1

CHAPTER I

INTRODUCTION

1.1 Research Background

Discourse analysis is an analysis of language which focuses on investigation

about how to use language correctly (Brown & Yule, 1983; Edwards, 2008;

Hyland, 2005). In terms of the investigation, a language can be divided into two

kinds, written language and spoken language. They are different in how to

produce the language. In discourse analysis, there are many fields of study. This

study focuses on genre analysis of text which can be analyzed based on the

rhetorical features. Metadiscourse as the subject of the study is known as one of

the rhetorical features (Swales, 1990).

In the past, some studies have been conducted in the development of

metadiscourse features of research articles (RAs) (Del Saz-Rubio, 2011; Hyland,

2005a; Loi & Lim, 2013; Mur-Dueñas, 2011; Peterlin, 2005). RAs as the

academic writing should be effectively understandable so as to deliver the idea

and the aim of the texts. Therefore, the use of metadiscourse is important as the

support in understanding the text. Metadiscourse is known as the reflective

language used to interact between readers and writers or speakers and listeners. It

depends on how the users display the metadiscourse itself. Especially in writing

text, it can be used in expressing the important meanings correctly, in organization

text, and interaction between them for understanding the text (Fa-gen, 2012). The

types of metadiscourse has been revised time to time. Early study as Crismore

1
 2

(1984) differs it into two parts: “informational” and “attitudinal”. The division is

based on the signal presence of author and the signal attitude of author. Recently,

Hyland (2005) argues that metadiscourse become two general categories

“interactive” and “interactional.” It divides it based on the text function. Firstly, it

includes such sub- classifications as: “transition markers,” “frame markers,”

“endophoric markers,” “evidential,” and “code glosses” and secondly, it includes

such sub- classifications as: “hedges,” “boosters,” “engagement markers,”

“attitude markers,” and “self-mentions.” Specifically, metadiscourse has many

variant usages in languages and disciplines difference in RAs. For instance, Mu,

Zhang, Ehrich, & Hong (2015) find that Chinese and English applied the different

attention in using metadiscourse based on Hyland’s model in RAs. The English

RAs used more interactional metadiscourse than Chinese RAs. The result

demonstrates that the English pays more attention to the interaction between the

writers and their readers. In addition, it shows the different ways of metadiscourse

features used between both of them. Thus, it is likely that the metadiscourse

features understanding in academic article will be advantageous in discovering

knowledge among languages and cultures.

Writing is a skill which plays an important role in EFL learning. Although

students can write, but their mastery in academic writing in some cases is not as

good as their mastery in general English writing. They may have problems in

producing cohesive texts or communicating their thoughts with their readers.

Since the writers are not familiar with concepts of cohesion and cohesiveness,

they may be unable to create a cohesive and coherent text. According to Hyland
 3

(1998), metadiscourse markers can be used as one of the important rhetorical tools

in producing texts and persuading writers. Familiarity with metadiscourse markers

can in part solve this problem. Although metadiscourse is not a new concept, it is

increasingly important in writing and reading research articles and also it is an

essential issue in listening and speaking.

This study also focused on Hyland’s metadiscourse model. Mu, Zhang,

Ehrich, & Hong (2015) state to investigate the use of metadiscourse in a genre-

based approach study it is important to determine the suitable metadiscourse

features in discipline of applied linguistics. Furthermore, it states Hyland's model

is a genre based and has been existed from other metadiscourse studies. In

addition, it claims the previous taxonomy had been built by Hyland’s model, then

it arranged the model of metadisourse more accurately. Also, it is the latest

metadiscourse model which is clear, simple, and inclusive (Abdi, Rizi, &

Tavakoli, 2010). More importantly Hyland’s model had been applied in such of

studies (Abdi et al., 2010; Del Saz-Rubio, 2011; Lee & Subtirelu, 2015; Loi &

Lim, 2013; McGrath & Kuteeva, 2012; Mu et al., 2015). Therefore, as reasons

mentioned above, it is clear that metadiscourse of Hyland’s model is suitable to

apply in this study.

This study investigated metadiscourse features among cross- linguistic in

RAs. To date, there are numerous studies identified that English compare in other

languages in using metadiscourse such as: Chinese (Mu et al., 2015), Turkish

(Ozdemir & Longo, 2014), Iranian (Gholami, Tajalli, & Shokrpour, 2014),

Spanish (Moreno, 1997; Mur-Dueñas, 2011; Vergaro, 2004), Slovene (Peterlin,

 4

2005), Persian (Zarei & Mansoori, 2011), and Spanish and Norwegian (Dahl,

2004). These studies indicate that it has been common to compare the use of

metadiscourse of English with other languages. However, to the best of the

researcher knowledge, comparing English native writers with Indonesian writers

in term of their use metadiscourse is hardly ever done. This fact refers as the gap

for the current research. Considering this gap, this paper attempts to compare the

use of metadiscourse by English native and Indonesian writers in their English

abstract of RAs. Moreover, this study investigates how English native and

Indonesian writers construct their metadiscourse choices through their knowledge

in their writings.

1.2 Research Problems

Based on the background of the study above, the problems of this study are

stated as follows:

1. What type of metadiscourse are dominantly used in local, national and

international journal article on english education?

2. What are the differences and similarities in the use of metadiscourse between

local, national and international journal article on english education?

1.3 Research Objective

According to research problem above, the objectives of research are as

follow:
 5

1. To identify What type of metadiscourse are dominantly used in local, national

and international journal article on english education

2. To identify What are the differences and similarities in the use of

metadiscourse between local, national and international journal article on

english education

1.4 Scope and Limitation

As noted above, discourse analysis has many fields of studies to conduct.

However, this study focuses on one of the crucial part of discourse analysis,

namely metadiscourse. Metadiscourse has been developed time to time. It has

many types across experts. Previously, early study categorizes metadiscourse into:

“informational” and “attitudinal”. This study focuses on the latest metadiscourse

categories based on Hyland’s model (i.e. interactive and interactional

metadiscourse).

This paper aims to investigate and to identify the use of metadiscourse

between Indonesian and native English writers in their RAs. Although there are

many parts of RAs, this paper specifically focuses on the abstract of RAs. The

data utilized in this study is gathered from different texts based on local and

international journal articles in English language. Furthermore, the journals only

focus on the field of TESOL and applied linguistic.

 6

1.5 Research Significance

1. Discourse analysis teachers

As metadiscourse and corpus are acknowledged as two main parts of

discourse analysis, this paper can be one of references to support the discourse

learning. Hence, it is expected that this study helps the teacher to introduce

and to teach the importance of metadiscourse role in the discourse learning

itself.

2. English teachers

As English teachers, there are four main skills that they should teach to

their students comprehensively such as reading, speaking, listening, and

writing. Different skills need to have different way of how to teach them,

especially in writing skill. When students learn how to write, the coherence is

one of important things to make good writings. Because metadiscourse has the

function to keep the interaction between readers and writers in understanding

the text, teachers can introduce how to use metadiscourse in writings.

In order to make the metadiscourse easy to be understood by the

students, teachers should be aware of the way how to teach the metadiscourse.

They should well understand the differences of definition through

metadiscourse features and how they are applied in the text.

3. Students

In writing class, every student will produce their writings. In order to

make their writings good and coherent, they have to know how to make

relation between them, as the writer and their readers. The use of
 7

metadiscourse is essential to increase the understanding of the text. Therefore,

to make a good text based on correct use of metadiscourse, the students have

to understand the different function of metadiscourse features and how they

would be applied in the text.

4. Research in the field

Metadiscourse is the completely a new concept in the text analysis area.

Despite the importance of metadiscourse several studies have been

investigated in different angles recently, it is commonly done the investigation

about the use of metadiscourse between two different languages. Surprisingly,

the metadiscourse investigation for Indonesian is hardly ever done. In this

perspective, the future researcher will consider this gap for the study to

contribute the knowledge of metadiscourse use in Indonesian context. As a

good reference, this study can be one of it in the metadiscourse investigation

studies. Further, this study could be one of research article that introduces the

Indonesian context in metadiscourse research.

1.6 Definition of Key Terms

1. Discourse analysis is an analysis of language which focuses on investigation

about how to use language correctly (Brown & Yule, 1983; Edwards, 2008;

Hyland, 2005). The two of essential aspects of discourse which are used in

this study are metadiscourse and corpora, the former is use to express the

important meanings correctly, to organize text, and to make interaction

 8

between them for understanding the text (Fa-gen, 2012) and the latter is to

give us the information how the language is used.

2. Interactional metadiscourse is commonly used to involve the readers in the

text and to give them opportunities to respond what the perspective of the

writer is (Hyland, 2005).

3. Interactive metadiscourse uses to make the readers find the coherence and the

conveyance of the text (Hyland, 2005).

 9

CHAPTER II

REVIEW OF RELATED LITERATURE

2.1 Genre Analysis

Genre analysis is one of the types of discourse analysis. It is employed to

analyze the genre of the text both oral text and written text. To analyze the genre,

it focuses on the idea of texts in genre theorists which will be similar or different

and depends on the classification between the genres. In order to organize those

different classifications, research has set up to distribute the key linguistic and

rhetorical features in the particular genre (Hyland, 2005).

Rhetorical features is kind of art persuasive features focusing on the

arguments in no formal proof. In the recent past, rhetoric tended to involve

negative connotations and to give some immoral suggestions, but today rhetoric

has been an important aspect of working in text analysis and written

communication (Hyland, 2005). The use of rhetorical features divided into:

macro- textual level analysis and micro- textual level analysis. The first is used to

analyze the move structure of the text which assigns the pragmatic function and

builds the schematic structure thus mean to achieve the communicative purposes.

The second is used to analyze the interaction of the text which is concentrated on

the use of reference, mood and modality, and metadiscourse (Vergaro, 2004).

Thus, metadiscourse as the subject which is applied in this study known as one of

the rhetorical features (Swales, 1990).

9
 10

2.2 Metadiscourse

In writing text, there are two levels that the writers have to reach. On one

level, they have to supply the information or subject what they are going to write.

On this level, they should expend the propositional content. When the writers are

in the propositional information level, they will search and supply many

information what they are going to write. Then, they will sort out that information

to choose the suitable one for their texts. On the other level, they do not have to

expend the propositional, but they should expend metadiscourse level. The writers

would know how to help the readers classify, organize, react, and evaluate in what

they wrote (Crismore, 1984; Vande Kopple, 1985). In addition, Fa-gen (2012)

states that metadiscourse is used in expressing the important meanings correctly,

in organization text, and interaction between them and their readers for

understanding the text.

Metadiscourse is widely used in the terms of discourse analysis and

language education, referring in interaction between the text producers and the

texts or between the texts producers and users (Hyland, 2005). In discourse

analysis, sometimes metadiscourse become the main object of study to analyze the

genre of texts by analyzing the function. In contrast, in language education, for

writing skill, learning about metadiscourse is truly important for teacher and

students. In order to produce the good writing and coherent, both teacher and

students should understand the use of meadiscourse in constructing the

relationship between writers and readers.

 11

2.2.1 Metadiscourse Types

The model of metadiscourse has been revised time to time. It has been

changed in form of models and functions. There are many studies contribute the

change of the metadiscourse model. Early study, Crismore (1984) defines the

metadiscourse seeks to involve the reader as an active human being and indicates

the presence of author in the texts. It differs the metadiscourse model into two

different types, they are: “informational” and “attitudinal” (p. 10). The different of

metadiscourse is based on the function. The first indicates the existence of an

author, it consists of the explicit rhetorical devices. The function of explicit

rhetorical is referential or ideational, that is, to give author’s idea about content

explicit and structure.

They include devices for the effective communication such as: to state the

author purposes, goals, or aims; to decide main statement as a discourse topic

consists of a single topic and the predication or comment (called theses,

controlling idea, topic or main idea statement); to decide the justifications

between main statement; to label the discourse as the description, argumentation,

exposition; or to announce the topic shifts. On the other side, the second is to

show the author’s attitude by using word choices in the propositional content (by

using word amazingly, interestingly, etc.) or about how firm the propositional

content is (by using word true, probably, uncertain, etc.) In addition, the

metadiscourse are used to convey the attitude and comment of author such as:

pronoun of reader preference (I, we); mental state and process verb (feel, think
 12

realize, argue, etc.) and adverbial construction that qualify the whole sentence

(probably, naturally, it is true that, etc.) See appendix 1 for detailed description.

The next studies, Vande Kopple (1985) declares that metadiscourse divided

into “interpersonal” and “textual” metadiscourse (p. 86). Interpersonal

metadiscourse is used to help the reader react and evaluate what the idea in the

text. The sub- classifications are: “validity markers,” attitude markers,” and

“commentaries” (p.84-85). Next, Textual metadiscourse is used to make the text

makes sense in context to explain the function as the massage, being more

understandable and coherent. It also helps the readers to organize how the text

organized. It includes sub- classification as: “textual connectives,” “code glosses,”

“illocution markers,” and “narrators” (p.83-84) See appendix 2 and 3 for detailed

description.

The next previous study, Crismore, Markannen, & Steffensen (1993) claim

that the types of metadiscourse as the same as what the previous study stated.

Metadiscourse became two different types “interpersonal” and “textual” (p.47-

54), but it is different in the sub- classifications of each metadiscourse type. For

textual metadiscourse, it dropped the sub- categories of “narrators” and “textual

connective”. It changed “textual connectives” became “textual marker” (p.48)

which had other subcategories namely, “logical connectives,” “sequencers,”

“topicalizers,” and “reminders” (p.48-49). It shifted “code glosses,”

“announcements,” and “illocution markers” (p.49-50) in one category namely,

“interpretive markers” (p.49). For interpersonal metadiscourse, it separated

validity markers and became three subcategories they are “hedges,” “certainty
 13

markers,” “attributor,” and it shifted “commentary,” and “attitude markers” (p.50-

54) for the subcategories. See appendix 4 and 5 for detailed description.

Recently, Hyland (2005) argues that textual metadiscourse is not as

metadiscourse function. So-called ‘textual’ device organizes text to relate

statements about the ideas and metadiscourse relates statement of the text for the

readers; Their functions are not independently of two functions. Further, textual

function in a text will be organized when the propositional function and

metadiscourse collaborated as one. Therefore, in this model it focuses

interpersonal function of metadiscourse which it divided into two general

categories “interactive” and “interactional” (p.49). For each type of metadiscourse

has different sub- classifications.

From those previous studies, it showed that metadiscourse revised time to

time, depend on the function or the classification itself. Especially, in this study, it

focuses on the metadiscourse classification of Hyland’s model which is divided

into interactional and interactive metadiscourse. Besides it is flexible to identify, it

is understandable with each metadiscourse signals that they have.

1. Interactive Metadiscourse

These features are used to organize the propositional ideas of the text.

Then, the readers can find the conveyance and the coherence of the text. Here,

the author’s purpose is to shape and constrain a text, find the needs of the

readers so that they can set up their arguments to interpret in what they read

(Hyland, 2005).
 14

Interactive metadiscourse has 5 sub- classifications, they are: “transition

markers,” “frame markers,” “endophoric markers,” “evidential,” and “code

glosses” (p. 50-52). For each member has different function and signals. See

table 6 for detailed description.

As stated earlier, these markers were identified and categorized

based on Hyland’s (2005) classification of interpersonal metadiscourse

markers (interactive and interactional). It was subdivided into categories for

which the description is provided in Table 1 and 2.

Table 1. Interpersonal model of metadiscourse, interactive markers

(Hyland, 2005)
Macro category Subcategory Examples
Transitions a) Addition And, furthermore. moreover,
b) Comparison also, in addition anyway
c) Consequence in contrast, however, but, on
the other hand, on the contrary
consequently, after all, then,
therefore, as a consequence
Frame markers a) To sequence (in) Chapter X, first, next,
b) To label stages lastly, I begin with, I end with
c) To announce goals all in all, at this point, in
d) To shift topic conclusion, on the whole
my focus, goal, objective is to,
I seek to, my purpose is to
back to, in regard to, return to,
turn to
Endophoric markers Noted above/see Fig./in
Section 2
Evidentials According to X/(Y, 1990)/Z
states
Code glosses Namely/e.g./such as/in other
words/that is/ to put it simply
For example/for instance

Hyland (2005) classifies interactive metadiscourse into five major

categories which are transitions, frame markers, endophoric markers, evidentials,

 15

and code glosses. Two of them consist of some subcategories to mark their

significant forms in text. Each main category performs special function.

Transitions involve an array of devices, mainly conjunctions, applied to mark

additive, contrastive, and consequential steps in the discourse, as was against to

the external world. Frame markers refer to text boundaries or items of schematic

text structure, including items used to sequence, to label text stages, to announce

discourse goals, and to represent topic shift. Endophoric markers create more

materials available and important to the reader in recovering the author’s goals by

pointing to other parts of the text. Evidentials show the textual information’s

source which originates the outside of current text. Code glosses refer to the

ideational information’s restatement.

Here, there are some examples from both corpora which are italic and

underlined.

Examples:

A. Interactive Metadiscourse Markers

1. Transitions

a) addition

In addition to analysis of structural conditions, in the other side it

is important how the intellectual discourse mission to expand public

sphere.

b) comparison

He states most of different features of individualism which are

clarified in contrast to previous conditions, he implies organizational
ordered causalities (on the first step, the effects of financial economic),
 16

but he teaches it more than other things with real issues of humanity in
all individual level.
c) consequence
As a result, high serum cholesterol level can be achieved due to

hepatic dysfunction.

2. Frame markers

a) Sequencing

The idea of individuality from Simmel’s point of view is

designed as below: first, individuality as a separate entity (but not

unique or full) and consequently, it is understood as the idea of

quantitative individualism.

b) Label Stage

In summary, this is the first study demonstrating the effect of

exercise training on visceral fat ABCG1, ABCG5, and visfatin genes

expression.

c) Announce Goal

One existing study has focused on the effects of unsaturated free

fatty acids on plasma concentration of adipokine peptide: Cooper et al.

showed that dietary fatty acid composition significantly reduces

plasma PYY concentration and can increase plasma ghrelin

concentration that is not significant.

d) Shift Topic

With regard to the socio-economic situation of the Iranian

society in this period, the following factors can be considered

 17

influential in the increase in the number of employed population in the

agricultural sector, decrease in the degree self-sufficiency in the

production of agricultural products, and increase in the import of

agricultural products.

3. Endophoric markers

A high score in this part can be interpreted in this way, which

management style in an organization is very important; because if it is

assumed that all facilities be provided for research activities; but there

were no comprehensive management to organize them, facilities will not

be helpful.

4. Evidentials

According to the literature, the prevalence of cardiovascular

involvement in patients with SLE has been estimated to be more than 50

and it was shown that the left-sided heart valves are affected most

commonly.

5. Code glosses

Similar to other research studies, we also observed the highest

prevalence of LBP in last year among health care workers (e.g. hospital

staff, nurses, and dentists).

2. Interactional Metadiscourse

These features concern on the way how the author’s comment on their

massage. It used to open opportunities for the readers to respond the author’s

perspective. Metadiscourse here to evaluate and to engage, to express

 18

solidarity and to respond the dialog of imagination that written in the text

(Hyland, 2005).

There are 5 types of interactional metadiscourse: “hedges,” “boosters,”

“engagement markers,” “attitude markers,” and “self-mentions” (p.52-53). For

each type has different function and signals. See table 7 for detailed

description.

Table 2. Interpersonal model of metadiscourse, interactional markers

(Hyland, 2005)
Macro category Subcategory Examples
Hedges a) Epistemic verbs May/might/it must be two
b) Probability o’clock...Probably/perhaps/maybe...
adverbs unlikely, unclear
c) Epistemic
Expressions
Boosters a) To sequence Certainly, demonstrate, really
b) Emphatics totally, always in fact; definitely;
c) Amplifying it is clear that
adverbs
d) Cognitive verbs
Attitude markers a) Deontic verbs Have to/need to/we must know...
b) Attitudinal Unfortunately/remarkably…It is
adverbs absurd/it is surprising... I feel/I
c) Attitudinal think/I agree/I believe
adjectives
d) Cognitive verbs
Self-mentions first-person pronouns I, me, my, mine
Engagement a) Second-person You, your, yourself
markers pronouns by the way, you may notice
b) Interjection consider, note that, see, look at
c) Imperative verbs must, have to, should
d) Necessity modals

He also organizes interactional metadiscourse into five main categories with

specific functions which are hedges, boosters, attitude markers, self-mentions, and
 19

engagement markers. They were divided into some subcategories. Hedges

(epistemic verbs, probability adverbs and epistemic expressions) signal the

writer’s reluctance to display propositional information categorically. Boosters

(emphatics and amplifying adverbs) represent certainty and emphasize the force

of propositions. Attitude markers (deontic verbs, attitudinal adverbs, attitudinal

adjectives and cognitive verbs) present the writer’s appraisal of propositional

information, showing surprise, force, approval, importance, and so on.

Engagement markers (second-person pronouns, interjection, imperative verbs and

necessity modals) explicitly refer to the readers, either by focusing their attention

selectively or by including them as text participants through second person,

pronouns, imperatives, question forms, and so on (Hyland, 2001). Self-mentions

(first-person pronouns) offer the extent of the presence of the writer in terms of

possessives and first person pronouns, schematic text structure, including items

used to sequence, to label text stages, to announce discourse goals, and to

represent topic shifts.

1. Hedges

a) Epistemic verbs

The collapse of traditional cooperatives (Boneh) as collective

production units could demonstrate the downfall of traditional

partnership units.

b) Probability adverbs
 20

Since the K+-ATPase has a significant role in insulin secretion

of the pancreas; hyperglycemia indicates that insulin secretion process

may be affected by MnO2.

c) epistemic expressions (probability adjectives)

This means that if a person has high level of cultural capital, it is

probable to have quicker initializing of official values of society.

2. Boosters

a) Intensifier adverb

Certainly there are different statistics in this regard because of

prohibition and limitation resulting from that.

b) Intensifier adjectives

According to the results of cultural capital and self-satisfaction,

it is obvious that there is a weak and significant relation between these

two variants.

c) Intensifier verbs

These individuals are able to think and infer innovatively or have

personal philosophies about the existence and truth.

3. Attitude markers

a) Deontic verbs

The concept of intangible success factors is used to refer to

individual intangible assets and also the activities related to improving

or utilizing the assets, i.e. any intangible phenomena that are to be

measured.
 21

b) Attitudinal adverbs

However, nasal trauma was significantly less in HHHFNC group

compared with NCPAP.

c) Attitudinal adjective

It is important to understand the cause of changes in body weight

gain and glucose level and their correlation induced by MnO2

particles.

d) cognitive verbs

It is believed that mitral valve repair can be justified in specific

patients.

e) Self-mentions

The authors concluded that napping is associated with elevated

prevalence of diabetes and impaired fasting glucose.

4. Engagement markers

a) Reader pronoun

You et al. studied the effect of GA in two types of lung cancer

cells (A549 and Calu-6).

b) Interjections

Television and satellite networks have led to formation of

personality and identity of individuals and they are influenced by

multiple and sometimes conflicting information. TV, print, and

Internet advertising platforms that enable a variety of new modes and

 22

by the way, they prepare the ground for a new fashion trend toward

familie s and especially young people.

c) Directive imperatives

Note the use of Newton’s third law: the force of boundary on

fluid is minus the force of fluid on boundary.

d) Necessity modals

It is, therefore, an inevitable need to understand scientifically

such a growing in Iran, keeping an eye on its implications.

2.3 Local, National and International Journal Articles

Javid (2014) stated National Journal is a Journal where you can present

your research paper and is held within your country only, and only the researcher

within your country are allowed to present research papers. International Journal

is a Journal where you can present your research paper and is held in your

country or some other country , but the researchers from other countries are also

allowed to present research papers. Journal papers are the research papers being

published in Journals as a hard copy, while the conference papers may be

available online in their database.

A regularly published collection of articles that focus on topics specific to a

particular academic discipline or profession. Journals might be published

monthly, bi-monthly, quarterly, semi-annually, or even annually. Probably the

most common publication frequency is monthly and quarterly. Journal articles are

typically of substantial length (often more than 10 pages) and usually reflect
 23

research, whether it be surveys of existing research or discussions of original

research. Most journal articles will be prefaced with an abstract and will include

extensive documentation within the article or at the end of the article. Most

research begins with a survey of existing literature on a topic and proceeds with

the development of new ideas or new research into a topic. Articles are usually

written by experts in their fields, although journals might also publish letters from

their readership commenting on articles that have been published in previous

issues. Journals might also include opinion articles or editorials. Examples of

journals include Journal of the American Medical Association, American

Sociological Review, Psychological Reports, Publications of the Modern

Language Association, Educational Research Quarterly, and Evolutionary

Biology.

Academic journals first appeared in the 17th century, beginning with

the Journal des Sçavansin 1665 and followed by the Philosophical Transactions

of the Royal Society of London a year later. Their importance as a means of

disseminating knowledge has grown considerably since then. After 150 years of

sharing the spotlight with letters and scholarly treatises, by the early 1800s,

academic journals had become the fastest and most effective means of

disseminating research findings. The better part of the 20th century was devoted

to consolidating their position within the system of knowledge dissemination1 and

in particular in the natural and medical sciences.

Nowadays, academic journals are often divided into two categories: the so-

called international journals, typically published in English, and national and local
 24

journals, published in a language other than English. However, for the last 20

years or so, researchers have been encouraged to publish in international journals,

usually at the expense of their national counterparts despite the latter being just as,

if not more important, in some fields. It is important here to draw a distinction

between research topics in the natural sciences and those in the social sciences and

humanities. While natural sciences are by definition “international” – the electron

behaves the same way in Quebec as in Japan – the history, social structure or

education system of these two societies are very different. One of the central roles

of national journals is to provide a venue for the publication of research findings

related to local or national topics, which are likely to have a direct impact on the

societies concerned.

2.4 Corpora Analysisa method

Corpora analysis is a methodology to investigate of linguistic phenomena in

communicative situation by software via computer. It is known by corpus- based

studies focus on the analysis of corpus in the text. Additionally, corpus (pl.

corpora) refers to a collection of text which is electronically encoded which seems

a representative of some subject of language and uses for linguistic analysis

(Baker, 2006). The software that can be used to encode and to analyze the corpora

they are: Wordsmith Tools 4.0, Wordsmith Tools 6.0, Lextutor, Web Corp

Lingist’s Search Engine (WebCorpLSE), AnConc, ConcGram, etc.

In discourse analysis, corpora, having one major advantage, had been really

helpful. It can reduce researcher bias. When carrying out discourse analysis, the
 25

problematic such of cognitive biases can be present. In case, in kind of research to

select newspaper article, the article will be selected in the subjective point of view

and will ignore other newspaper presenting different perspective. By using

corpora, the researcher can place a number of restriction on cognitive biases

problem. Corpora researchers can be selective in choosing which aspects of their

research (Baker, 2006).

Another aspect which is essential in corpora analysis is sampling. The use

corpora in discourse analysis, it is possible to use small data for the representative

(Baker, 2006). Then, the sampling is needed. It is based on what corpus

depending from the genre of the text that will be analyzed. If it is the analysis of

pronoun in the newspaper, the sampling will be taken from the pronoun markers

in the text. Therefore, corpora have text collection which are balanced for the text

that are representative based on the genre of the text.

A number of studies have used corpora on their research. It is about no

metadiscourse analysis such as: grammar (Friginal & Mustafa, 2017), metaphor

(Deignan & Potter, 2004; Semino, Heywood, & Short, 2004), apologies (Lutzky

& Kehoe, 2017), pronoun (Kim, 2009) and sales promotion (Cheung, 2011). As

Kim (2009) declares the use of two different personal pronouns in Korean and

English corpus plural personal pronoun (uli and we) and second person pronoun

(tangsin and you) for ‘Reader- Involvement Evoking acts’ (RIE acts) around

newspaper science popularizations. Information about the corpora is English 115

texts and 8855 sentences for Korean 241 texts and 8917 sentences. It is different

because of the average length of Korean article is shorter than English and
 26

English newspapers are published more frequently than Korean. The corpus

software used in this study is Wordsmith Tools 4.0. This study finds that

quantitatively there is the significant differences in using pronoun between two

different languages. The lower frequencies of the Korean personal pronoun

because the omission of pronoun is allowed in Korean.

2.5 Research Article

Metadiscourse has been investigated in many descriptive stand points and

has been shown as the essential feature in some types of academic discourse.

There are numerous studies which is aimed the metadiscourse role in pedagogy

(Gholami et al., 2014). Several studies which is included to find the effectiveness

of metadiscourse on the writing abilities (Crismore et al., 1993; Lee & Deakin,

2016; Rustipa, 2014; Steffensen, Margaret S.; Cheng, 1996), reading

comprehension (Crawford-Camiciottoli, 2003), listening comprehension

(Heshemi, Khodabakhshzade, & Shirvan, 2012), and lecture comprehension

(Ghaemi,F., & Naderi Anari, 2014). For instance, Heshemi et al. (2012) state that

the role of metadiscourse in the effectiveness of listening comprehension seems

have been ignored. Then, the study aimed to know the effective of metadiscourse

role in listening comprehension among EFL students. It finds that the role of

metadiscourse in improving listening comprehension is significantly effective

depended on the student EFL awareness of metadiscourse use in the monologues.

In recent years, many considerable studies of text analysis have been

conducted. Among them, some studies occurred to clarify the characteristic genre
 27

of the text or text types in term of discourse, structure, and metadiscourse

properties. In some studies, metadiscourse had been explored in various genres

such as: academic research articles (Abdi et al., 2010; Abdollahzadeh, 2011; Del

Saz-Rubio, 2011; Gillaerts & Van de Velde, 2010; Koutsantoni, 2006; Salas,

2015), newspaper discourse (Dafouz-Milne, 2008; Le, 2004), post- graduate

dissertation (Bunton, 1999), commercial websites (González, 2005),

advertisements (Saadi & Roosta, 2014), and academic writing (Mur Dueñas,

2007; Nasiri, 2012). As Dafouz-Milne (2008) declares how the role of textual and

interpersonal metadiscourse to persuade among newspaper readers. This study

finds interpersonal metadiscourse is not only the one which persuade the readers,

textual metadiscourse in the form of logical markers, is essential to give the

persuasive effect in the texts. Nevertheless, it seems that the readers prefer both

textual and interpersonal metadiscourse to guide them through the text.

Most of the comparative studies which is focused on metadiscourse feature

among cross- linguistics. Those studies are the comparison between English and

other language in the use of metadiscourse such as: Chinese (Loi & Lim, 2013;

Mu et al., 2015), Turkish (Ozdemir & Longo, 2014), Finnish (Crismore et al.,

1993), Spanish (Lafuente-millán, 2014; Lee & Casal, 2014; Martı n-Martı n,

2003;

Moreno, 1997; Mur-Dueñas, 2011; Mur Dueñas, 2007; Pérez-llantada,

2010), Slovene (Peterlin, 2005), Persian (Gholami et al., 2014; Zarei & Mansoori,

2011), and Spanish and Norwegian (Dahl, 2004). As a case in point, Mu et al.,

(2015) examined the use of metadiscourse in different languages in research

 28

articles (i.e. English and Chinese). Before, this kind of study had been conducted

by (Loi & Lim, 2013), which only focuses on the introduction of research articles

(RAs). It finds that Chinese and English applied the different attention in using

metadiscourse in RAs. The English RAs used more interactional metadiscourse

than Chinese RAs. The result demonstrates that the English pays more attention to

the interaction between the writers and their readers. In addition, it shows the

different ways of metadiscourse features used between both of them.

Furthermore, fewer works have been compared the metadiscourse features

in different genres of specific languages, they are: L1 and L2 undergraduate

student writings (Lee & Deakin, 2016), EAP lessons and university lectures (Lee

& Subtirelu, 2015), PhD theses and research articles (Kawase, 2015), and

paradigmatic and disciplinary of text influenced by metadiscourse (Cao & Hu,

2014). In a similar case, Kawase (2015) states that no study examined how and

why metadiscourse is employed differently in the introduction of PhD theses and

RAs which are written by the same authors. Then, this study finds the different

amount of metadiscourse that used in those texts. Especially in the use of

endophoric markers, all authors used it to present previews in their these, but none

of them used it in their RAs. However, this study suggests that to make the greater

use of metadiscourse in PhD theses than in RAs may not be applicable with regard

in introduction. Moreover, the variation of metadiscourse that used in the

introduction of local, national and international journal article on english

education
 29

2.6 Revies of Relevant Studies

Before going further to this study, the researcher found some previous

studies related to this research. The first is Representation of Rhetorical Move of

Thesis Abstracts in English Teacher Education Departement by Lathifatul

Fajriyah (2015). In this study, Fajriyah analyzed the move structure in 22

thesis abstracts of undergrduate students in English Teacher Education

Department. She found out the rhetorical moves of thesis abstracts and differences

moves among thesis abstracts of English Teacher Education Department. Even

her study and this study analyze the similar data, however, this study tries to

examine difference aspect of abstract, which is metadiscourse categories.

The second is study conducted by Mirshamsi and Allami (2013) entitled

Metadiscourse Markers in the Discussion/Conclusion Section of Persian and

English Master's Theses. In Mirshami and Allami study, they used Hyland’s

model of metadiscourse to examine the Master theses. They observed about

interactive and interactional metadiscourse markers used in discussion and

conclusion sections of master’s theses. They tried to find out the similarities and

differences in the use of metdiscourse markers in three categories, they are native

English speakers, native Persian speakers, and non-native English speakers.

And the result showed that native English writers used more interactive and

interactional metadiscourse markers than native Persian and EFL learners.

Eventhough their study and this study observe the same scope of study,

which is metadiscourse markers which adopted by Hyland’s model, there are

 30

several differences between both studies. Mirshamsi and Allami used discussion

and conclusion sections of master’ theses as a data, while this study using

introduction of local, national and international journal article on english

education as a document.

The third is study by Al-Shujari, Ya’u and Buba (2016) entitled Role of

Moves, Tenses, and Metadiscourse in the Abstract of an Acceptable Research

Article58. In their study, they investigated 60 research article abstracts that

selected from two diciplines, which were applied linguistic and teaching English

as a second language. All the seleted articles were taken from Pertanika journal of

social science and humanities (official journal of University Putra Malaysia)

which published between 2011 aand 2015. They analyzed the rhetorical

moves, preferred tenses of each move, and metadiscoure categories which used

Hyland’s model metadiscourse in 60 research abstracts.

What differ between their research and this research is the aspects they

analyzed were move structure, preferred tenses and metadiscourse categories in

abstracts, while this reseach only concerning on metadiscourse categories. And

the subject fields’ of Al-Shujari, et.al are applied linguistic and English as a

second language, whereas this research’ subject fields is English Education. The

fourth is Metadiscourse in Indonesian EFL Learners’ Persuasive Texts: A

Case Study at English Department, UNISBANK, study by Kathrina Rustipa

(2014). Her study explored about the use of metadiscourse markers in persuasive

writing of 7 Indonesian EFL learners. She used Hyland’s model of metadiscourse.

She tried to figure out the metadiscourse markers applied by Indonesian EFL
 31

learners in persuasive writing. And her study 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).

This previous study and this study analyze the same scope of study, whic is

metadiscourse markers. Rustipa study’s, however, has significantly differences

with this study, they are different in the type of corpus. This study uses

introduction of local, national and international journal article on english

education as document.

And the last is Metadiscourse in the introductions of PhD theses and

research articles by Tomoyuki Kawase (2015). Kawase’s study examined how

eight writers construct metadiscourse in the introductions of their PhD theses

and research articles that they later produced based on the theses. By doing this,

he believed that it examined the assumption that variations in the use of

metadiscourse in those texts could be attributed to the nature of the genre. The

research findings showed that the majority of writers made greater use of

metadiscourse in their research article introductions. Both Kawase’ study and this

study examine metadiscourse markers which used Hyland’s model of

metadiscourse as the means to analysis. What makes it different is the corpus

used. His study concerned on introduction sections of PhD theses and research

articles, while this study focusing on undergraduate’ thesis abstracts.

 32

CHAPTER III

RESEARCH METHOD

3.1 Research Design

This study is a corpus- based studies or corpora analysis. This kind of study

is suitable in order to determine the use of linguistic features in a text likewise

metadiscourse features. Baker (2006) states unlike qualitative research design,

corpora analysis implements more quantitative design for instance by using

frequency of linguistic features in texts. Additionally, Biber, Conrad, & Reppen

(1998) point out that corpus- based studies depends on both quantitative and

qualitative research design. It states ‘associations patterns’ refers to quantitative

relations using to measure the characterizations exist in different pattern

associated with contextual patterns. Nevertheless, functional interpretation

(qualitative design) is also an important design in corpora analysis.

Therefore, the use of corpora analysis in this study, to identify and

investigate the metadiscourse markers by the combination of quantitative and

qualitative research design as the research methods. By quantitative design, the

different and similar amount in the use of metadiscourse signals between different

journals will be exposed. The data of this method would be the different amount

of metadiscourse in form of number. On the other side, by qualitative design is

applied to analyze the function of the corpus in the text based on the

metadiscourse function theories. Then, the way how they use those metadiscourse

will be delivered in form of description.

26
 33

3.2 Research Object

In order to compare metadiscourse features between Indonesian and

English native authors in their introduction in english educational journal, this

study was based on two comparable corpora each consisting of 60 introduction in

english educational journal. They were chosen in the field of 20 joall journal , 20

teflin journal and 20 asian efl journal.

The first corpus came from 20 introduction in english educational journals

which were specifically defined as only written and edited by English native

authors. The editors and the writers came from English speaking countries. The

journals that had been chosen for this study such as asian efl journal.

Meanwhile, the second came from 20 local journals which were specifically

defined as only written and edited by Indonesians. The journals that had been

chosen for this study such as teflin journal

The third corpus came from 20 national journal abstracts which are

specifically defined only written and edited by original lecturer English writers.

Editors and writers come from English-speaking countries. The journals that had

been chosen for this study such as joall journal. For clear description, see Table 1.

Table 1. Description of Corpus

No Journal Frekuensi Persentage

(F) (%)
1 Joall article 20 33,3

2 Teflin article 20 33,3

3 Asian efl article 20 33,3

 34

3.3 Data Collection

In order to investigate the distribution of interactional metadiscourse

markers in the introduction sections of the research articles, a manual corpus

analysis was carried out primarily to provide a qualitative and comprehensive

picture of how metadiscourse markers are used in the specific genre of local,

national and international article.

To this end, the corpus was examined for all instances of interactional

metadiscourse markers listed by Hyland (2005). For the analysis of the

aforementioned categories, the whole corpora were examined word by word,

rather than selecting a number of typical hedges, boosters, etc. from a list.

Following Crismore et al. (1993), since the size of the introduction sections

across the fields is inevitably unequal, the frequency of metadiscourse markers

was calculated per 1,000 words to make the length of the texts consistent and

ensure the comparability of the results. After determining the types of

metadiscourse markers employed in the introduction sections of the sample

articles, the gleaned data were quantitatively analyzed in order to identify their

frequency of occurrence in the texts and examine whether there was a statistically

significant difference between the two sets of corpus data.

3.4 Data Analysis Procedure

1. Based on the aim of the study, the articles that were investigated were

international journal articles national journals and local journals.

2. For international journal articles, as mentioned above 20 Introduction in

english educational were taken from different journals. Therefore, the

 35

total number of Introduction in english educational were 20. These 20

Introduction in english educational built corpus of international journal

articles.

3. For national journals, 20 Introduction in english educational were taken

from different journals. The total number of Introduction in english

educational were 20. These 20 Introduction in english educational built

corpus of national journals.

4. Similarly, for local journals, 20 Introduction in english educational were

taken from different journals. The total number of Introduction in

english educational were 20. These 20 Introduction in english

educational built corpus of local journals.

5. Based on the purpose of this study, it was only focused on the abstract of

Introduction in english educational, then the other indicators of

Introduction in english educational were ignored. These 40 abstracts

were as the corpora of this study.

3.5 Technique of Data Analysis

As mentioned above, in corpora analysis, both quantitative and qualitative

research design were crucial.

1. To answer the first aim it used quantitavie research. The analysis of

the selected texts was closely done based on Hyland’s (2005) taxonomy.

The data were collected through extracting metadiscourse markers from

the Iranian papers written in English. In this study, metadiscourse markers

were counted and classified in their corresponding groups based on the

 36

category proposed. Each sample text was read and the number of

metadiscourse markers was counted by the first rater one by one. Then,

each sample text was reread by the second rater sentence by sentence to

identify metadiscourse markers once again to make sure that any kind of

mistake did not happen during first count. Finally, each metadiscourse

marker was checked again to make sure it was correctly classified. After

collecting data, the SPSS version 22 software was applied to achieve

quantitative analysis. Furthermore, to examine whether there were any

significant differences in the use of metadiscourse markers in the

discussion sections of these research articles, a Chi-square analysis was

run.

2. Therefore, in order to answer the second aim, it used qualitative research

design. Moreover, there are several steps that to investigate how the

metadiscourse markers were used. First, for each metadiscourse category

had the most frequently appeared marker. Those most frequently appeared

for each metadiscousre category as the representative in analyzing the

function. For both different corpora had different representative in their

each metadiscourse category. Next, for all representative markers were

analyzed based on their function in each metadisourse category.

Sometimes, one or two markers commonly use in different categories,

they were analyzed based on the function in sentence or texts and based

on metadiscourse use theory.

37
 38

A COMPARATIVE STUDY ON METADISCOURSE MARKERS USED IN

LOCAL, NATIONAL AND INTERNATIONAL ON ENGLISH
EDUCATIONAL JOURNAL

PROPOSAL

Presented as a partial requirement for Master Degree in post Program of

English Education of Bengkulu University

BY

HAIRI YANTO

NPM. A2B0170008

POST GRADUATE PROGRAM OF ENGLISH EDUCATION

DEPARTEMEN OF TEACHING AND ART
FACULTY OF TEACHER TRAINING AND EDUCATION
BENGKULU UNIVERSITY
2019