Metadiscourse in Indonesian EFL Learners ’ Persuasive Texts : A Case Study at English Department

Writing competence facilitates the students’ success in their study. In order to produce a coherent text, a writer needs to employ sufficient metadiscourse markers. This study attempted to investigate the use of metadiscourse markers in 7 Indonesian EFL learners’ Persuasive writings. The study is descriptive, applying a quantitative method to support a qualitative analysis. The research results reveal that the occurrences of textual marker types in EFL learners’ Persuasive texts are overall closely similar to those at considered as standard proficient writing (extract from BAWE corpus), while those of interpersonal marker types are different from the standard proficient writing . Since the occurrences of interpersonal marker types are different from those at BAWE corpus in terms of hedges, boosters, engagement markers, the teacher needs to give more practice in teaching the interpersonal metadiscourse markers. Metadiscourse analysis can be a strategy to make the students aware of the important role of metadiscourse markers.


Introduction
Writing activities play a crucial role in college students' learning process.They start writing from the early up to the late stage, e.g., writing summary, research proposal, research report, etc.That is why, Geiser and Studly (2001) in McNamora, Crossley, and McCarthy (2010, p. 58) state that for students, writing ability is among the best predictors for success in course work during their years of study.Rao (2007) explains that EFL writing is useful in two respects, i.e., it encourages students' thinking, organizing ideas, developing their abilities to summarize, analyze, and criticize.And it strengthens their learning and thinking.From the statements above, it is clear that writing is essential for academic survival.Schleppegrel (2004, pp. 88-89) states that as students move to high school and beyond, they are needed to write Persuasive writings, a genre through which they present a point of view and support it with arguments.The Persuasive writing is symbolic of students' success with language at school where they are expected to provide reasoned, concrete arguments.Syntactic complexity is greater in Persuasive writing than in other genres.When students are writing Persuasive or Argumentative discourse, they are engaging in an activity which inherently requires logical interrelationships among propositions.Consequently, the complex skill in using metadiscourse markers is badly needed.Intaparawat and Steffensen's research "Comparison of metadiscourse use between good and poor ESL undergraduate writers" (1995) as cited by Heng and Tan (2010) discloses that more metadiscourse features are applied in qualified essays.Thus, genre of Persuasive writing and metadiscourse are two things that college students need to master in order to produce qualified writing.
Metadiscourse is a range of devices that writers use to explicitly organize their texts, engage readers, and signal their attitudes to both their material and their audience (Hyland & Tse, 2004, p. 156).It is the set of linguistic resources that every language has as part of the textual metafunction for linking one part of a text to another (Olateju, 2006).So, it is like glue that holds sentences and paragraphs together.It is textual and interpersonal in nature.Vande Kopple (2002) states that metadiscourse satisfies the interpersonal and textual function of language.
Based on the background above, this study attempted to investigate metadiscourse markers in Indonesian EFL learners' Persuasive texts.The problems clarified in this study are as follows: What are textual metadiscourse markers applied by EFL learners?What are interpersonal metadiscourse markers applied by EFL learners?What metadiscourse markers are dominantly applied?How are the occurrences of metadiscourse markers in EFL learners' Persuasive texts compared with those in BAWE corpus?
The studies on metadiscourse markers have been conducted by several researchers.Hyland and Tse (2004) investigated metadiscourse markers in Ph.D. dissertations and Master's theses.The research results reveal that one discourse marker is used in every 21 words.The metadiscourse markers used in Master's theses are balanced overall between the textual and interpersonal types, while in the doctoral texts the number of textual metadiscourse markers is ten per cent more than that of interpersonal metadiscourse markers.Noorian and Biria (2010) did a comparative study on interpersonal metadiscourse markers between American and Iranian Persuasive journalism texts.The research results reveal that interpersonal metadiscourse markers are employed far more frequently by the American writers and the difference showed is statistically significant.American writers employ 161 interpersonal metadiscourse markers per 1000 words, while Iranian writers employ 88 interpersonal metadiscourse markers per 1000 words.
Heng and Tan (2010) compared the metadiscourse markers use between the writing samples of Malaysian undergraduates at tertiary level and that in standard proficient writings (extract from British Academic Written Essays-BAWE corpus).Heng and Tan obtained the BAWE corpus, as the benchmark, from the text corpora available on the internet.BAWE corpus is claimed to be the standard for student academic writing and it was constructed as a research reference for the researchers to use.Their research reveals that BAWE corpus has higher frequency of use in textual metadiscourse (379 occurrences per 10,000 words compared to 242 occurrences per 10,000 words in interpersonal metadiscourse).It means that the metadiscourse markers use is 621 per 10,000 words, meaning that 1.24 metadiscourse marker occur per 20 words.It also means that the metadiscourse markers comprise 39 percent of interpersonal types and 61 percent of textual types.Since my research investigated college students' Persuasive writing, I used this finding as reference to analyze my research finding.Heng and Tan also reveal that Malaysian undergraduate corpus exhibited a higher frequency of interpersonal metadiscourse markers when compared to textual metadiscourse markers (354.2 occurrences per 10000 words compared to 319.3 per 10000 words).Since Malay and Indonesian are of the same language family, replication of similar study is necessary.
Referring to the research problems formulated above, this study is aimed at listing the types of textual and interpersonal metadiscourse markers, searching the dominant types, and describing the EFL learners' Persuasive texts in reference to the BAWE corpus.
The current study is expected to give practical contributions that may provide Indonesian teaching experts with beneficial information about the textual and interpersonal metadiscourse markers employed by the Indonesian students in their Persuasive texts.This information may be taken into consideration for developing English teaching materials to Indonesian learners, especially for teaching writing, and simultaneously may inspire the teacher concerning how to teach metadiscourse material.

Review of Related Literature
The theoretical framework employed in this study comprises Metadiscourse, Textual Metadiscourse Markers, Interpersonal Metadiscourse Markers, and Persuasive Discourse.

Metadiscourse
Written language is like oral language in the sense that both of them reflect three kinds of meanings: ideational, textual, interpersonal.Noorian and Biria (2010) explain that the theoretical basis for the term 'metadiscourse' derives from Halliday's classification of language macro functions.That is, when we use language we nearly always try to give expression to our experience, to interact with our audience, and to organize our expression into cohesive and coherent text.
Similar to Halliday's idea, Sinclair (1993) and Hyland (1985) as cited by Moreno (2003, p. 113) explain that in communicating meaning there are two basic components, namely that involved in creating meaning and that involved in sharing meaning.In communicating, on one level we expand propositional content while on the other level (metadiscourse level) we do not add propositional material but help our readers organize, classify, interpret, and evaluate the material.Thus, metadiscourse markers have two main functions: textual and interpersonal.And these metadiscourse markers are of the same importance as the propositional content.Hunston and Thompson (2001) explain that textual metadiscourse markers help to organize the discourse by pointing out topic shifts, signaling sequences, cross referencing, connecting ideas, etc., while interpersonal metadiscourse markers highlight the writer's attitude or appraisal.The category of textual and interpersonal metadiscourse markers based on Hyland and Tse (2004) will be clarified below.
Transitions are logical connectors that express semantic relation between main clauses or sentences.They are realized mainly with conjunctions used to mark additive, contrastive, consequential, temporal.Examples of transitions are in addition/but/thus/and/therefore/however/as a consequence/still.
Frame markers refer to discourse act (announcing discourse goal), sequences, text stages, topic shift.Examples of frame markers are finally/firstly/secondly/to conclude/my purpose here is to/to give an example/to sum up/I state again that/My question is/What I am emphasizing is/now let's turn to.Evidentials refer to source of information from other texts, providing support for the speaker's/writer's arguments.These metadiscourse markers are considered as important persuasive tools in the genre of opinion since they provide support and justification for the writer's arguments.Examples of evidentials are according to X/(Y, 2000)/Z states.
Code glosses signal the restatement of ideational information in other ways, providing additional information or examples for words or propositions that a text producer predicts the reader may find problematic.Thus, code glosses explain or expand propositional information to assist the interpretation and ensure the writer's intention is understood.Examples of code glosses are namely/e.g./such as/in other words/by this I mean/X meansY/that is/in other words.
Hedges mark the writer's reluctance to propositional information.Hedging shows the degree of tentativeness, possibility and/or politeness that the writers use in their texts.The function of hedges is to tone down assertions.Camiciottoli (2003) argues that hedges mitigate the writer's authorial position which makes the text more friendly.Noorian and Biria (2010) state that the ability to hedge effectively and successfully is a rather difficult skill, especially for EFL students, and need to be considered seriously by both teachers and students.Hedges are realized in epistemic verbs, probability adverbs, epistemic expressions.Examples of hedges are might/possible/perhaps/about/it is likely/maybe.
Boosters imply certainty and emphasize the force of proposition, expressing full commitment to the truth value of proposition.The function of boosters is to increase the force of assertions.Examples of boosters are in fact/definitely/it is clear that/certainly/of course/obviously/I know/clearly.
Attitude markers express writer's attitude to propositional content, conveying surprise, obligation, agreement, importance.Thus, attitude markers indicate the writer's affective.Attitude markers are realized in deontic verbs (should, have to), attitudinal adverbs, adjectival constructions, and cognitive/mental verbs.Examples of attitude markers are unfortunately/I agree/surprisingly/undoubtedly/most importantly/I hope/remarkably/I think/I believe/I feel/My opinion is.
Engagement markers build relationship with the readers by explicitly address the readers.They are realized through second person pronoun, imperative, question forms.Examples of engagement markers are consider/note that/you can see that/you may think/you may ask/you may not agree.
Self mentions reflect the degree of author presence in terms of the incidence of first person pronouns and possessives.Examples of self mentions are I/we/my/our.Kinneavy (1980) in Takagi (2009, p. 58) classifies discourse into four categories based on its objective, i.e., reference, literary, expressive, and persuasive which is the feature of Argumentative discourse.If the aim of a discourse is to reproduce reality, it is called reference.If the central attention of the discourse is the reader (decoder) and the discourse aims to move him or her into action, it is called persuasive.If the language (signal) is the main focus, the discourse is categorized as literary.And if the central attention is the writer (encoder) and his or her self expression is the most important, the discourse is considered expressive.A discourse can have more than one aim.

Persuasive or Argumentative Discourse
Making persuasion is a regular feature of our daily lives.The need for Persuasive discourse arises from our desire to persuade or convince others of a point of view or course of action.There are several genres that are persuasive, e.g., Analytical Exposition, Hortatory Exposition, Argumentative Discourse, Discussion.Whatever the name, the discourse is analytical discourse, i.e., the one that analyzes, argues, and persuades.This study focused on investigating EFL learners' Discussion texts.Gerot and Wignell (1995, pp. 197, 209) classify Exposition into Analytical Exposition and Hortatory Exposition.Hortatory Exposition differs from Analytical Exposition in that the latter argues that X is the case.Hortatory Exposition argues that X ought or ought not to be or should or should not be the case.Richardson (2007, p. 150) claims that Argumentative Discourse is designed to convince readers of the acceptability of a point of view and to provoke them into an immediate or future course of action by presenting logical, interrelationship of propositions or arguments/reasons.Thus, the key words of Argumentative Discourse are argument and persuasion.
Discussion text is a text that presents the points for and against the topic of the text.The purpose of Discussion text is to present to the audience different opinions on a topic and, the end, the writer's opinion.The schematic structure of Discussion text is as follows: 1) A general statement to introduce the reader or listener to the topic of discussion… 2) A series of paragraphs that have points for and against the topic-the text may have paragraphs on the for side followed by paragraphs on the against side, and within the paragraphs there should be evidence to support the point of view.
3) A concluding paragraph that sums up the discussion and gives the opinion of the author of the text.(Anderson & Anderson, 2003, pp. 120-121) Therefore, in order to construct a discussion text an analysis of pros and contras is needed in order to appropriately conclude.
Some language scholars name the stages of schematic structure of Persuasive or Argumentative Discourse differently.The schematic structure recommended by a number of scholars in Philosophy is Toulmin's Model because it has more stages and the name of each stage is based on its function in the Persuasive or Argumentative Discourse.
According to Toulmin, Rieke, and Janik, the schematic structure of Persuasive or Argumentative Discourse comprises 6 stages as follows: Data, Claim, Warrant, Backing, Qualifier, Reservation or Rebuttal (1984).
Data are some facts or observation about situation under discussion.They answer the question, "What information do you have to go on to reach your conclusion?"A given element is understood as datum if it supports a Claim, i.e., it must be specific, related to the Claim for persuasive results.Thus, it functions as the ground for the Claim.Inch, Barbara and Danielle (2006, p. 10) state that data/evidence contains proposition that helps the arguer focus the direction of the Persuasive or Argumentative Discourse.Proposition functions like thesis in an essay.They further explain that a propositional arena is the ground for dispute that includes all the issues for controversy.
A Claim is the end point of a Persuasive or Argumentative Discourse that the arguer wants accepted, that Toulmin calls as a conclusion.A Claim in an essay is considered strong if it is relevant to the task, suggests a specific and clear problem, and presents a consistent point of view.
The relation between Data and Claim is characterized by a rule of inference, i.e., a Warrant that serves to link the information set forth in the Data and Claim.Warrant expresses the reasoning used to link the Data to the Claim.Thus, Warrant is reasoning or rational link between the evidence and the Claim in the Persuasive or Argumentative Discourse.Inch et al. (2006, p. 164) state that when we reason, we make connections, distinctions, and predictions; we use what is known or familiar to reach a conclusion about which is unknown or unfamiliar.Warrant plays an important role in Persuasive Discourse because the logical dimension of an Argumentative Discourse mainly lies in the links or connections the arguer makes in reaching a conclusion.In other words, Warrant answers the question, "How do you reach the conclusion, based on the data/evidence?" Backing of a Persuasive Discourse is some knowledge structure from which an arguer derives the Warrant.So, Backing is data/evidence used to support the Warrant.It is the foundation of arguments.Inch et al. (2006, p. 164) clarifies that the sources of Backings vary depending upon the domain.For example, one common type of Backing in law is the set of relevant previous cases or pieces of legislation.In science, one may back a Persuasive Discourse by reference to theory, by reference to experimental practice, etc. Toulmin (1958) terms Backing as unassailable facts.
A Persuasive Discourse often contains Qualifiers, i.e., phrases or terms that signal our stance toward our Claim or our degree of certainty.Toulmin explains that when arguers make Claims, they attribute greater or lesser degrees of strength to them.The words "probably, certainly, possibly, strongly" are examples of qualifiers.They function to indicate the degree of certainty of how arguers feel regarding their Claims.
Reservation states the circumstances or conditions that undermine the Persuasive Discourse.It is the exception of the rule expressed in the Warrant.Real Persuasive Discourse, unlike the analytic Persuasive Discourse of standard logic allows for exceptions.So, this stage is optional in Persuasive Discourse.
According to Toulmin (1958) of these six parts, three are the most important and must appear in Persuasive or Argumentative Discourse, i.e., Data, Claim, and Warrant (D-C-W) which constitute the inferential core of a Persuasive or Argumentative Discourse.

Method
This study is descriptive in nature.And to support the description, simple quantification was done.Firstly, the metadiscourse markers of Indonesian EFL learners' Persuasive texts were identified, then they were analyzed, and interpreted.
The data were collected by asking 14 students to write Discussion texts.The researcher provided two topics and then each of the students chose one to be written for maximally 90 minutes.The topics provided were: a) The good and bad effects of watching TV, b) There are two arguments whether students should be given a lot of homework or not.From the 14 EFL learners' Persuasive texts, 7 of which were taken as the data of the study.
After the data had been collected, they were analyzed using the following steps: identifying, classifying, and interpreting.Firstly, the researcher identified the metadiscourse markers of EFL learners' Persuasive texts.Secondly, she classified the metadiscourse markers into textual and interpersonal.The textual metadiscourse markers were further classified into transitions, evidentials, code glasses, frame markers, endophorics; while the interpersonal metadiscourse markers were further classified into hedges, engagement markers, booster, attitude markers, self mentions.This classification is based on Hyland and Tse (2004).Thirdly, she found out the textual and interpersonal metadiscourse markers dominantly employed by the learners.Then, they were interpreted.
In analyzing the data, the researcher also compared the research findings with the metadiscourse used in standard proficient student writings (extract from British Academic Written Essays-BAWE corpus) revealed by Heng's and Tan's study (2010).The similarities and the differences between the metadiscourse markers used in EFL learners' Persuasive texts and those in BAWE corpus were revealed.This was aimed at knowing the position of Indonesian EFL Persuasive texts in reference to the native students' Persuasive texts.

Results
Based on the data analysis, some findings are put forward.There are 153 metadiscourse markers in 1617 words of 7 EFL learners' Persuasive texts, consisting of 89 textual markers and 64 interpersonal markers.Thus, the average use of metadiscourse markers in the EFL learners' Persuasive texts is 1617 words: 153 metadiscourse markers = 10.57.It means that one marker is used in every 10.57words.
The average use of interpersonal metadiscourse markers is 1617 words: 64 metadiscourse markers = 25.27,meaning that one interpersonal marker is used in every 25.27 words.While the average use of textual metadiscourse markers is 1617 words: 89 metadiscourse markers = 18.17, meaning that one textual marker is used in every 18.17 words.
The comparison between textual and interpersonal metadiscourse markers employed by the EFL learners is 89: 64, meaning that the use of textual metadiscourse markers is (89:153) X 100% = 58.17%,while the use of interpersonal metadiscourse markers is (64:153) X 100% = 41.83%.Thus, the use of textual metadiscourse markers is 16.34% higher than that of interpersonal metadiscourse markers that will be discussed in the following section.

Discussion
In the EFL learners' Persuasive texts, transitions are the most numerous textual markers (65 or 73%), followed by code glosses (11 or 12%), evidential markers (9 or 10%), and finally frame markers (4 or 4%), while none of the learners employ endophoric markers.From 65 or 73% of the transitions, causal markers dominate the number, i.e., 32 or 49%, followed by additive markers 15 or 23%, adversative markers 13 or 20%, and finally temporal markers 5 or 8%.This finding is as what is expected by Schleppegrel (2004, p. 88), Macken (1991, p. 5) that syntactic complexity is greater in Persuasive than in Narrative or Descriptive writing.When individuals engage in Persuasive or Argumentative Discourse, they are engaging in an activity which inherently requires the logical interrelationship of propositions.Persuasive as a factual text used to put forward a point of view involves logical rather than temporal sequencing.The highest frequency of transition markers indicates that the writers attach prime importance to the organization of the texts, to make the connection between different bits of information in the text.
Evidential markers considered as important persuasive tools in the genre of opinion appear in the third position of textual markers.They refer to source of information from other texts, providing support for the speaker's/writer's arguments.These markers are important in a situation in which objectivity is essential, in which the source of every proposition needs to be determined.Statements with evidential markers usually appear in Backing stage of Persuasive Writing.Toulmin (1958) terms Backing as unassailable facts.Thus, it is a good phenomenon that these markers appear in EFL learners' Persuasive texts.However, from 7 texts as the data of the research, 3 of which do not contain evidential markers.Therefore, these markers need to be emphasized in teaching-learning processes.
The students are skillful in using markers of code glosses which are 12% of the textual markers.The code glosses provide additional information or examples for words or propositions that the writer predicts the reader may find problematic.All of the data in this research employ markers of code glosses such as e.g., that is, for example, such as.
Frame markers are used only by 2 learner writers in the forms of illocutionary markers and sequencing such as in short, first, the second, whereas the other 5 writers do not employ frame markers.The lack of frame markers usage may cause in the unsmooth topic shift.Therefore, teachers should make the students aware concerning this impact.
None of the learner writers employ endophoric markers.This is perhaps because their texts are short enough from 131 up to 316 words.Endophoric markers are reminders referring to the information in the other parts of the text.Since their texts are short, reminders are perhaps not so needed.
The number of occurrences of textual metadidiscourse category in EFL learners' Persuasive texts can be summarized as follows.The frequency use of transition, code glosses, evidentials, frame markers, and endophoric markers is sequentially 73%, 12%, 10%, 4%, zero.These occurrences are similar to those at standard proficient writing (extract from BAWE corpus) revealed in Heng's and Tan's study.In BAWE corpus, transitions dominates the occurrences, i.e., 63.8%, code glosses number of occurrences is 11.3%, evidential markers' occurrences are 17.7%, frame markers' occurrences are 3.1%.The difference is that endophoric markers do not occur in the data of EFL learners' Persuasive texts, while in BAWE corpus, the endophoric markers have 4.1% occurrences.
The length of EFL learners' Persuasive texts and that of BAWE corpus is also similar, i.e., less than 500 words.The length of EFL learners' Persuasive texts is 131 up to 316 words, while that of BAWE corpus is 145 up to 425 words.
In EFL learners' Persuasive texts, hedges and boosters occur 9% and 6% respectively.According to Heng and Tan (2010, p. 131), the careful balance of the use of both hedges and boosters is important as they reflect the writers' ability to balance a show of their confidence with caution.Dafouz (2008) confirmed the key role of hedges in Persuasive texts where the author needs to balance between commitment to his/her ideas and respect and dialogue with the reader.Hedging shows the degree of tentativeness, possibility and/or politeness that the writers use in their texts.Camiciottoli (2003) argues that hedges mitigate the writer's authorial position which makes the text more reader friendly.Heng's and Tan's study (2010) reveals that hedges are the most numerous interpersonal markers in BAWE corpus (47.7%).Noorian's and Biria's study (2010) discloses similar result.However, in EFL learners' Persuasive texts only 9% of the interpersonal markers are found.This minimum use of hedges strengthens Noorian's and Biria's statement (2010) that the ability to hedge effectively and successfully is a rather difficult skill for the students.
In BAWE corpus, the occurrences of boosters which are also called as certainty or emphatic markers are much smaller than those of hedges, i.e., 20.3%.While in EFL learners' Persuasive texts, the occurrences of boosters such as actually, really, of course are only 6%.Boosters allow the readers to find out about the writer's opinion.The less use of boosters indicates that EFL learners are perhaps less assertive in their Persuasive texts.
EFL learner writers attempt to engage their readers by addressing the audience, using rethorical question, using imperative form.It is useful to involve the readers in the process of argumentation.However, in EFL learners' Persuasive texts, engagement markers have the lowest frequency, i.e., 5% while in BAWE corpus, the occurrence of engagement markers reaches 18.9%.From the data of the research, less than 50% of the learner writers employ engagement markers.This indicates that EFL learner writers are less skillful in engaging the readers.
Heng's and Tan's study (2010) reveals that the lowest use of interpersonal metadidiscourse markers in BAWE corpus is self mention, i.e., 5.4%.On the contrary, markers of self mention dominate EFL learners' Persuasive texts.The high frequency of self mention markers signals the personal presence of the learner writers in the texts, meaning that they prefer adopting personal style.This finding coincides with Mauranen's (1993) research as quoted by Noorian and Biria (2010) that Anglo-American writers frequently signal their personal presence in academic texts.It seems that personal markers and inclusive expressions play an important role in American opinion articles since they allow writers to express their opinion in a more personal way and help the reader find out about the writer's stance.Dafouz' study (2008) pointing out that these markers are persuasive tool in the eyes of the reader.The use of attitudinal markers is a key strategy to convey the writer's affective values towards the propositional content, to express the writer's opinions and feelings in a personal way.However, the occurrences of attitude markers in EFL learners' Persuasive texts are much more than those in BAWE corpus as revealed in Heng's and Tan's study (2010).

Conclusions
From the data analysis, some conclusions can be drawn.The research results reveal that 1 discourse marker is used in 10.57 words or 1.88 discourse marker in 20 words.It is a bit higher than that in BAWE corpus as revealed by Heng's and Tan's study, i.e., 1.24 per 20 words.The average use of interpersonal metadiscourse markers is 1 marker per 25.27 words, while that of textual metadiscourse markers is 1 marker per 18.17 words.Thus, the comparison between textual and interpersonal metadiscourse markers employed by the EFL learners is 58.17% and 41.83%, meaning that the use of textual metadiscourse markers is 16.34% higher than that of interpersonal metadiscourse markers.This finding is similar to that in BAWE corpus, i.e., the use of textual metadiscourse markers is 22% higher than that of interpersonal metadiscourse markers.
The occurrences of textual marker types in EFL learners' Persuasive texts are overall closely similar to those at standard proficient writing (extract from BAWE corpus), except in the endophoric markers, i.e., the endophoric markers do not occur in the data of EFL learners' Persuasive texts, while in BAWE corpus, the endophoric markers have 4.1% occurrences.
The occurrences of interpersonal marker types are different from those at BAWE corpus.The differences are EFL learners' Persuasive texts employ minimum use of hedges, boosters, engagement markers.

Suggestions
Based on the conclusions concerning the use of metadiscourse markers by EFL learners, some suggestions are put forward in the following.
Forty-one percent of the data do not employ evidential markers, thus, these markers need to be emphasized in the teaching-learning process.The students need to be reminded that evidential markers are considered as important persuasive tool in the genre of opinion because they strengthen the writer's arguments.The teacher should remind the students that the lack of frame markers usage may cause in the unsmooth topic shift because only twenty-nine percent of the learner writers employ frame markers.None of the learner writers employ endophoric markers.Thus, the teachers need to remind them that endophoric markers are usually needed in longer texts as reminders referring to the information in the other parts of the text.
Since the occurrences of interpersonal marker types are different from those at BAWE corpus (considered as standard proficient learner writing), the teacher needs to give more practice and time allotment in teaching interpersonal metadiscourse markers.It is suggested that English teacher inform the students concerning the importance of interpersonal metadiscourse in Persuasive writing to provide support and justification for the arguments.The teacher should remind the students that the careful balance of the use of both hedges and boosters is important as they reflect the writers' ability to balance a show of their confidence with caution.The teacher needs to remind them that it is useful to involve the readers in the process of argumentation, thus, they need to use more engagement markers because less than 50% of the learner writers employ engagement markers.
In short, the findings of the study might have pedagogical significances, especially for L2 writing teaching for Indonesian EFL learners because metadiscourse markers are indispensable from writing.And according to Crismore, Markkanen, Steffensen (1993), metadiscourse is area that is difficult to gain by foreign language learners.The research results may give information for the students concerning the differences between the metadiscourse markers employed in Indonesian EFL learners' persuasive writings and that employed in native learners' persuasive writings reflected in BAWE corpus.
It is also suggested that the students be informed the way native writers organize their writings using metadiscourse markers.Thus, metadiscourse markers should become one of the writing teaching materials.Metadiscourse analysis can be a strategy to make the students aware of the important role of metadiscourse markers.Teachers can provide authentic texts for their students and ask them to identify and count the metadiscourse markers and discuss them in groups.This may make the students familiar with metadiscourse markers and hopefully they will apply the metadiscourse markers in their writings.
Endophoric markers are reminders referring to the information in the other parts of the text.Examples of endophoric markers are noted above/see Fig./in section 2/as I mentioned before.
Based on the data analysis, EFL learner writers employ 22 (34%) attitude markers comprising deontic verbs (should, have to, must), cognitive verbs (I feel, I think), adjectival constructions (It's very fun, it is very interesting, It is common, It is wrong, It is true, it is necessary), attitudinal adverbs (in my opinion).The presence of attitude markers coincides with