NS sub-corpus while the smallest number of these features are found in the MED NS sub-corpus. This difference can be attributed to the specific nature of FLT which often requires clear explanations of language concepts and in-depth knowledge. In contrast, multinational authors of medical papers may rely more on seeking support and validation from their colleagues, making them use logical reasoning markers more extensively.

 

Table  11.  Appeals to shared knowledge across the sub-corpora (raw number)

Shared knowledge appeals	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
logical reasoning	8	18	11	7	19	25	26	18
routine conditions	2	0	1	4	3	4	2	5
familiarity & tradition	10	16	16	16	26	32	26	32

 

Table  12.  Appeals to shared knowledge across the sub-corpora (normalised per 10,000 words)

Shared knowledge appeals	FLT MULT	FLT NS	MED MUL	MED NS	MULT	NS	FLT	MED
logical reasoning	3.6	8.1	5.4	3.0	4.5	5.5	5.9	4.1
routine conditions	0.9	0.0	0.5	1.7	0.7	0.9	0.5	1.2
familiarity & tradition	4.6	7.2	7.9	6.9	6.2	7.1	5.9	7.4

 

Routine conditions prevail in the MED NS sub-corpus while no instances of this strategy are found in the FLT NS sub-corpus. It contradicts the supposition of Zou & Hyland [36] that these markers are used in research on education. Due to the specificity of medical research this engagement feature denotes the routine conditions under which the research was carried out.

Familiarity & tradition are distributed almost equally in three sub-corpora as can be seen in Table 12 except for FLT MULT. It can be explained by the desire of the non-native authors to avoid wordiness due to limited knowledge of the cultural aspects and traditions of the country requiring the authors to use them correctly. If authors are unsure about their knowledge of the culture, they may avoid using such markers to prevent potential misinterpretations or inaccuracies. Moreover, both MED MILT and MED NS show the highest rate of frequency. These results are in line with Zou & Hyland [36] who supposed that hard disciplines are likely to formulate knowledge in an established knowledge context.

Thus, logical reasoning in Example 6 concerns the coherence of the argument. Hyland & Jiang [21] believe that this way of engagement is the most manipulative as it shifts the focus from the discourse and makes the reader anticipate possible criticism and shape their own understandings, as it can be seen in Example 7:

6) Clearly many factors affect performance when learning L2 vocabulary, and they may operate differently depending on whether a person is learning for a receptive or a productive test situation (FLT NS).

7) Some of the concern over the apparent difference in the unit-dose effect arises from confusion between the definition of an insulin unit, both in terms of clinical “effect” and unit “amount” (MED NS).

Routine conditions are concerned with usual circumstances or behaviours in the real world. Examples 8 and 9 show that these engagement features refer to shared knowledge of conventional research procedures and methods of analysis:

8) The presence of SIH for hospitalized patients is conventionally defined as BGL ≥ 10 mmol / L. In this cohort with a widely varying HbA1c range, glucose was poorly associated with outcome and BGL ≥ 10 mmol / L not a reliable indicator of the presence of SIH (MED NS).

9) Failure to learn English, therefore, is not typically considered as a failure with costly negative consequences (FLT MULT).

Familiarity & tradition are concerned with usual community practices and beliefs. Examples 10 and 11 appeal to the reader’s supposed ’familiarity with the background of the research:

10)  Not surprisingly, the upfront culture methods had lower LOD due to the pre-amplification that occurs naturally in culture prior to testing (MED NS).

11)  CLIL is usually associated with L2-medium instruction, although it can be implemented with different intensity of the L1 involved (FLT MULT).

Questions. Questions are considered to be the most widespread means of dialogic engagement [40] involving the readers into the discussion, making him/her an equal participant in the argument and showing willingness to share knowledge. Still, the corpora reveals that the use of questions in RAs is not frequent despite their rhetorical nature. As in case of personal asides, the Anglophones use questions more frequently than their multilingual counterparts (1.1 vs 0.5). Similarly, the FLT NS community tends to use more questions due to the communicative focus of the discipline. In the medical corpora of both language groups the use of questions is very low (MED NS 0.4 vs MED MULT 0). Examples 12 and 13 show the use of this engagement marker as a one-side exposition to the dialogue, making the reader feel that their views matter:

12)  How well do English teachers succeed in bridging the gap between elementary and secondary school? (FLT MULT).

13)  What about our immersed learners? (FLT NS).

Directives. Directives tend to instruct the reader and engage him/her into discussion through imperatives and modals. Among all the engagement markers in our study, this group is the greatest. It shows significant differences in the frequency of use in two disciplines (MED 7.6 vs FLT 24.5). Also, the scholars using ELF for conducting their research employ directives less than the natives (NS 18.1 vs MULT 14.0). The differences within the disciplinary groups are less significant (FLT NS 29.0 vs FLT MULT 20.1; MED NS 7.8 vs MED MULT 7.4). Such frequent occurrence in academic discourse reveals that directives appear to be a good rhetorical tool allowing the author to impart his knowledge and findings to the reader clearly⁸ [18; 21]. Hyland [41] claims that frequencies for directives seem to be higher in the hard sciences helping the authors share and present information succinctly by directing readers through an argument. On the contrary, Dontcheva-Navratilova argues that in soft sciences the use of directives can be explained by their frequent use to indicate intertextual reference [20]. Nevertheless, our research shows that the FLT sub-corpus uses directives more frequently. The way directives are realised across the corpora are shown in Tables 13 and 14.

 

Table  13.  Directives across the sub-corpora (raw number)

Directives	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
Imperatives	24	43	1	1	25	44	67	2
Adj to V	4	7	2	2	6	9	11	4
Modals	16	14	12	15	28	29	30	27

 

Table  14.  Directives across the sub-corpora (normalized per 10,000 words)

Directives	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
Imperatives	10.9	19.5	0.5	0.4	5.9	9.7	15.2	0.5
Adj to V	1.8	3.2	1.0	0.9	1.4	2.0	2.5	0.9
Modals	7.3	6.3	5.9	6.5	6.6	6.4	6.8	6.2

 

Among all the directives, imperatives are employed most frequently. Table 14 shows their distribution highlighting that in the FLT context they outnumber those used in the MED sub-corpus, with their use in the MED sub-corpus being the lowest. Again, the NS sub-corpora include more imperatives than the ELF sub-corpora. The most frequently used imperatives are see, consider, as in the Example 14 and 15 below:

14)  Some outcomes were not assessed in the entire cohort, because of missing data (Table 2) (MED MULT).

15)  Recognise Inequalities in Knowledge of Specialised Discourses (FLT NS).

Predicative adjectives express the author’s opinion about necessity or importance of the findings in the study (Example 16). Also we can observe a cross-disciplinary difference in the use of predicative adjectives (FLT 2.5 vs MED 0.9), while the rate of occurrence in different language contexts remains relatively close (MULT 1.4 vs NS 2.0). There is no difference in the use of these engagement features within the MED sub-corpus, while we can observe a slight shift towards NS within the FLT sub-corpus (FLT MULT 1.8 vs FLT NS 3.2).

16)  Finally, it is important to note that our account does not question the importance of high verbal analytic ability in later-onset learners; rather we suggest that the effects of explicit aptitude reported in previous studies should not be interpreted to entail a general loss of implicit learning ability (FLT NS).

Obligation modals have a great impact on the reader and are employed by the authors equally in our sub-corpora (MULT 6.6 vs NS 6.4; MED 6.2 vs FLT 6.8). Both medical papers and those on teaching language tend to impose obligations due to the nature of the field.

17)  These results indicate that alongside MyD88 signalling, the role of other receptors and pathways in recognition of S. suis and induction of inflammation should be evaluated (MED NS).

18)  However, it must be noted that the two other longitudinal studies we found were carried out with older students (age range: 11–15 years) and used different instruments (FLT MULT).

Functions of directives across sub-corpora are shown in Tables 15 and 16.

 

Table  15.  Functions of directives across the sub-corpora (raw number)

Directives	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
Textual acts	24	38	1	1	25	39	62	2
Cognitive acts	20	21	14	17	34	38	41	31
Physical acts	0	5	0	0	0	5	5	0

 

Table  16.  Functions of directives across the sub-corpora (normalised per 10,000 words)

Directives	FLT MULT	FLT NS	MED MULT	MED NS	MULT	NS	FLT	MED
Textual acts	10.9	17.2	0.5	0.4	5.9	8.6	14.1	0.5
Cognitive acts	9.1	9.5	6.9	7.3	8.1	8.4	9.3	7.1
Physical acts	0	2.3	0	0	0	1.1	1.1	0

 

Textual acts guide the readers through the body of the RA or refer them to another text as in Example 19. They are mostly used in FLT context by native speakers (Table 16). This can be explained by the interpretive character of the discipline, moreover it helps to direct the readers in the scope of the research helping them to understand the full picture.

19)  VII Discussion A time-consuming practice for L2 writing instructors, WCF continues to be perceived by students as something that can help them develop accuracy (see for review Bitchener & Storch, 2016; see also Storch & Wigglesworth, 2010) (FLT MULT).

20)  Cognitive acts lead the readership through a line of reasoning to steer them to certain conclusions (see Examples 20–23). Again, they are more frequent in FLT (FLT 9.3 vs 7.1), establishing no difference between different language contexts and disciplines. These engagement features tend to make the reader think, reflect on some aspect of the argument.

21)  Finally, it is important to note that our account does not question the importance of high verbal analytic ability in later-onset learners; rather we suggest that the effects of explicit aptitude reported in previous studies should not be interpreted to entail a general loss of implicit learning ability (FLT NS).

22)  Thus, when analyzing immunity elicited by vaccination it is essential to focus on the fine granularity (e.g. S. Typhi-specific TM subsets, whether the responses are S versus MF, which combinations of cytokines are produced following antigenic stimulation) to better characterize the responses and properly study differences between immune compartments (MED NS).

23)  However, limitations must be acknowledged in the current study (MED MULT).

24)  However, at this point, the possible reasons for higher adjective frequency counts leading to slower response times need to be considered (FLT MULT).

Our results prove the supposition of Zou & Hyland [36] that physical acts are mostly used in papers on education (FLT 1.1 vs MED 0.0) and help create space to foresee the author’s outcomes and involve the reader’s actions (Example 24). Native speakers use these markers more frequently (Table 16).

25)  Pay Critical Attention to Inequalities of Power (FLT NS).

The obtained results, overall, reflect the findings of previous studies that indicate that disciplinary variation is influenced by the epistemology, methods, and conventions specific to different disciplines⁹  [19; 24; 36]. However, it is necessary to trace the internal logic of deepening the author-reader engagement in the manuscript, especially considering potential asynchronous readers. This calls for further research on the topic across various disciplinary corpora.

There are noticeable differences at the disciplinary level, with FLT employing engagement markers more frequently. The differences are evident between the Anglophone authors and the authors that use ELF for communicating research across disciplines. Both sub-corpora use the specified markers similarly in medical papers, while in FLT papers, native speakers use a greater number of markers, positioning features being particularly prominent.

The comparative analysis of engagement markers in RAs by Anglophone scholars and non-native scholars using ELF showed that the differences in overall frequency of engagement categories between these contexts are not significant. In this regard, the standards of academic culture specific to a particular country and the characteristics of the journal for which the manuscript is prepared for publication come to the forefront.

The analysis of subcategories of engagement revealed significant differences between RAs written by the authors using ELF and RAs published by the Anglophones in terms of directive types and functions. The orientation of English-speaking authors, driven by the norms of academic writing¹⁰ , towards a heterogeneous audience determines the frequency with which they provide instructions to readers and the number of steps they take to direct readers in the logic of their arguments. Non-native English authors using ELF in FLT RAs differ in levels of development in academic culture and may not pay as much attention to understanding the heterogeneity of the academic community to which they address their research results. They may not place great importance on building solidarity and referring to shared sources, methods, and understanding. However, the situation is opposite in medical papers due to the conventions of medical research. Thus, it can be argued that while both anglophone and non-native scholars using ELF equally strive for interaction with their readers, they are inclined to diverge in their preferences for the specific implementation of engagement markers.

According to Lafuente-Millán, national culture is an important variable in the author’s rhetorical choices [24]. However, the global academic English language, which incorporates not only the canons of academic English but also the canons of national academic language paradigms, is gradually shaping a new space of communication. The ability of communicants, whether they are native or non-native English speakers, to engage in the construction and dissemination of scientific knowledge in a global dialogue calls for an examination of the features of academic ELF. Additionally, it necessitates the establishment of a universal framework that transcends national sub-corpora, promoting unity, transparency, and effective communication without further divisions based on national boundaries. The establishment of global conventions for such communication should involve collaboration between native and non-native speakers. Hence, researchers should pay close attention to the strategies employed by non-native speakers in using language to engage readers within the framework of authorial logic.

One limitation of this study is the characteristics of the corpus used, which consisted only of international second-quartile journals. A parallel analysis of national journals would have allowed for a comparative examination of the influence of journal status on the engagement strategies used by authors to involve readers in their discourse. For example, it would have provided insights into how authors might feel more at ease when addressing a national audience. Would this lead to a more frequent mention of the potential reader as a member of a group sharing common interests and values? Would it reveal a dependency on the audience (addressing fellow citizens vs. an international audience)? Nevertheless, the aim of this study was to track the authors' behaviour in the context of using English as an academic lingua franca. It focused on the effectiveness of reader engagement conventions in facilitating the understanding of presented information for non-native speakers, prioritizing the principle of comprehensibility over grammatical accuracy in English proficiency [27].

Discussion and Conclusion

This study has adopted a contrastive approach to the investigation of cross-disciplinary and interlanguage variation in medical and FLT NS RAs written by Anglophone and non-native scholars using ELF. The results of this study have demonstrated that disciplinary factors and the status of English as a lingua franca influence the choice of engagement markers in RAs in different ways. Disciplinary conventions appear to have a significant impact on the rhetorical choices that play a crucial role in persuading the academic community to accept the proposed interpretations and claims. However, the specific engagement markers chosen by authors are determined not only by cultural considerations but also by the language status for the manuscript author.

This conclusion may seem obvious and not requiring further evidence. However, it pertains to the creation of a universal language for academic scientific communication in the context of a heterogeneous readership and research territory. The next stage can be refining our understanding of how the trends in the development of a universal language for academic communication and disciplinary involvement contributes to persuasive communication in an academic context. This interplay should be studied further and these findings need to be expanded and tested through larger-scale studies that take into account the interaction of engagement markers and positioning in different disciplines and linguocultural contexts.

 

¹           Swales J. Genre Analysis: English in Academic and Research Settings. Cambridge: Cambridge University Press; 1990; Swales J. Research Genres: Explorations and Applications. Cambridge: Cambridge University Press; 2004. https://doi.org/10.1017/CBO9781139524827

²           Hyland K. Disciplinary Discourse: Social Interactions in Academic Writing. Longman; 2000.

³  Fløttum K., Dahl T., Kinn T. Academic Voices – across Languages and Disciplines. John Benjamins Publishers; 2006. https://doi.org/10.1075/pbns.148

⁴  Jenkins J. English as a Lingua Franca in the International University. London and New York: Routledge; 2013. https://doi.org/10.4324/9780203798157

⁵  Mauranen A. Exploring ELF: Academic English Shaped by Non-native Speakers. Cambridge: Cambridge University Press; 2012. https://doi.org/10.1002/tesq.88

⁶  Kirkpatrick A. Introduction. In: Kirkpatrick A., ed. The Routledge Handbook of World Englishes. Routledge; 2010. p. 1–14.

⁷  Dash N.S., Arulmozi S. Features of a Corpus. In: History, Features, and Typology of Language Corpora. Springer, Singapore; 2018. https://doi.org/10.1007/978-981-10-7458-5_2

⁸  Hyland K. Disciplinary Discourse: Social Interactions in Academic Writing.

⁹  Fløttum K., Dahl T., Kinn T. Academic Voices – across Languages and Disciplines; Gray B. Linguistic Variations in Research Article. When Discipline Tells Only Part of the Story. Amsterdam & Philadelphia: John Benjamins; 2015. https://doi.org/10.1075/scl.71

¹⁰  Academic writing in the United States has a long and rich history. The standards set by academic writing in the U.S. have had a significant influence on the development of academic writing in other countries. However, it was European countries that first embraced and incorporated these standards into their own academic communication practices, while other countries have adopted them to varying degrees, integrating them into their national traditions of scholarly discourse.

 

Sobre autores

Elena Tikhonova

MGIMO University

Email: etihonova@gmail.com
ORCID ID: 0000-0001-8252-6150
Scopus Author ID: 57208387246
Researcher ID: B-1951-2015

Cand.Sci. (History), Associate Professor

Rússia, 76 Vernadsky Avenue, Moscow 119454

Marina Kosycheva

HSE University

Autor responsável pela correspondência
Email: mkosycheva@hse.ru
ORCID ID: 0000-0003-0328-3109
Scopus Author ID: 57219614422
Researcher ID: ABB-3567-2020

Cand.Sci. (Philol.), Associate Professor, Associate Professor of the School of Foreign Languages

Rússia, 20 Myasnitskaya St., Moscow 115054

Tatiana Golechkova

New Economic School

Email: tgolechkova@nes.ru
ORCID ID: 0000-0002-1639-4556

Cand.Sci. (Philol.), Associate Professor of the Department Humanities and Languages

Rússia, 3 Nobel St., Moscow 121205

Bibliografia

Arquivos suplementares