Introduction

The global spread of English has enriched the varieties of English and the number of speakers and learners of English as a foreign language will likely continue to grow in the coming years (Graddol, 2006; Selvi, 2014). The expansion of English in outer and expanding circle countries has made new speakers of English emerge, as well as “new Englishes variety” as it has come into contact with speakers’ first language. Kachru (1992) introduced the classification of English speakers with his “three concentric circles” model; the Inner circle – English as the first language, the outer circle – English as a second language, and the expanding circle – English as a foreign language. The teaching of English in the expanding circle is often oriented to achieving competence and proficiency like that of native speakers. It is evident, for instance, in the assessment of speaking accuracy or pronunciation which considers the local accent to be deficient (Sung, 2014). Besides, it is also apparent in the teaching of paragraph writing that dictates non-native English students to follow the paragraph writing model of the “West”: topic sentence, supporting sentence, and concluding sentence. Under this rule, students are not allowed to use an interrogative sentence in the concluding sentence (Phan, 2018) although such a usage may exist in other languages and may still be considered logical in their own way (see Phan, 2009).

As a result, ELT practices in the expanding circle frequently adopt a monolingual approach and are dependent on native-English norms, which directly or indirectly perpetuate English domination and supremacy (Phan & Le, 2013; Wahyudi, 2018). EFL students in the South East Asian region who have a different writing style, for example, being indirect, often find it difficult when writing in English to successfully produce a native-like writing model. To overcome this obstacle, many scholars (cf, Halliday & Matthiessen, 2014; Hyland, 2005) proposed guidelines of producing coherence in writing (and speaking) by introducing cohesive devices, cohesion in English, and metadiscourse markers. Apart from different terms, they share similar functions as to produce persuasive and coherent writing or speaking in the English language context (Hyland, 2005; Hyland & Tse, 2004) perceived necessary for EFL learners.

The term metadiscourse, which first appeared in Harris (1970), has been employed to study language in-use in numerous contexts, especially in writing (Hyland, 2005). Since then, metadiscourse is understood as “writing about writing” which does not refer to the subject matter being addressed (Williams, 1982. p. 212), to be a “linguistic unit that does not add propositional content” (Kopple, 1985, p. 83), and as “a linguistic element which goes beyond propositional content” (Khedri et al., 2013, p. 320). Hyland (2004), in addition, views metadiscourse as “self-reflective expression referring to the evolving text, writer, and imagined reader” (p. 133). Hyland can be considered one of the most prominent scholars in metadiscourse research, which is later connected to the systemic functional linguistic paradigm, presenting ideational, textual, and interpersonal functions of the text. From that point, the metadiscourse plane is categorized into interactive and interactional metadiscourse (Hyland and Tse, 2004).

Metadiscourse markers are a prominent vehicle to transfer, interact, or communicate thought to the readers or listeners (cf. Hyland, 2017; Hyland and Tse, 2004). In other words, they have a pragmatic and social function to provide interlocutors with adequate resources to interpret meanings of their utterances. Nonetheless, Hyland (2017) has distinguished metadiscourse from metalanguage or metapragmatics. The former is synonymous with using a language to describe another language, while the latter is on par with “speakers’ judgment of communicative behavior” (p. 17). He also emphasizes that metadiscourse markers have fluid forms and functions or categories. The same form of metadiscourse markers might function in different ways. This rationale underscores the complexity and fluidity, in terms of categorization and analysis, of metadiscourse markers. This indirectly suggests to researchers not to rely solely on existing form or function but to look comprehensively at the components they appear with.  

Numerous studies examine metadiscourse markers in written texts in different domains, for example in research articles (Gholami & Ilghami, 2016; Hu & Cao, 2015), academic writing (Hyland, 2004, 2005; Park & Oh, 2018), and in argumentative writing of non-native speakers (Taka?, 2018; Tan & Eng, 2014; Yeganeh & Ghoreyshi, 2015). These studies have found that metadiscourse use in academic writing signals speakers’ attitude in guiding, and to another extent, convincing readers by formulating well-constructed and connected discourse units. In addition, this use may also be influenced by a number of factors such as the field of study (Heng & Ebrahimi, 2013; Hyland, 2005), genre (Kuhi & Behnam, 2011), and culture (Kim & Lim, 2013). Studies on metadiscourse markers also take non-native English as the research scope. Farahani and Sabetifar (2017) studied EFL students’ writing and pointed out that in some particular cases, Iranian EFL writers and native English writers differ in using metadiscourse markers. This finding is relevant with Gholami and Ilghami (2016) who studied academic writing within the field of biology from both native and non-native English speakers in a way that, for the sake of making coherence, non-native writers tend to use interactive and interactional metadiscourse markers more frequent than native writers

Scholars also pay attention to Indonesian EFL writing competence. Aunurrahman et al. (2017) investigated 36 Indonesian undergraduate students’ writing. By employing Halliday & Matthiessen’s (2014) functional linguistics, they demonstrated that students often make use of inaccurate grammatical patterns and lack of complexity and coherence. They contended that such a case might be influenced by a limited understanding of the use of proportional metadiscourse markers in writing. Gunawan and Aziza (2017) confirmed that such a meaning making process could also appear due to cultural factors. Their research on undergraduate theses found that themes, be they topical, interpersonal, or textual, are salient tools in the teaching of good academic writing across boundaries. In regard to the metadiscourse markers we examine, this finding is worth considering to shed light on how different cultural backgrounds may also lead to typical lexical choice in writing, especially in navigating coherence and cohesion.    

In addition to written discourse, student presentations are a potential domain to investigate interactional or interactive functions of metadiscourse markers which remain rarely explored. Adel’s (2010) work on metadiscourse taxonomy in spoken and written discourse can be cited as the most influential example for the present study since it provides us clues and directions on how to deal with the use of metadiscourse markers by Indonesian EFL students in the present study. This taxonomy is later employed in a presentational setting by Correia et al. (2015). Speakers, like writers in academic writing, also need to engage with the audience and to make their speaking coherent. This paper therefore aims to explore and compare the use of metadiscourse in undergraduate students’ writing and oral presentations. The data range from seventy students' argumentative essays and eight student presentations in Kencana University (pseudonym), one of the Islamic universities in Malang, East Java, Indonesia. Setting, genre, and EFL factors in the use of such metadiscourse will be taken into account during the analysis.

Metadiscourse Markers

Metadiscourse refers to the strategies used by authors/speakers to build and organize their discourse and implications of the message that is being transmitted (Hyland, 2005: 18). In writing, the use of metadiscourse devices can help readers interpret the proposed propositions or claims. Metadiscourse markers, as proposed by Hyland (1998, 2004), are divided into two categories: interactive and interactional metadiscourse markers. Interactive metadiscourse functions to guide the audience to get the idea of the discourse. It includes the use of five elements:

• transitions which connect some clauses (e.g., additionally, but, and),
• frame markers which are defined as the act of discourse and the stages of sequences (e.g., in conclusion, the aim is),
•  endophoric markers that guide readers to refer to other parts of the text/speech (e.g., as explained above, the following paragraph),
• evidentials, which refers to other information that is written/spoken in other texts/speeches (e.g., according to, based on), and
• code-glosses, which express the propositional meanings (e.g., for example, for instance).

In addition, interactional metadiscourse markers are sub-divided into hedges and boosters. The former function keeps the commitment of the write/speaker such as I guess, maybe, etc.; while the latter is used to accentuate the certainty delivered by the writer/speaker (e.g., in fact, exactly). Other categories are a) attitude markers which stand to show the writer’s position toward a particular suggestion (e.g., luckily, unconsciously, b) Self-mentions that are used to point on his/herself (e.g., I, we, my, our), and c) Engagement Markers that are used to explicitly involve the reader (e.g., you, yours, we (inclusive)).

Earlier studies of metadiscourse markers seem to highlight their occurrence, function, and role in written academic English. For instance, Tan and Eng (2014) and Aull and Lancaster (2014) examined academic writing texts. Tan and Eng (2014) investigated the use of metadiscourse markers by Malaysian undergraduate students as tools for persuading. The study compared high and low English proficiency students (HEP and LEP).  Aull and Lancaster (2014) compared the writing of first-year and upper-level students. They demonstrated that first year students, despite different disciplines, shared linguistic stance markers with hedges/boosters, code gloss, and adversative metadiscourse markers, among others. Lee and Subtirelu (2015) investigated spoken metadiscourse in university lectures, how metadiscourse is used in EAP lessons and by university lectures in different disciplines. Their study further highlights the complexity of metadiscourse markers use in which academic trajectory and teaching models might influence way lecturers guide and unfold students’ ideas. Unlike those studies, this paper endeavors to understand similarity and or difference between metadiscourse marker use in spoken and written language in non-native English learners/speakers contexts.

Metadiscourse in an EFL context is important to investigate, especially in the case of non-native English speakers’ academic writing and presentation. There are many academics who pay meticulous attention to this case. Among them are Taka? (2018) in Croatian EFL persuasive essays, and Park and Oh (2018) in Korean EFL argumentative texts, which scrutinize the relation between metadiscourse, coherence, and writing proficiency.

The present study draws from EFL students’ metadiscourse use in written discourse, and aims to widen the research domains by taking spoken data of EFL students in an Indonesian context into account.

Method

This study employs a case study approach. A limited quantification process was employed to study the frequency of the examined metadiscourse markers. This study aims to explain one or more cases detailed (Johnson & Christensen, 2004) and to substantiate textual analysis and interpretation on metadiscourse markers in both spoken and written data (Creswell, 2008). To present the variations of metadiscourse marker occurrence in EFL students’ academic essays, the frequency of each category was computed per 1000 words. In addition, the researchers computed the frequency per 60 seconds to figure out the distribution of metadiscourse in academic presentations.

Data Collection

Argumentative and academic writing in Kencana University are offered in the third and fourth semesters. The data in this study were gathered from 70 essays with a total of 31,551 words. Metadiscourse markers are likely to be found since they can help writers to convince readers and direct audience involvement in discourse (Ó’del, 2010).  The spoken language data were taken from the speech of eight undergraduate English department students in Kencana University while they performed presentations in English in the following courses: Public Speaking and Academic Writing. These courses are offered in the fourth semester. Students have not taken English language proficiency examinations such as TOEFL or IELTS. However, we believe that their English level might be in B1 given that they have completed A1-A2 level in the first semester prior to enrolling in the language skill courses given from semesters two to four.

The topic of metadiscourse markers is not specifically studied as a course, but as part of paragraph writing topic on transitional phrases or markers. Students already have a preliminary understanding of using transitions in the text to produce coherence. Regarding ethics, we first contacted the lecturers of the corresponding courses, and later their students, and asked for their consent to conduct this research and make use of students’ writing as our source of data.

In the speaking class, students are required to prepare a group and an individual presentation consisting of storytelling, promoting the university, and a mini-debate simulation. We recorded the presentations using a phone recorder and transcribed verbatim. Student presentations ranged from three minutes to ten minutes long. Sentences and utterances where metadiscourse markers appeared (Hyland, 2004, 2005) were regarded as the data for analysis. Surrounding sentences or utterances were beneficial elements in providing detailed context. As for quantifying the frequency of the metadiscourse in the present study, we took all data, both from students’ presentation and writings. The frequency of each sub-category of metadiscourse markers is counted per 1.000 words and 60 seconds (see the data analysis below).

Data Analysis

The data were categorized based on Hyland’s (2005) taxonomy and data which did not fall under the metadiscourse markers category were not selected for the analysis. Afterwards, selected data were categorized into interactive metadiscourse markers (transition markers, frame markers, endophoric markers, code-glosses and evidentials) and interactional metadiscourse markers (hedges, boosters, attitude markers, self-mentions, and engagement markers). The first author first coded and categorized the data based on Hyland’s (2005) taxonomy. The second and third author then double checked them and made some changes when necessary, for instance, ensuring the data for analysis were relevant and matched with given criteria of Hyland’s (2005) metadiscourse markers. We then present a table of frequency before comparing how those categories appear in text and speech. The frequency of occurrence is counted per 1.000 words and 60 seconds with the following formula:

After quantification, the researchers analyzed and discussed the results in accordance with the proposed theory of metadiscourse markers (Hyland, 2005) and relevant studies in of such cases in EFL contexts.

Results and Discussion

The distribution and frequency of metadiscourse markers are shown in Table 1 for both written and spoken data. Table 2 illustrates the distribution of each category of interactional metadiscourse markers. It is worth noting that the interactive devices dominate the occurrence in the present data, with 1616 or 64.6% of the total of cases in the written data and 278, equal to 67.6% of all cases found in the spoken data.  Interactional devices were only 885 cases in the academic writings and 133 cases in the academic presentations. In this section, the explanation of the findings, including the comparison between the written and spoken discourse, are classified in two main discussions: interactive and interactional metadiscourse markers.

Transition markers ranked the highest frequency with 1116 cases (35.3 per 1000 words) in the written data and 160 cases (3.65 per 60 seconds) in the spoken data. Code-glosses appeared the second leading category in interactive devices of the written data with 216 cases found (6.84 frequent per 1000 words), but not in spoken data which appeared in 52 cases only (1.18 frequent per 60 seconds). In addition, evidential markers made a small distinction with the code-glasses in the written data, 6.65 hits per 1000 words with 210 cases. Surprisingly, this category was not found in the spoken data.

As for frame markers, they are the fourth most-used interactive metadiscourse markers. Frame markers occurred in written data almost as frequently as in spoken texts, with 61 cases in the earlier domain (1.39 frequent per 60 seconds) and 63 cases in the latter (6.84 frequent per 1000 words). Finally, endophoric markers do not seem to make a huge difference in either the spoken or the written data. Our findings demonstrated that they appeared in 11 cases in students’ essays (0.34 frequent per 1000 words) and only five cases in the academic presentations (0.11 frequent per 60 seconds), which later put it in fifth place.

Table 1: Interactive metadiscourse markers

In the interactional metadiscourse category, hedges are the most used markers with 523 cases, with 16.57 numbers of frequency per 1000 words. However, in spoken discourse, they are used less with 45 cases (0.95 frequent per 60 seconds). Boosters had a slightly significant difference between the spoken and written data. It became the second leading category in students’ essays with 239 cases (7.57 frequent per 1000 words), but it remained in the third/fourth rank with a similar number of cases with engagement markers (15 cases which equals to 0.34 frequency). In contrast, self-mentions led the count in the spoken discourse with 58 cases, both inclusive and exclusive self-mentions with 1.32 frequency per 60 seconds. On the other hand, this category placed in the third rank in the written discourse with a large number of distinct boosters (80 cases with 2.53 frequency per 1000 words).

Following the prior categories, engagement markers and attitude markers seemed to occur rarely in both spoken and written discourse. Attitude markers were found in 27 cases, 0.85 frequency per 1000 words. However, this category was not even found in the spoken data. Meanwhile, engagement markers occurred less in the written data with 16 cases (0.5 frequency per 1000 words), but it placed the third/fourth place of spoken discourse with 15 cases (0.34 frequency per 60 seconds). Finally, to reach a better understanding regarding each sub-category of interactive and interactional metadiscourse, we provide a rigorous discussion as follows.

Table 2: Interactional metadiscourse markers

Interactive Metadiscourse Markers

The interactive devices refer to the elements that represent a writer/speaker’s strategy to manage his/her propositional claims. This type of text management aims to guide and convince the audience to reach a coherent understanding of the proposed claims (Hyland, 2004, 2005). The following section will discuss each of the interactive sub-categories in detail. The categories are listed based on Hyland’s interactive metadiscourse taxonomy.

Transition Markers (TM)

Transition markers seem to be the most common metadiscourse markers used by EFL students. This shows the tendency of the internal cognitive connection usage as a tool for students in organizing their arguments within discourse. Hyland (2005) divides the categories of Transition Markers into three specific sub-types: addition, comparison and consequence which functions in the external and internal relation. Additional markers are used to extend the information given by adding particular elements and “to express semantic relation” (Hyland, 2005). Meanwhile, comparison markers are represented by the use of elements which give contrary or equivalent information in accordance with the previous sentences. On the other hand, consequence markers refer to the elements that justify the conclusion or counter an argument as pointed out in Table 3.

The use of Transition Markers in written and spoken language reflects a significant distinction. Figure 5 reports the distinction made in each discourse. The result shows that addition markers led the count with 701 cases in students’ argumentative essays, 22.21 points of frequency per 1000 words. This number holds a big gap with the consequence markers as this second leading category only has 7.38 total of frequency. On the other hand, the result differs in the spoken data which reflects more usage of consequence rather than of addition markers. Nevertheless, the gap between the two categories is a small distinction; consequence markers with 1.64 frequency and addition markers with 1.46 frequency per 60 seconds. For the comparison markers, they were the third most used transition markers in written and spoken academic context.

Since additional markers are used to give extended propositional information, their occurrence is led by the use of and in both written and spoken discourse as with example 1. However, the use of and dominates in spoken context, with 64 cases. In other words, the other addition markers such as furthermore, moreover, and in addition, for example, were not found in the spoken data. Nonetheless, as seen in the Example 1, there is an excessive occurrence of and which leads to the overuse of transition markers. The overuse of metadiscourse markers has been investigated in previous research (cf. Hinkel, 2002). Korean, Japanese, Chinese, and Arabic EFL students tend to overuse the phrase-level transitions (and, but, etc.), while native English students attempt to use long and complex transitions such as furthermore, moreover, etc. in their writing (Hinkel, 2002; Hyland, 2005).

Table 3: Categories of transition markers

Such cases also appeared in Indonesian EFL students’ presentations in which they often use and as a form of phrase-level transition to connect the idea that appears before and after it. The additional marker and guides readers to understand the speech - that what appears earlier and what comes after are equally important (Hyland, 2005). What is surprising is the fact that they also employ it in their writing as in example 2. After careful quantification analysis, we found that and occurred on average six times every two paragraphs throughout the essays, up to two times in every two sentences and is often used to begin the sentence as with Example 3. Meanwhile, the longer additional markers appear only once in every two paragraphs. In sum, both Example 1 and Example 2 report that most of the students have not performed effective communication.

Example 1

There are syntactic determination and non-syntactic determination and the one with understood subject given by non-subject NP and understood subject not given by non-subject N: [S]

Example 2

It has been the way for students to communicate, interact and socialize, and has become an integral part of their social life. [W]

Example 3

Moreover, the attraction will interest people to come, watch and assemble at sidewalks. [W]

Transition markers remain the second most used category in written text and were used by students in various forms. Nonetheless, so, along with and, convincingly dominated the spoken data as in example 4. The example shows the repetition of the word so in every new sentence that the students convey. On the other hand, the occurrence of the similar words such as therefore or in conclusion is not found in the spoken data. Furthermore, the word so was also found in 10 of 70 essays as seen in Example 5. This might indicate that some students cannot separate the written and spoken context.

Meanwhile, the comparison and contrast category remains the less used Transition Marker in either spoken or written data. The statistics of the two data also indicated a substantial use of the word but which represents a difference of arguments or information within a discourse. In Example 6, the writer tried to argue that religion is not only a choice, and added a contrast on the argument by extendedly explaining that it is also a result of individual decision. However, unlike the previous most-used markers in the two former categories, the word but is not overused in both types of data. This indicates the assumption that students are mostly directive with an argument without often involving a contrary opinion.

Example 4

So, we talk about the subject less that appears in non-finite. So, this is the non-finite. So, there is supposed to be someone brings something in here. So, in syntactic determination… [S]

Example 5

So, the role of government is very needed to take the right consideration of some statement or perspective of people who agree… [W]

Example 6

The religion column must be attached because religion is not just a choice but is a result of the individual decision as identity. [W]

Example 7

In addition, submitting the final evaluation of students to each region as well as a form of justice in testing the ability of students …, so that testing must also be adapted to their abilities. [W]

The last example above reports the use of three types of Transition Markers in such a long single sentence. The first marker that appears is the additional marker in addition used to begin the sentence. This marker contradicts the previous argument in a new argument in the sentence. Meanwhile, the transition as well as is categorized as a form of comparison which changed the statements of final evaluation as an association with form of justice. Following the comparison marker, the word ‘so’ as a spoken form of conclusion marker appears to justify the conclusion of the essay discussion.

Frame Markers (FM)

This category of metadiscourse often appears as a guide for the readers or the audience to understand the text by giving signals of the text elements in the structure. Frame markers include the words of label, sequence and argument shift, and prediction (Hyland, 2005). In the spoken and written data, the most frequent frame markers that occur in the text are the words of sequence (example 8 and 9). This type of frame marker represents the explicit additive relations between one point and another. Meanwhile, example 10 illustrates the frame markers use in labeling the stages of a text with particular words such as in sum, in conclusion, to summarize, etc., which were not found in the spoken data. In any case, Hyland (2005) also proposed the mention of discourse goals such as I hope to, I argue, etc. as a form of frame marker. However, the researchers could not find such phenomena either in written or in spoken data.

Example 8

… describe the three factors that led to the growing of schools with full day system. First, many parents are busy working outside the house … Second, teachers in schools face difficulties … Third, the curriculum demands that the fit current development era takes longer in integrating students. [W]

Example 9

The next part will be explained by her. [S]

Example 10

In sum, I do think that Sandi’s statements are not operative at all. [W]

Hyland (2005) demonstrated that the signals of topic shifts might be represented by well, okay, now, etc. By so doing, a writer or speaker is able to lead the audience’s focus of understanding in perceiving the arguments. Unlike frame markers category, this type of markers was more likely to be found in the spoken context. Specifically, these words occur in the beginning of their sentence as pointed out in example 11.

Example 11

Okay. So, let’s refresh ourselves and listen to our explanation. [S]

Endophoric Markers (ENM)

Endophoric markers refer to the elements which are used to guide the reader/audience into other points somewhere within the text. The use of endophoric markers aims to point out the emergence of particular points that the writer or speaker needs to emphasize those points in the beginning. According to Hyland (2005), endophoric markers in soft disciplines refers to the actions of reinforcing the discourse to premeditatedly provide a quick access for the readers/audience to reach the relevant points between some parts of the text. In addition, these markers noticed the writer/speaker’s strategy to provide the following contributory arguments by referring to the earlier sections or statements of the text. Example 12 points out the use of endophoric markers in students’ argumentative essays.

Example 12

Regarding this matter, Indonesian government should require sex education since elementary school. [W]

Example 13

Back to the early explanation which the judge is also a human. [W]

Evidentials (EV)

This category holds as the third most used interactive metadiscourse in the written text. In contrast, evidentials did not even occur in any spoken data. In the data, evidential devices occur in different ways. First, the writer/speaker might lead into some claims by adding the words based on, according to, etc. (example 14) Secondly, they directly mention the owner of the statement (example 15). Thirdly, in the written context, the writer might announce the superiority of their arguments by writing convincible phrases mentioning prior studies (example 16). Besides, according to Khedri et al. (2013), the in-text citation (example 17) is categorized as evidential markers since it strengthens writer’s claims with the convincible frameworks in the literature. With the use of evidential devices, a writer or a speaker expected that the arguments they provided in their argumentative essays or public speech would be logical and acceptable as they try to provide information from other (relevant) text (Hyland, 2005). Unfortunately, the evidential devices were not found in the academic presentation due to, perhaps, the setting of the presentation which is more descriptive than argumentative.

Example 14

According to Britannica Dictionary, Gadjah Mada rose to power on his intelligence, courage, and loyalty to King Jayanagara during a rebellion led by Kuti in 1319. [W]

Example 15

Central Bureau of Statistics stated that the population of Indonesia in 2018 reached 256 millions, of which 133.17 million men and 131.88 million women. [W]

Example 16

Several studies in Indonesian hospitals show that the methods and pregnancy devices used are drugs, herbs, injections… [W]

Example 17

…, there is also legal behavior in response to social structures to achieve goals, be reminded that a person’s age will influence the actions that he will take (Cloward & Ohlin, 1960). [W]

Code-glosses (CG)

This category appears as the second most used marker in the written data, while it appears as the third most used marker in the spoken data with a slight number difference with the Frame Markers. The use of Code-glosses is intended to give the readers or the audience an optional meaning of referred information. Writers or speakers are supposed to provide additional information or prediction to encourage proper understanding in their audience or readers. Our data showed that a student is more likely to give additional preferences of meanings by connecting them with the use of or (example 18), which according to Hyland (2005) serves to help readers grasp the function. This indicates the writer or the speaker’s effort to guide the readers or audience to reach the conceptual understanding regarding the topic. Fortunately, the word or is not used as highly as the word ‘and’ and ‘so’ in the previous category. The measured data showed a slight distinction between ‘or’ and the other Code-glosses words.

Example 18

They demand their right to freely determine gender or sex orientation and the right to show their identity as LGBT people in the general public. [W]

Example 19

Social media, for example, has frighteningly stolen the life of Millennials. [W]

Moreover, the students often give an optional concept of understanding by guiding the readers or audiences to illustrate the example. The most used phrase in the written and spoken data is for example. Hyland (2017) categorized these words as exemplificatory markers with other words such as like and such as (example 20 and 21). Exemplificatory markers occur as a process of communication in which an example is attached to demonstrate the primer unit of a discourse (Hyland, 2017). In the spoken discourse, the most used exemplificatory marker is like. Nonetheless, unlike its use in written text, the word like tends to be used to lead the audience into a particular point of the material (example 22). In addition, other lexical elements like in fact, especially, this is called, etc. are also categorized as code-glosses (Hyland, 2005).

Example 20

They support any government’s censorship in many media such as TV, websites, news, etc. [W]

Example 21

…, it becomes one noun phrase like ‘the man who lives next door’. [S]

Example 22

The prohibition on illegal abortion among teenagers is a good rule because it has the purpose of banning bad actions outside marriage, especially for teenagers. [W]

Example 23

While in fact, online transportation is totally contrasted with this policy, they even use their personal vehicle as a transportation for their job. [W]

Interactional Metadiscourse Markers

This metadiscourse category refers to the strategies used by a writer/speaker to narrate his/her commitment. It consists of five categories: hedges, boosters, engagement markers, attitude markers and self-mentions. Atolaibi (2018) mentioned that the interactional metadiscourse markers represent the writer/speaker’s strategies of stance-taking. Therefore, using the interactional devices, the reader might be led and convinced to understand the ideas written/spoken in the text. Besides, the interactional devices allow the reader/audience to get involved in the discourse. The writer/speaker will try to emphasize their claims, connecting themselves with their audience, figuring each of the information out that the audience will focus and pay attention, convincing the audience and leading them to the understandings. In addition, the writer/speaker might use these devices to socially engage with the readers/audience (White, 2003, as cited in Hyland, 2005).

Hedges (H)

The first mostly used interactional metadiscourse markers in EFL students’ written and spoken discourse is the hedging device. Hedges are employed in a text to give an understanding toward the writer/speaker commitment within the text. This category leads the main category of written data, with 522 cases in the argumentative essays, while it is in the second place of spoken discourse with 45 cases in total. According to Salichah et al.(2015), hedges can occur in five categories. The first category is modal verbs (example 24). These modal devices represent the ability and possibility expressed towards the readers/audience. On the other hand, the second category of hedges is epistemic adjectives or adverbs. Similar with the modal verbs, epistemic adjective and adverb words reflect the uncertainty or probability of a particular claim (example 25 and 26).

Example 24

If I mention the name “Genghis Khan”, what would be the biggest thing that related to? I’m sure “China” will be the surest certainty answer. [W]

Example 25

…, but in that time their children likely have to complete their homework or take rest to prepare tomorrow’s learning process. [W]

Example 26

…, with the plurality of people in Indonesia, it is possible for customary law in each region. [W]

Another category is epistemic lexical verb which was also found in the present study (example 27). Specifically, the epistemic lexical verbs are distinguished in some sub-types. They show speculation of the author such as the word suggest, the act of subtraction with the use of conclude, infer, etc., or the author’s commitment by using the words claimed, showed, etc. They also include verbs of perception such as seem, probably, etc. and the elements of narrating such as attempt, tend, seek, etc. (Salichah et al., 2015). We found no narrating related-examples since data employed here were argumentative essays and academic presentations in which the students were not supposed to mention or explain their objectives as a researcher is supposed to do with their journal articles.

Example 27

For further development, I suggest Sandi to consider his idea and make it as clear as possible before announce it publicly. [W]

Example 28

It seems like those kinds of stuff are just as essential as oxygen for their life sustainability. [W]

Boosters (B)

In general, boosters are employed to convince the reader/audience by writing or uttering the certainty towards them. It provides the author/speaker a chance to express their claims and prevent the countered views on their perspectives (Hyland, 2017). In the written data, the booster elements were the second most-used category with 237 cases. Nonetheless, this category was not much found in the spoken data; it was in the third place with only 15 cases in total. The rank might be caused by a different genre of text. In the written data, the students wrote their arguments and aimed to convince the readers with their stance in their academic writing. Meanwhile, the tone of their academic presentation was informative and descriptive. Therefore, they might have no persuasive intention as they did in their academic writing.

There are several categories of boosters: universal and negative pronouns, amplifiers and emphatics. The universal pronouns (every, no, all, etc.) are used to hyperbolically narrate an overstated claim that might invite exaggerated impressions coming from the audience (example 29). Meanwhile, the amplifiers are used to exaggerate the effect of a claim and increase the lexical intensity of its. The amplifiers are mostly represented by adverbs such as very, certainly, extremely, etc. (example 30). Finally, the emphatic markers such as indeed, for sure, etc. refers to the elements which emphasize the claim of a writer/speaker in the text (example 32).

Example 29

Nobody could save Millennials, but they themselves. [W]

Example 30

However, basically, every child in this world was born with various potentials in each of their souls. [W]

Example 31

These derivations certainly will be a negative thing that the commercial sex workers do. [W]

Example 32

Indeed, education is a primary thing which has to be gotten by every child in our nation. [W]

Furthermore, Lee & Subtirelu (2015) mentioned that the word know is categorized as a booster. This word is used to emphasize the writer/speaker’s claim that it might encourage the readers/audience’s understanding. It is often following the pronouns like you, we, or I. Nonetheless, this word is more often found in the spoken data rather than the written data as in example 33.

Example 33

In the previous semester, we have learned about relative clause, right? Do you know what are those relative clauses? [S]

Attitude markers (AM)

The attitude markers refer to the devices that show the writer/speaker’s emotion. Therefore, these devices might show their acceptance, obligation, frustration, importance, etc. (Hyland, 2005). To highlight, the attitude markers are rarely used either in spoken or written discourse. It might appear in three main forms. First, it appears as an attitude verb such as agree, support, prefer (example 34). Secondly, it might be in the form of sentence adverbs such as hopefully, fortunately, etc. (example 35). Finally, some adjectives such as illogical, appropriate and remarkable are also classified as attitude markers (example 36).

Example 34

I disagree with the argument of the opposing side. [W]

Example 35

Unfortunately, all those brilliant superiorities do not balance with their morality. [W]

Example 36

It is appropriate to call Genghis Khan as one of the most feared conquerors of all time world history. [W]

Self-mentions (SM)

Self-mentions are often used with first-person pronouns (I, we) and possessive adjectives (mine, me, our, ourselves). A writer/speaker uses these devices to highlight their own presence in the text and represents their self-representations (Hyland, 2005: 53). In this case, there is a distinction between the spoken and the written data. In the spoken data, self-mentions are used more often with 1.32 frequency of occurrence in 60 seconds, making this category the second most used category of interactional markers. However, in the written data, the students do not seem to use self-mentions often; it ranks in the third place with 80 cases.

Example 37

The first and the second gonna be explained by me. [S]

Example 38

I think it is better to limit the number of private transportations instead of adding more transportations. [W]

Example 39

..., we may assume that the one of the most possible reasons toward the increase of traffic jam is caused by them. [W]

Engagement Markers (EGM)

A writer/speaker might use engagement markers to explicitly address the audience in their texts. The use of these devices is aimed to directly involve the audience to participate in the discourse and highlight their presence in the text. This category is mostly signaled by the use of second-person pronouns and possessive adjectives, including you, yours and yourselves (example 40). In addition, engagement markers might also be presented using interjections such as well, hello, by the way, etc. Additionally, Hyland (2009) provides another strategy of engagement by positioning the audience and pulling them into the discourse by giving imperative words (see, highlight, note, etc.) or modals such as should, must, have to, etc. which are directed to the audience (example 41).

Example 40

You will no longer be able to subsidize yourself. [W]

Example 41

Take a look to the example of relativised elements… First example, “some friends who saw her”. So, ‘who’ here functions as relative clause that modify “some friends”. Second example… [S]

Conclusions

We have examined the frequency distribution and use of metadiscourse markers in Indonesian EFL students’ academic writing and classroom presentations. Our study demonstrated that there is subtle difference of metadiscourse markers used by Indonesian EFL in speaking and writing which might be influenced by different factors such as setting and genre. Nonetheless, Transition Markers are consistently found to be concomitant with earlier existing studies denoting them as pivotal and frequent markers as to keep a text coherent (i.e., Hu & Cao 2015; Hyland, 2005; Park & Oh, 2018). Although the form that falls within this category slightly varies, which might confirm Jenkins’ (2014) notion of English as a Lingua Franca feature, Transition Markers in our study have showed their important role as providing readers or listeners with the most relevant (Sperber & Wilson, 1996).

The data also elucidate the higher occurrence of interactive metadiscourse markers. These findings resonate with Hyland and Jiang’s (2018) recent investigation of 2.2 million words of academic corpus which suggests that, over the past fifty years, there has been a significant increase in the interactive category. Another study by Gillaerts and Velde (2010) who examine metadiscourse markers in abstracts of research articles has also reported the changes of interactional type. The present study has therefore corroborated the dynamics of English metadiscourse, especially as used by speakers in the expanding circle (Kachru, 1992). Besides, it also has filled the gap of metadiscourse studies in writing that have been outnumbered as we examine and compare spoken and written discourse.  

EFL students’ use of metadiscourse markers remains a significant issue to address. It is important to provide students with proper understanding on how to use such markers for either spoken or written English purposes. It is also advisable to enrich students, non-native English speakers especially, with adequate understanding on the use of metadiscourse markers in academic settings, be it spoken or written since metadiscourse could help them improve the quality of the paper (Burrough-Boenisch, 2003; Flowerdew & Wang, 2016).

The present study showed that interactive and interactional metadiscourse marker (Hyland, 2005, 2017) frameworks can be tools to lead students to better understanding of metadiscourse markers use in academic settings. Regardless of the fact that in our study Indonesian EFL students tend to use interactive markers more frequently than interactional metadiscourse markers in both spoken and written text, we need to thoroughly examine how such usage is performed. We believe a corpus-based approach to the teaching of metadiscourse markers can be applied to provide students with the real-life use of those markers in spoken and written text. English as a lingua franca could also be employed to investigate the complexity, dynamics, and or hybridity of metadiscourse markers in the expanding circle to provide alternatives toward the existing model proposed by Hyland (2004, 2005) which was mostly based on English in the inner circle.  

References

Adel, A. (2010). Just to give you kind of a map of where we are going: A taxonomy of metadiscourse in spoken and written academic English. Nordic Journal of English Studies, 9(2), 69-97. http://ojs.ub.gu.se/ojs/index.php/njes/article/view/415/403

Aull, L. L., & Lancaster, Z. (2014). Linguistic markers of stance in early and advanced academic writing. Written Communication, 31(2), 151-183. https://doi.org/10.1177%2F0741088314527055

Aunurrahman, A., Hamied, F. A. H., & Emilia, E. (2017). Exploring the tertiary EFL students' academic writing competencies. Indonesian Journal of Applied Linguistics, 7(1), 72-79. https://doi.org/10.17509/ijal.v7i1.6860

Atolaibi, H. S. (2018). Metadiscourse in dissertation acknowledgments: Exploration of gender differences in EFL texts. Educational Sciences: Theory and Practice, 18(4). https://doi.org/10.12738/estp.2018.4.0247

Burrough-Boenish, J. (2003). Shapers of published NNS research articles. Journal of Second Language Writing, 12(3), 223-243. https://doi.org/10.1016/s1060-3743(03)00037-7

Correia, R., Eskenazi, M., & Mamede, N. (2015). Lexical level distribution of metadiscourse in spoken language. Proceedings of the First Workshop on Linking Computational Models of Lexical, Sentential and Discourse-level Semantics (LSDSem), 18 September, 2015, Lisbon, Portugal., 70-75. http://www.emnlp2015.org/proceedings/LSDSem/pdf/LSDSem08.pdf

Creswell, J. W. (2008). Educational research: Planning, conducting, and evaluating quantitative and qualitative research (3^rd ed). Pearson.

Farahani, M. V., & Sabetifard, M. (2017). Metadiscourse features in English news writing among English native and Iranian writers: A comparative corpus-based inquiry. Theory and Practice in Language Studies, 7(12). http://dx.doi.org/10.17507/tpls.0712.12

Flowerdew, J., & Wang, S. H. (2016). Author’s editor revisions to manuscripts published in international journals. Journal of Second Language Writing, 32, 39-52. https://doi.org/10.1016/j.jslw.2016.03.004

Gholami, J., & Ilghami, R. (2016). Metadiscourse markers in biological research articles and journal impact factor: Non-native writers vs. native writers. Biochemistry and Molecular Biology Education, 44(4), 349-360. https://doi.org/10.1002/bmb.20961

Gillaerts, P., & Van de Velde, F. (2010). Interactional metadiscourse in research article abstracts. Journal of English for Academic Purposes, 9(2), 128-139. https://doi.org/10.1016/j.jeap.2010.02.004

Graddol, D. (2006). English next (Vol. 62). British Council.

Gunawan, W., & Aziza, F. (2017). Theme and thematic progression of undergraduate thesis: Investigating meaning making in academic writing. Indonesian Journal of Applied Linguistics, 7(2), 413-424. https://doi.org/10.17509/ijal.v7i2.8350

Halliday, M. A. K., & Matthiessen, C. M. I. M. (2014). Halliday’s introduction to functional grammar (4^th ed.). Routledge.

Harris, Z. S. (1970). Linguistic transformations for information retrieval. In Z. S. Harris (Ed.). Papers in structural and transformational linguistics (pp. 458-471). Springer.

Hinkel, E. (2002). Second language writers' text: Linguistic and rhetorical features. Routledge.

Hu, G., & Cao, F. (2015). Disciplinary and paradigmatic influences on interactional metadiscourse in research articles. English for Specific Purposes, 39, 12-25. https://doi.org/10.1016/j.esp.2015.03.002

Hyland, K. (1998). Hedging in scientific research articles. J. Benjamins.

Hyland, K. (2004). Disciplinary discourses: Social interactions in academic writing. University of Michigan Press.

Hyland, K. (2005). Metadiscourse: Exploring Interaction in Writing. Continuum.

Hyland, K. (2009). Academic discourse: English in a global context. Continuum.

Hyland, K. (2017). Metadiscourse: What is it and where is it going? Journal of Pragmatics, 113, 16-29. https://doi.org/10.1016/j.pragma.2017.03.007

Hyland, K., & Tse, P. (2004). Metadiscourse in academic writing: A reappraisal. Applied Linguistics, 25(2), 156-177. https://doi.org/10.1093/applin/25.2.156

Hyland, K., & Jiang, F. (K.) (2018). "In this paper we suggest": Changing patterns of disciplinary metadiscourse. English for Specific Purposes, 51, 18-30. https://doi.org/10.1016/j.esp.2018.02.001

Jenkins, J. (2014). English as a lingua franca: Attitude and identity. Oxford University Press.

Johnson, B., & Christensen, L. (2004). Educational research: Quantitative, qualitative, and mixed approaches. Pearson.

Kachru, B. B. (1992). Teaching world Englishes. In B. B. Krachu (Ed.), The other tongue: English across cultures, (pp. 355-366). University of Illinois Press.

Khedri, M., Heng, C. S., & Ebrahimi, S. F. (2013). An exploration of interactive metadiscourse markers in academic research article abstracts in two disciplines. Discourse Studies, 15(3), 319-331. https://doi.org/10.1177%2F1461445613480588

Kim, L. C., & Lim, J. M.-H. (2013). Metadiscourse in English and Chinese research article introductions. Discourse Studies, 15(2), 129-146. https://doi.org/10.1177%2F1461445612471476

Kuhi, D., & Behnam, B. (2011). Generic variations and metadiscourse use in the writing of applied linguists: A comparative study and preliminary framework. Written Communication, 28(1), 97-141. https://doi.org/10.1177%2F0741088310387259

Lee, J. J., & Subtirelu, N. C. (2015). Metadiscourse in the classroom: A comparative analysis of EAP lessons and university lectures. English for Specific Purposes, 37, 52-62. https://doi.org/10.1016/j.esp.2014.06.005

Park, S., & Oh, S-Y. (2018). Korean EFL learners' metadiscourse use as an index of L2 writing proficiency. The SNU Journal of Education Research, 27(2), 65-89.

Phan, L. H. (2009). Strategic, passionate, but academic: Am I allowed in my writing? Journal of English for Academic Purposes, 8(2), 134-146. https://doi.org/10.1016/j.jeap.2008.09.003

Phan, L. H. (2018). Higher education, English, and the idea of ‘the West’: Globalizing and encountering a global south regional university. Discourse: Studies in the Cultural Politics of Education, 39(5), 782-797. https://doi.org/10.1080/01596306.2018.1448704

Phan, L. H., & Le, L. T. (2013). Living the tensions: Moral dilemmas in English language teaching. In T. Seddon, & J. Levin (Eds.). World yearbook of education 2013: Educators, professionalism and politics: Global transitions, national spaces and professional projects (pp. 220-235). Routledge.

Selvi, A. F. (2014). Myths and misconceptions about non-native English speakers in the TESOL (NNEST) movement. TESOL Journal, 5(3), 573-611. https://doi.org/10.1002/tesj.158

Sperber, D., & Wilson, D. (1996). Relevance: Communication and cognition (2^nd ed). Blackwell.

Sung, C. C. M. (2014). Accent and identity: Exploring the perceptions among bilingual speakers of English as a lingua franca in Hong Kong. International Journal of Bilingual Education and Bilingualism, 17(5), 544-557. https://doi.org/10.1080/13670050.2013.837861

Taka?, V. P. (2018). Sentence-initial interactive metadiscourse and text coherence in EFL argumentative essays. Lonaka Journal of Learning and Teaching,, 9(2), 13-29. https://journals.ub.bw/index.php/jolt/article/view/1361/860

Tan, H., & Eng, W. B. (2014). Metadiscourse use in the persuasive writing of Malaysian undergraduate students. English Language Teaching, 7(7), 26-39. https://doi.org/10.5539/elt.v7n7p26

Wahyudi, R. (2018). Situating English language teaching in Indonesia within a critical, global dialogue of theories: A case study of teaching Argumentative Writing and Cross-Cultural Understanding courses [Unpublished PhD dissertation]. Victoria University of Wellington. http://hdl.handle.net/10063/7609

Williams, J. M. (1982). Style: Ten lessons in clarity and grace. Longman

Yeganeh, M. T., & Ghoreyshi, S. M. (2015). Exploring gender differences in the use of discourse markers in Iranian academic research articles. Procedia - Social and Behavioral Sciences, 192, 684-689. https://doi.org/10.1016/j.sbspro.2015.06.104