Enhancing Discourse Markers Awareness among Iranian EFL Learners: A Product - Process Based View

Document Type : Original Article


Imam Khomeini International University, Qazvin, Iran


While several researchers have studied discourse markers to determine their roles in language skills, still research needs to address these devices in relation to second language learners writing proficiency Hence, the present study had two main goals: identifying the usage of discourse markers used in Iranian writing compositions without instruction, and describing how treatment of discourse markers functions in learners’ writing.. Participants in this study were 60 upper-intermediate and intermediate learners studying English as the second language at Safir, Qazvin branch. Descriptive composition writing was assigned to students to write on the topics, selected by the researchers and based on Fraser's (1990, 1999) taxonomy of discourse markers. The results indicate that within the explorative section, with the proficiency level rising, the frequency and the type of the discourse markers used subconsciously without the treatment rise. At the same time, the treatment and discourse marker awareness among the learners can be of significant value in both the quality and quantity of discourse markers. Of course, the type of the treatment must be adjusted to the proficiency level of the learners. This is in keeping with the previous research that reported a positive impact of instruction of DMs on success in language writing (Feng, 2010).


1. Introduction

The field of second language (L2) writing has come of age. The formal study of L2 writers, writing, and writing instruction has a relatively short but fruitful history going at least as far back as the 1960s. Research on L2 writing has grown exponentially over the last 40 years and, during the late 1980s and the early 1990s, second language writing began to evolve into an interdisciplinary field of inquiry with its own disciplinary infrastructure. As Adel (2006) points out, nowadays writing in English plays a significant role not only in the academic context, but in the daily life as well.

Like Sarig’s (1995) study, the present study adopts Schiffrin’s (1987) term, i.e. discourse markers. On the one hand, the word “discourse” makes it clear that the items taught in this study function at the discourse level, i.e. above sentence boundaries. On the other hand, the word “marker” is more general than the term “connective,” and thus could comprise the different communicative functions that the items described in this study serve. Although, in the present study, the word “connectivity” not only refers to the connectivity between segments of texts but also to the connection established in texts between the writer, the reader, and the text, in many studies in the literature connecting function is associated only with linking segments of texts. This makes the word “markers” more appropriate to use in this study where discourse markers are signals not only of semantic relations but also of intentions, plans, attitudes, as well as social relations.

With this in mind,the present study has two main goals: identifying the usage of discourse markers used in Iranian writing compositions without instruction, and describing how treatment of DMs functions in learners’ writing. Regarding instruction and treatment, according to Nunan (1991), writing is not only the process the writer uses to put words to paper but also the resulting product of that process. This process and product are also conditioned by the purpose and place of writing (its audience and genre). Writing in a second language is further complicated by issues of proficiency in the target language, first language literacy, and differences in culture and rhetorical approach to the text. Instruction in writing can effectively improve student proficiency in a number of key areas. Regarding the importance of the study, it should be mentioned that an effective text is a text that achieves the text-producer’s communicative goals. Theoretically, discourse markers (DMs) are a class of verbal and non-verbal devices which provide contextual coordination for ongoing talk (Schiffrin, 1987). They help writers provide writing which is effective and satisfactory. Text-producers are not only engaged in presenting the propositional content of their texts as making sense, but they are also concerned with presenting this content in a way that fulfills their communicative intentions. Texts achieve the communicative intentions of their producers when text-receivers are able to recognize these intentions and to perceive the texts accordingly.

Discourse markers are essential tools for achieving the text-producers’ communicative goals. Because discourse markers are communicative tools that signal how text-producers organize, develop, and evaluate their ideas in texts, the use of these items is closely tied to the communicative norms that govern text production and perception in the language in which they operate. Therefore, the type, frequency, and distribution of these items that render texts acceptable and natural differ among languages. Accordingly, what is expected and accepted in using discourse markers in a certain language may not be thus in other languages. This language-specific use of discourse markers poses a challenge for L2 learners and translators. It has been noticed that the lack of competence in the use of discourse markers is an important factor for the lack of coherence in L2 writing, especially at the advanced levels. For example, Siepmann’s (2005) study indicates that very advanced German-native learners of English experience difficulties using discourse markers in a native-like manner. Moreover, since DMs facilitate communication, it is logical to suppose that the lack of DMs in an L2, or their inappropriate use could, to a certain degree, hinder successful communication or lead to misunderstanding. L2 students must learn to signal the relations of their utterances to those which precede and follow. Therefore, in terms of communicative competence, L2 learners must acquire the appropriate use of DMs of the L2. It is plausible to suppose that those nonnative speakers who are competent in the use of the DMs of the L2 will be more successful in interaction (both oral and written) than those who are not (Blackmore, 1993).

It is also important to know which type of writing instruction can be more effective regarding the proficiency level of learners. The product-oriented approach to the teaching of writing emphasizes mechanical aspects of writing, such as focusing on grammatical and syntactical structures and imitating models. This approach is primarily concerned with "correctness" and form of the final product. Moreover, this approach fails to recognize that people write for an audience and for a purpose and that ideas are created and formulated during the process of writing. However, the process-oriented approach emphasizes that writing itself is a developmental process that creates self-discovery and meaning. While the mechanical aspects of writing are important, they should not interfere with the composing process. This composing process requires much revision and rewriting. The teacher intervenes and guides students during the composing process but initially does not emphasize "correctness" and the final product comes only toward the very end of the writing process. Instead of worrying about form, students concentrate on conveying a written message. Hence the product of writing will improve with the discovery involved in composing. Product-oriented approaches to writing largely concern the forms of the written products that students compose. The writing exercises applied in this approach typically deal with sentence-level writing and paragraph-level organization. Students are often given a framework which illustrates a pattern of rhetorical organization; then, they are asked to fit their ideas into this framework. Both the content and the form which the students deal with are largely controlled by the teacher. Since the main focus of these approaches is on written form, grammar is emphasized and a particular effort is made to avoid errors.

Process-oriented approaches concern the process of how ideas are developed and formulated in writing. Writing is considered a process through which meaning is created.

This approach characterizes writing as following a number of processes: First, a writer starts writing ideas as drafts. Subsequently, he checks to see whether the writing and the organization makes sense to him or not. After that, he checks whether the writing will be clear to the reader. This approach focuses on how clearly and efficiently a student can express and organize his ideas. Students are first asked to go through the writing processes trying to organize and express their ideas clearly. The assumption is that what the student as a writer is going to say will become clearer through these processes. Students are also taught writing devices used in marking the organization and in making the general coherence clearer. 

2. Theoretical background

In this study, discourse markers are approached from a discourse perspective. Because their operational scope is considered to be “units of discourse such as sentences and paragraphs, rather than inter-sentential predications,” as Matras (1997, 177) puts it, they are best seen as providing information at the discourse level, and not at the sentence level. Focusing on the text in order to examine the function of discourse markers is based on text linguistics which views connected discourse as central to understanding language and grammar. This enables linguistic analysis to reach beyond the traditional level of sentence to the text as a whole (Hatim, 1997; Hoey, 2001, 2002; Longacre, 1995, 1996; Morgan and Sellner, 1980).

Modern linguistics, however, emphasizes the grammatical importance of connected discourse. For functional linguists, grammar cannot be understood “just by looking at it from its own level;” rather, it has to be approached “from above” (Halliday 2004, 31). This, Grimes (1975) argues, is “because certain factors are needed for the understanding of elements in sentences that are not available within those sentences themselves but only elsewhere in the discourse” (8). To dissociate grammatical phenomena from the structure of texts is to miss the essence of their use. Hatim (1997), states that, “Individual words, which are conventional elements of language,” he states, “have not been created in order that what they represent may itself be discovered (or known), but in order that they may be joined with one another, thus expressing, by their interrelations, valuable information. This is one honorable science and a great foundation principle.”

Another crucial assumption underlying this study is that a text is cohesive. Cohesion is defined by Halliday and Hasan (1976) in their seminal work as a semantic relation that is realized through the lexico-grammatical system. Brown and Yule (1986) and Morgan and Sellner (1980) represent a trend in the field of discourse and text analysis to treat coherence as being less dependent on the language of text itself, and thus not being a product of cohesion. In this study, however, formal markers provide well-defined tools for the study of coherence in text. Just like Halliday and Hasan (1976), the present study considers that explicit linguistic realizations of the underlying functional relations are crucial for the identification of a text as a coherent unit.  Alongside   formal   connectedness, coherent text must also display underlying relations between propositions and relevance to the situation.

Halliday and Hasan (1976) focus on cohesion across sentence boundaries. According to Halliday and Hasan (1976), conjunctive relations are based on the assumption that there are in the linguistic system forms of systematic relationships between sentences. These functional relations, where one segment of text elaborates, extends or enhances a previous segment of the text (Halliday, 2004,).  Conjunctions are classified into four basic categories that correspond to the semantic relationships they signal: additive, adversative, causal, and temporal.

Text is cohesive. Whether this is a by-product of the need to be coherent, as Morgan and Sellner (1980) have argued, or a prerequisite of coherence, as was originally argued in Halliday and Hasan (1976), seems irrelevant. Almost certainly the relationship works both ways.  On occasion, writers (and more rarely speakers) consciously produce cohesive devices in order to clarify or emphasize, i.e. to create coherence… On other occasions, a writer's or speaker's coherence is reflected automatically in the language they use, i.e. in cohesion. Either way, that is a feature of text that cannot be denied and one, furthermore, that continues to be the subject of study.

Also, discourse markers are extremely useful tools for clarifying the writers/speakers communicative intentions.  They signal  how  the  text-producers  “intend  a  message  to relate  to  the  foregoing  or  following  discourse  or  to  a  particular  aspect  of  the communicative situation” (Kroon 1997, 17). Underlying the description of discourse markers in this study, therefore, is the assumption that they perform not only connective but also communicative functions. This means that the writer/speaker employs them to signal to the reader/hearer what he is doing in the text and influence his understanding of what s/he says. The reader/hearer, in turn, uses these linguistic expressions to postulate the writer/speaker’s goals and intentions which can guide his interpretation process.

3. Discourse markers in literature

There is still no generally accepted list of members belonging to this linguistic group in English because researchers do not have the same conception of what counts as one (Brinton 1996, 31; Lenk 1998, 39). Compare for example the discrepancies between the list of discourse markers given by Schiffrin (1987, 2001) and the one given by Lenk (1998). Schiffrin’s list includes “Oh, well, but, and, or, so, because, now, then, I mean, and y’know,” whereas the focus of Lenk’s study of discourse markers are: “anyway, actually, however, incidentally, still, and what else.

Östman (1995), for example, prefers the term “pragmatic particles” because it covers the essence and flexibility of this group of elements better than any of the other terms. He considers the attribute “pragmatic” crucial in defining this group since it is indeed functional, i.e. pragmatic, aspects which we should associate with these particles. For Grosz and Sidner (1986), the term “cue phrases” is more preferable because each one of these devices cue the hearer to some change in the discourse structure. Blakemore (2001), however, who characterizes these items as “indicative” words with no “propositional meaning” uses the term “discourse markers” to underline the fact that the role of these expressions must be analyzed in terms of what they indicate or mark rather than what they describe.

     Despite the large disagreement in this area of linguistic studies, it is possible, Schourup (1999) argues, to identify a small set of characteristics most commonly attributed  to  discourse  markers  and  to  items  referred  to  by  other  closely  associated terms. The most common features that he realizes in these expressions from numerous studies in the discourse markers literature are: “multi-categoriality, connectivity, non- truth conditionality, weak clause association, initiality, and optionality”.

     DMs have been investigated in classroom oral discourse (Hays, 1992), informal settings (Lee, 1999; Muller, 2004; Trillo, 2002), reading (Abdullah Zadeh, 2006; Jalilifar & Alipour, 2007), lectures (Dailey-O’Cain, 2000; Perez & Macia, 2002), academic genres (Abdi, 2002; Blagojevic, 2003; Bunton, 1999; Longo, 1994; Mauranen, 1993; Ventola & Mauranen, 1993), and student writings (Connor, 1984; Field & Yip, 1992; Intraprawat & Steffensen, 1995; Johns, 1984; Johnson, 1992; Karasi, 1994; Norment, 1994; Steffensen & Cheng, 1996). These studies have targeted their use patterns of frequency.

Johns (1984) analyzed English essays by tertiary-level teachers following Halliday and Hasan’s (1976) model. In a similar study Connor (1984) compared six essays written by English native and ESL students, following Halliday and Hasan’s (1976) framework. Field and Yip (1992) compared 67 Hong Kong students with 29 Australian students writing on an argumentative topic. Johnson (1992) analyzed 20 expository essays in Malay, 20 essays in English By the same group of Malay students, and 20 essays in English by native speakers. Karasi (1994) analyzed 135 expository essays by Singapore secondary students. Norment (1994) studied 30 Chinese college students writing in Chinese and English on both expository and narrative topics following Halliday and Hasan’s (1976) framework. Intraprawat and Steffensen (1995) analyzed the DMs in persuasive essays by ESL university students. And Steffensen and Cheng (1996) analyzed texts written by students who worked on the propositional content of their essays and were taught using a process approach and students who concentrated on the pragmatic functions of DMs by enjoying direct teaching of DMs.

    In most of these studies,results have shown that conjuncts were overused and lexical cohesion was moderately used by native speakers (Johns, 1984), that non-native students of English used more conjunctions than Australian students did, (Field & Yip, 1992), that there was a difference between text types in the use of cohesive devices (Norment, 1994), that differences between essays that received good ratings and essays that received poor ratings were found in the number of words, T-units, and density of DMs (Intraprawat & Steffensen, 1995), and that students receiving direct instruction on DMs used them more effectively and also became more sensitive to their readers’ needs thereby making global changes that improved their papers (Steffensen & Cheng, 1996).

     The results of some studies were also contradictory. For example, while Connor (1984), Johnson (1992), and Karasi (1994) found no discrimination between native and ESL students in the frequency of ties, Norment (1994) discovered a correlation in the frequency of ties and the quality of writing. Results of the above studies, in general, suggest that language learners underutilize DMs (compared with native speaker use) especially for their pragmatic functions.

As elaborated on above,while  several  researchers  have  studied  discourse  markers  from  the  descriptive  and  contrastive  perspectives,  to determine the role of DMs and language skills, still research needs to address these devices in relation to second language learners writing proficiency.

The present study, then, reports on the use of discourse markers in academic compositions of Iranian students at upper-intermediate level of language proficiency taking a more bottom-up and mixed process-product view. The research is based on the premise that the knowledge derived from this investigation will provide insights into the nature of the academic compositions of EFL students.Indeed, the investigation of the role of discourse markers treatment in second language writing ability continues to be revealing for the better understanding of the nature of second language writing.

Based on what was said above, the following questions were posed:

1. To what extent do upper-intermediate and intermediate learners use discourse markers in their writings without instruction?

2. To what extent, can treatment enhance the use of discourse markers in writing compositions among Iranian upper intermediate and intermediate EFL learners?

4. Methodology

4.1 Participants

Participants in this study were 60 upper-intermediate and intermediate learners studying English as the second language at Safir, Qazvin branch. The average age of the learners were 25 and they all had Farsi as their mother tongue. Learners had been involving with writing different samples of writing before the term started and they were familiar with different types of writing but they hadn’t been taught discourse markers exclusively. All participants in this study were female.

4.2 Instruments

This study made use of two instruments for the purpose of gathering the data. Descriptive composition writing was assigned to students to write on the topics, selected by the researchers. The rationale for the selection of descriptive composition was that the students were more familiar with this kind of writing as describing places and events is a common practice in writing classes; therefore, the effect of rhetorical structure on learners' performance in writing was minimized.

     Fraser’s (1990, 1999) taxonomy of discourse markers was chosen for classification on the ground that it conforms to written discourse and that it seems to be the most comprehensive classification in written discourse. Fraser’s (1999) taxonomy includes three main subclasses. The first are contrastive markers that signal that the explicit interpretation of the second sentence contrasts with an interpretation of the first sentence. The second subcategory is elaborative markers that signal a quasi-parallel relationship between the sentences. And the third subclass, inferential markers, that signals how the following sentence is a conclusion derived from the preceding sentence.

Besides, Fraser (1999) distinguishes additional subclasses of discourse markers that specify that the following sentence provides a reason for the content presented in the previous sentence. (e.g., after all, because, for this /that reason, since). While Fraser’s first class of DMs involves the relationship between aspects of the explicit message of the second segment and either an implicit or explicit message of the first segment, his second class of DMs are distinguished by focus on topic, and so he calls them topic relating markers.

4.3 Procedure and data analysis

The whole research was done during two intensive terms of study which took 40 sessions, each session one hour and half, in three months. Classes were held three times a week, each day twice. There was a week or so break after the first twenty sessions.

To answer the first research question which seeks to analyze the extent to which learners apply discourse markers in their writing at intermediate and upper-intermediate level, an experiment was conducted within 10 successive sessions. Without any instruction, each week the participants were given a topic on which to write compositions individually in 45 minutes for two and half weeks. The total number of compositions delivered to the researcher was 600 papers. During and after writing they received no feedback.  Then, to reduce the danger of subjectivity in analyzing the compositions and to increase the reliability of results, two other applied linguistics instructors analyzed twenty compositions and calculated the number and type of DMs. After discussing the tiny differences ,agreement was reached by the four raters accordingly. Next, the compositions were carefully scrutinized to determine the extent that DMs were functionally appropriate in the context of use.

To answer the second research question, participants were divided into four groups which are shown schematically as follows:

Intermediate group (process-based) (A)

Intermediate (product-based) (C)

Upper-intermediate (process-based) (B)

Upper-intermediate (product-based) (D)

The students had writing courses with a product view in groups A and B. In remaining thirty sessions, the instruction and writing samples started. The researchers started to allocate 45 minutes each session to instruct DMs providing learners with lots of examples and referring them to many sources to see and evaluate how they are actually used in the text. Also learners were provided with different exercises to match related sentences using DMs and feedback was given on their writings explicitly. The courses were designed mostly depending on a text book; the students were taught the rules for composing, including the usage of discourse markers with some reading assignments. Then, they wrote their five-paragraph essays. The students prepared their papers considering the reading texts they had read previously in class. And then, they handed in their papers to be marked by the teachers. Consequently, they were handed the papers back to exchange views and discuss on if necessary. Accurate writing was primarily emphasized as the product view does.

In groups C and D, process-oriented writing courses were designed for students. During this period, the emphasis was on the process. Classes became ‘writing workshops’ for the students to develop their thoughts. They shared their work with others, and the teacher intervened, just like a coach, during all stages of composition development -prewriting, writing and rewriting.In an attempt to observe a comparison and contrast between the 5- paragraph essays written after product-viewed and the process-viewed writing courses, in terms of discourse marker usage, 60 essays (30 essays written after the product-viewed courses and 30 written after the process-viewed courses) have been analyzed and evaluated. Therefore, the data collected from the papers were analyzed and documented mainly in terms of the number of sentences written, the amount of the discourse markers used, and the variety of these expressions preferred.   

5. Results

For the first research question, the proportion of DMs to the total words in all compositions was manually calculated. Table 1 introduces the frequency and ratio of DMs in compositions written by the two groups.

Results showed that the subjects in this study employed a variety of DMs with some types used more frequently than others. Upper-intermediate students were considered the greatest users of DMs with the ratio of 44.52. Data revealed a positive relationship between language proficiency and the rate of DMs use. That is, the more experience one was in English, the greater the frequency of DMs.

Table 1

Distribution of All DMs in Compositions


Total number of compositions

Number of words

Number of DMs

Ratio of DMs per 1000





















The Chi-Square analysis was used to find out if the differences in the use of DMs between the groups were statistically significant. As shown in table 2, analysis revealed statistically significant differences between the groups in the use of DMs in their compositions at 0.05 level of significance.

Table 2

The Chi-Square Analysis of the Results







Asymp.   Significance


The statistically significant results acknowledged a direct relationship between the use of DMs and the composition writing experiences of the three groups. In order to be precise regarding the use of DMs, the frequency and percentage of the subcategories of DMs was also determined. As shown in Table 3, Elaborative markers were most frequently employed (40%), followed by inferential markers (23%), contrastive markers (9.9%), and causative markers (5.6%). This order was observed by the groups. The extensive use of elaborative markers may be explained because descriptive writing in general requires elaboration of ideas which depends on the use of elaborative markers to signal the relationships between segments. Zhang (2000) reported a similar result in her study on cohesion extensive use, even overuse, of additive conjunctions such as and, also, besides, and in addition. One finding was that as the writing experience increased, the rate of DMs other than elaborative markers increased, but elaborative markers decreased. That is, there was a negative relationship between increase of composition writing ability and the use of elaborative DMs.




Table 3

The Percent of DMs Usage Types







343 (45.7%)

126 (16.8%)

76 (10.1%)

23 (3%)


467 (37%)

342 (27.1%)

123 (9.7%)

90 (7.1%)


810 (40%)

468 (23%)

199 (9.9%)

113 (5.6%)


The second research question tried to answer the following question:

To what extent, can treatment enhance the use of discourse markers in writing compositions among Iranian upper intermediate and intermediate learners?

After collecting and rating writing samples, the following results were obtained. In Group A and C in which intermediate learners had product-based and process –based teaching, respectively, results were as follows:

Table 4

Number of Sentences Used in Writings of Intermediate Learners


                                  Number of sentences

After product-based view (group C)









After process-based view (group A)









Table 4 clearly indicates that learners in group A outperformed those in C regarding the number of sentences they used in their writing (2267 versus 1980).

However, the point is that whether this outperformance can be traced in using DMs too or not is shown in Table 5.

Table 5

Number of DMs Used in Writings


                      Number of sentences

After   product-based view (group C)









After   process-based view (group A)









As it is shown, surprisingly, the intermediate learners instructed under product-based view outdid their counterparts in the other group in their total use of DMs in their writings. It is important to highlight that the increase in the number of sentences used by intermediate learners in the process-based classes did not corroborate the rise in the number of DMs used when they are compared with product-based group. Contrary to the common belief that the frequency of complex sentences would result in an increased use of discourse markers, in dealing with the intermediate level, which is the case in this study, the opposite was proved.

Also the variety of DMs used was calculated to see any improvements in using different types of DMs

Table 6

Variety of DMs Used by Intermediate Learners

Variety of DMs used





After product-view courses




After process-view courses




Table 6 shows intermediate learners in product-view courses were able to use more DMs in their samples as compared to the other group.

To see how the upper-intermediates were influenced by the treatment, we can refer to the table 7 which represents the total number of the sentences used by the participants.

Table 7

 Number of Sentences Used in Writings of Upper-Intermediate Learners


                                  Number of sentences

After product-based view (group B)









After process-based view (group D)









It is manifest from the table that among the participants of the study, those who went through a process-based approach were able to write more sentences, while this was not true of the participants who received a product-based treatment.

The number of discourse markers used by the upper-intermediate level by the two groups is shown in table 8.

Table 8

 Number of DMs Used in Writings of the upper-intermediate groups


                                  Number of sentences

After product-based view (group B)









After process-based view (group D)









The frequency of discourse markers among the upper-intermediate level students who received a process-based treatment was higher than the other group, i.e. the product-based.

To see the variety of discourse markers used by the upper-intermediate group after the treatment, the following table should be referred to.

Table 9

Variety of DMs Used by Upper-Intermediate Learners

Variety of DMs used





After product-view courses




After process-view courses




Table 9 indicated that upper-intermediate learners in process-view classes outperformed in variety the other class again. 

6. Discussion

As Kroll (1990) indicates, the ability to write well in a foreign language is a difficult and demanding task to master for EFL students because producing a successful written text requires not only the ability to control over a number of language systems, but also the ability to take into consideration the ways the discourse is shaped for a particular audience and for a particular purpose (Kroll, 2001).

Iranian learners face lots of problems in their writings due to disorganized format of compositions in Farsi or lack of instruction in educational system. Writing coherently seems to be a real challenge for students. It is of high importance to teach strategies for improving rhetorical and textual aspects of compositions instead of the sole sticking to the propositional content. Besides this, Traugott (1995) argues that “cohesion alone is not enough to make a text coherent” (P.126). Texts have an internal logic, which the reader recognizes even without the aid of explicit cohesive devices. Students need to know  that there are a number of other linguistic devices that affect the extent to which groups of sentences hold together and form a complete and cohesive text such as reference words ( e.g. pronoun reference, article reference, ellipsis etc.), lexical sets, lexical repetition, as well as conjunctions (Jalilifar, 2008).     

 One of the aims of learner training is to help learners become independent in the learning process and become more confident with writing task of language learning.Indeed, this study was carried out to see whether receiving any form of such instructions influences writing proficiency. Finding indicates that the more learners receive instruction about knowing of writing in a foreign language, the more  proficient they can become of the writing skill.

As tables 1-3 indicates proficiency level of learners is important in the awareness level of learners regarding DMs. It shows that even without explicit instruction of DMs, upper-intermediate learners unconsciously are able to apply more of them as compared to the intermediate learners. It can be due to increasing the cognitive/academic language proficiency during the time of the study. Furthermore, the linguistic knowledge of learners also improves and they get more confident in writing so this lets them write more freely which means they can also focus on expanding ideas and supporting them more to make their writing more communicative. In other words, linguistic cognitive demanding of the text decreases and it creates more attention to the function of the writing. However, learners with lower proficiency level have to focus more on the forms so it limits them to use more words and even reasonable ideas. It is also possible to know more DMs in their competence, but due to high cognitive-demanding task of writing, they are not still able to use them all.

Tables 4-6 revealed the result that training intermediate learners explicitly can increase the number of DMs they use in their writing; however, the number of sentences would increase by implicit teaching. It may first look baffling. But it seems that in intermediate level, learners are more willing to apply those strategies regardless of the quality of what they use. It also seems that learners consider DMs as some abstract forms and not as cohesive devices and in most cases they may wrongly apply them in their compositions. Therefore, the increase in number of DMs does not necessarily endorse the improvement of learners’ ability to write better, either. Then, it can be argued that learners are able to improve their writings quantitatively and not necessarily qualitatively.

Tables 7-9 reveal that, as it is expected, implicit learning of DMs led to better learning which was manifestly clear both in the number of words used and DMs. It can be discussed that since writing is more a cognitive skill than a mechanical one, cognitive development of learners during the course of instruction can make them ready to write with better quality. Also they are already familiar with DMs and when they are presented in the context, learners can understand the whole picture; therefore it is not really needed to repeat what they already know. 

This is in keeping with previous research that reported a positive impact of instruction of DMs on success in language writing (Feng, 2010). Obviously, learners who can cope with writing task, dealing with creating cohesion text, feel more successful in writing in a foreign language. This finding is further supported by former research (Traugott, 1995).

On the basis of the findings of this paper it is concluded that instruction of discourse markers can be one of the basic processes in developing the writing ability, and learners profit from it and use it in an efficient way.

The findings of this study are supported by other researchers as well. Moradan’s study (1995), investigated the effect of explicit teaching of DMs on the appropriate use of DMs by students in their writings and found that the students’ conscious awareness of forms and implications of DMs improved their appropriate use of DMs. Innajih (2007) investigated the effect of explicit instruction of DMs on the reading comprehension of the second language learners. The result showed that the treatment group performed better than the control group on the discourse cloze test. And also, Stoodt (1972, as cited in Innajih, 2007), in a cloze study, found a significant relationship between reading comprehension and the comprehension of DMs.

7. Conclusion

The results of the study indicate that within the explorative section, with the proficiency level rising, the frequency and the type of the discourse markers used subconsciously without the treatment rise as well. At the same time, the treatment and discourse marker awareness among the leaners can be of significant value in both the quality and quantity use of discourse markers. Of course, the type of the treatment must be adjusted to the proficiency level of the learners.

We may conclude from the results of this study that the level of language proficiency has a direct impact on learner awareness of the use of discourse makers.As expected, the upper-intermediate group gained more awareness concerning the discourse markers.

Finding of this study can be useful for material developers and syllabus designers and also ESP practionairs to tailor their plantings with the proficiency level of the learners. It makes it more important if it is taken into account that Iranian learners have difficulties with organized writings due to insufficient knowledge of transitors, fillers and DMs. These two approaches for teaching writing can also be helpful for teachers to improve the quality of writings by making learners aware of DMs and their applications in writings.

Abdi, R. (2002). Interpersonal meta-discourse: An indicator of interaction and identity. Discourse Studies, 4(2), 139-145.
AbdollahZadeh, E. (2006). The role of textual signals in L2 text comprehension. ESP Malaysia, 12, 1-18.
Adel, A. (2006). Meta-discourse in L1 and L2 English. Amsterdam: John Benjamins.
Blakemore, D. (1987). Semantic constrains on relevance. Oxford: Blackwell.
Blackmore, D. (2001). Discourse and relevance theory. In The handbook of discourse analysis, ed. Deborah Schiffrin, Deborah Tannen, and Heidi E. Hamilton, 100-115. Malden/ Oxford: Blackwell.
Blagojevic, S. (2003). Meta-discourse in academic prose: A contrastive study of academic articles written in English by English and Norwegian native speakers. ERIC Document Reproduction Service, No. ED424356.
Brinton, L. J. (1996). Pragmatic markers in English: Grammaticalization and discourse function. Berlin/New York: Mouton de Gruyter.
Bunton, D. (1999). The use of higher level meta-text in PhD theses. English for Specific Purposes. 18 (1), S41–S56.
Connor, U. (1984). A study of cohesion and coherence in English as a second language students’ writing. Papers in Linguistics: International Journal of Human Communication, 17(3), 301-316.
Dailey-O’Cain, J. (2000). The sociolinguistic distribution of and attitudes toward focuser like and quotative like. Journal of Sociolinguistics, 4, 60–80.
Feng, L. (2010). Discourse Markers in English Writing. Journal of International Social Research, Vol. 3/11, spring 2010. [Online] Available: http://www.sosyalarastimalar.com
Field, Y. & Yip, L. M. O.   (1992). A comparison of internal conjunctive cohesion in the English essay writing of Cantonese speakers and native speakers of English. RELC Journal, 23(1), 15-28.
Fraser, B. (1990). An approach to discourse markers. Journal of Pragmatics 14: 383- 395.
Grimes, J. E. (1975). The thread of discourse. New York: Mouton Publishers.
Haliday, M.A.K. (2004). An introduction to functional grammar. 3rd edition. Revised by Christian M.I.M. Matthiessen. London: Hodder Arnold.
Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. London: Longman.
Hatim, B. (1997). Communication across cultures: Translation theory and contrastive text linguistics. Exeter: University of Exeter Press.
Hays, P.R., (1992). Discourse markers and L2 acquisition. Papers in Applied Linguistics-Michigan, 7, 24–34.
Hoey, M. (2001). Textual interaction: An introduction to written discourse analysis. New York: Routledge.
Karasi, M. (1994). Cohesive features in the expository essays of secondary four (Express) and secondary five (Normal) students in Singapore. Unpublished M.A. thesis. Singapore: Nanyang Technological University.
Innajih, A. (2007). The effect of conjunctive types on the English language reading comprehension of Libyan university students. [Online] Available: www.ecls.ac.uk/publish/text/the Effect of Conjunctive/
Intarparawat, P. & Steffensen, M. S. (1995). The use of meta-discourse in good and poor ESL essays. Journal of Second Language Writing, 4, 253-272.
Jalilifar, A. R. (2008). Discourse Markers in Composition Writings: The Case of Iranian   Learners of English as a Foreign Language. Journal of CCSE, English Language Teaching, 1(2).
Jalilifar, A. R., & Alipour, M. (2007). How explicit instruction makes a differences: Metadiscourse markers and EFL learners’ reading comprehension skill. Journal of College Reading and Learning, 38(1), 35-52.
Johns, A. M. (1984). Textual cohesion and the Chinese speakers of English. Language Learning and Communication. 3(1), 69-73.
Johnson, P. (1992). Cohesion and coherence in compositions in Mayland English language. RELC Journal, 23(2), 1-17.
Kroon, C. (1997). Discourse markers, discourse structure and functional grammar. In Discourse and pragmatics in functional grammar, ed. John H. Connolly, Roel M. Vismans, Christopher S. Butler, and Richard A. Gatward, 17-32. Berlin/New York: Mouton de Gruyter.
Lee, I. (1999). The acquisition of colloquial features by Korean Americans. Journal of the Pan-Pacific Association of Applied Linguistics, 3, 71–87.
Lenk, U. (1998). Marking discourse coherence: Functions of discourse markers in spoken English. Gunter Narr Verlag: Tübingen.
Longo, B. (1994). Current research in technical communication: The role of meta-discourse in persuasion. Technical Communication, 4(2), 348-352.
Longrace, B. E. (1996). The grammar of discourse. London: Plenum Press.
Longrcae, B. E. (1995). Left shifts in strongly VSO languages. In Word order in discourse, ed. Pamela Downing, and Michael Noonan, 331-354. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Matras, Y. (1997). The function and typology of coordinating conjunctions: Evidence from discourse and language-contact situations. In Discourse and pragmatics in functional grammar, ed. John H. Connolly, Roel M. Vismans, Christopher S. Butler, and Richard A. Gatward, 177-191. Berlin/New York: Mouton de Gruyter.
Mauranen, A. (1993). Contrastive ESP rhetoric: Meta-text in Finnish-English economic texts. English for Specific Purposes, 12, 89-103.
Moradan, A. (1995). Significance of conjunctions as a cohesive device in teaching writing. Unpublished MA thesis. Allameh Tabatabai University.
Morgan, Jerry L., and Sellner, M. B. (1980). Discourse and linguistic theory. In Theoretical issues in reading comprehension: Perspective from cognitive psychology, linguistics, artificial intelligence, and education, ed. Rand j. Spiro, Bertram C. Bruce, and William F. Brewer, 163-200. New Jersey: Lawrence Erlbaum Associates, Inc.
Muller, S. (2004). Well you know that type of person: Functions of well in the speech of American and German students. Journal of Pragmatics, 36, 1157–1182.
Norment, N. (1994). Contrastive analyses of cohesive devices in Chinese and Chinese ESL in narrative and expository written texts. Chinese Language Teaching Association Journal, 29, 1, 49-81.
Nunan, D. (1991) Language Teaching Methodology. London: Prince Hall.
Osman, J. (1995). Pragmatic particles twenty years after. In Organization in discourse: Proceedings from the Turku conference, ed. Brita Wårvik, Sanna-Kaisa Tanskanen, and Risto Hiltunen. Turku, 95-108. Finland: University of Turku.
Perez, M. A. & Macia, E. A. (2002). Meta-discourse in lecture comprehension: Does it really help foreign language learners. Atlantis, 14(2), 3-21.
Sarig, L. (1995). Discourse Markers in contemporary Arabic. Zeitschrift für Arabische Linguistik 30: 7-21.
 Schiffrin, D. (1987). Discourse markers. Cambridge: Cambridge University Press.
Schourup, L. (1999). Discourse markers: Tutorial overview. Lingua 107: 227-265.
Siepmann, D. (2005). Discourse markers across languages: A contrastive study of second-level discourse markers in native and non-native text with implications for general and pedagogic lexicography. London/New York: Routledge.
Steffensen, M. S. & Cheng, X. (1996). Meta-discourse and text pragmatics: How students write after learning about meta-discourse. ERIC Document Reproduction Service, No. ED400709.
Traugott, E.C. (1995). Subjetification in grammaticalization. In D, Stein & S, Wright, (Eds.) Subjectivity and subjectivization: Linguistics perspectives (pp.31-54). Cambridge University press.
Trillo, J. R., (2002). The pragmatic fossilization of discourse markers in non-native speakers of English. Journal of Pragmatics, 34, 769–784.
Ventola, E. & Mauranen, A. (1993). Non-native writing and native revising of scientific articles. ERIC Document Reproduction Service, No. ED352149.