Discourse markers in EFL setting: Perceptions of Turkish EFL teachers

Discourse markers are seen as one of the fundamental units in spoken discourse due to their frequent and multifunctional use by native speakers of English. Discourse markers also have significance in foreign language instruction. In this respect, this study explored the perceptions of Turkish EFL teachers towards the use of discourse markers in terms of the pragmatic and pedagogic value and their representations in EFL classrooms. To this end, a questionnaire was administered to 104 EFL instructors working at seven different state universities in Turkey. The quantitative analyses of the data yielded that the teachers had positive perceptions towards discourse markers by finding them necessary for improving pragmatic competence of language learners. They also stated that discourse markers had teaching value so they highlighted them in their instruction. Moreover, the study also discussed the integration of discourse markers into language instruction. The study had some implications for further discussion over the issue of discourse markers in terms of investigating teacher talk, real classroom practices during language teaching instruction and teaching materials.


INTRODUCTION
Discourse markers (henceforth DM) are important features of spoken language in terms of their frequency and multi-functionality in spoken discourse.They are commonly preferred discourse items used by particularly native speakers of English.Due to their significance as 'sharing devices and intimacy signals in our everyday talk' (Quirk et al., 1985, p.179), their pragmatic and indispensable value is salient in spoken discourse.
DMs are defined as "sequentially dependent elements which bracket units of talk" (Schiffrin, 1987, p.31).In other words, DMs are seen as textual and cohesive elements in discourse by depending on the previous or following units.Thus, connectivity is one of the fundamental features of DMs as they signal relationships between the utterances.Moreover, DMs can exist in several grammatical classes; verbs (look, listen, see), conjunctions (and, but, nevertheless) or non-finite clauses (you know, i mean, to be honest).Aijmer (2002) points out that they should be studied from functional and pragmatic perspective rather than grammatical classifications.
According to Louwerse and Mitchell (2003), DMs mostly occur in spoken rather than written discourse.For example, Carter and McCarthy (2006)  common DMs in everyday spoken language which are single words such as anyway, cos, fine, good, great, like, now, okay, right, so, well and phrasal and clausal items such as you know, I mean, as I say, for a start, mind you.Furthermore, DMs are also known as optional elements in spoken discourse.Their removal from the utterance may not change the grammaticality of the utterance.They do not influence the truth condition of the proposition (Schourup, 1999).In other words, they do not contribute to the content of the utterance.However, these criteria do not make them unnecessary elements in an utterance.They are peculiar as they "reflect choices of monitoring, organisation and management exercised by the speaker" (Carter and McCarthy, 2006, p. 208).
DMs are multi-functional and multi-categorical elements.Fung and Carter (2007) gather the functions of DMs in four categories: interpersonal, referential, structural and cognitive.As seen in Table 1, classifying a DM into only one functional category is difficult.Speakers may use these items for several purposes.For example, you know can be used not only for marking shared knowledge between the speaker and the listener but also for assessing the listener's knowledge about the utterances (Fung and Carter, 2007, p.418).
Due to their significance in native discourse, DMs have been of interest in second/foreign language teaching.They have been seen as necessary items in spoken discourse to promote communicative and pragmatic competence of the speakers (Hellermann and Vergun, 2007;Lam, 2009;Müller, 2005;O'Keeffe et al., 2007;Wierzbicka, 1991).Trillo (2002) states that pragmatic distance is not displayed to non-native speakers of English.According to Trillo (2002), pragmatic distance is defined as "the variants in the social, cognitive and contextual dimensions of linguistic communication that govern and systematize social relations in speech" (p.771).Thus, the level of exposure of non-native speakers to pragmatic items in discourse may limit their naturalness in the target language, especially in EFL setting.Non-native speakers also use DMs in their own language significantly, but they do not feel independent and confident when using DMs in the target language.This inability to use DMs in their spoken interaction may make the speakers "potentially disempowered and at risk of becoming a second-class participant' (O' Keeffe et al., 2007, p.39).Wierzbicka (1991) states, from a pedagogical perspective, failing to master the use of discourse particles may seriously impair the communicative competence of learners.
Furthermore, Trillo's comparative study (2002) investigates the use of pragmatic markers in speech of native and non-native speakers of English, in the case of children and adults.It concludes that if there is no consistent teaching of pragmatic markers in language instruction, the use of pragmatic markers becomes fossilized in quantity and the diversity.Similarly, Müller's studies (2004Müller's studies ( , 2005) ) point out that there are fewer and different functions of DMs used by German speakers of English when compared to the native data.Moreover, Fung and Carter's (2007) study compares a pedagogic sub-corpus of spoken British English and a corpus of classroom discussions in Hong Kong secondary schools.Thus, they find out that native speakers use DMs for a wider variety of pragmatic functions when compared to Hong Kong learners of English who used yeah, really, say, sort of, I see, you see, well, right, actually in relatively restricted sense.
How DMs should be taught or promoted in language classrooms have also been discussed and searched.In this sense, the role of teaching materials and the teacher in a language classroom are significant.Lam's (2009) study on investigating well in 15 textbooks concludes that none of the textbooks has a separate section to describe well for improving oral skills.This study reveals that textbooks lack a detailed description and presentation of DMs as used in real-life contexts.Thus, "such textbooks are so detached from reality that they have ultimately lost their pedagogical value" (Lam, 2009, p.277).Sert and Seedhouse's (2011) review on conversation analysis also discusses the authenticity of dialogues in language teaching materials.They suggest that although each ordinary conversation may not always be the best material in teaching, the textbooks should be able to include naturally occurring talk with the examples of scripted conversations.Hellermann and Vergun's (2007) study on the DMs well, you know, like which are not explicitly taught at beginning to adult learners of English also suggest some implications.They conclude that the teachers in language classrooms play a fundamental role in promoting DMs to language learners.They suggest that although the teachers do not need to spend a particular time on DMs, learners should be made aware of DMs and their pragmatic functions.Thus, it has been discussed that the DMs should be taught explicitly or implicitly in language classrooms by the teacher.Moreover, teacher talk during instruction may also be guiding while learners are not focusing on these items.In this sense, the attitudes of the teachers towards DMs need attention.Fung's (2011) study investigates the attitudes of 132 Hong Kong teachers working in secondary schools towards the use of DMs in their classrooms.This study reveals that although the teachers have a certain positive sense of DMs in terms of pragmatic and pedagogic value, they state that DMs are underrepresented in materials and instruction.
As a result, the literature focusing on DMs in several aspects has been increasing and diversifying recently.Previous research has addressed several aspects of DMs such as DMs used by native speakers of English and their functions (Aijmer, 1987;Erman, 1987;Schiffrin, 1986;Schourup, 1999) and DMs in different languages rather than English (Bazzanella and Morra, 1990;Chen and He, 2001).Although the interest towards the pedagogical significance of DMs in ESL/EFL classrooms has been increasing (Fung and Carter, 2007;Lam, 2009;Müller, 2004;Trillo 2002), little attention has been given to the perceptions of teachers towards DMs (Fung, 2011).However, teachers' beliefs can provide insights into the pedagogic practice of DMs.While the performance of language learners in using DMs and their awareness has been studied mostly, awareness, experience and viewpoints of EFL teachers about DMs should also be studied.Therefore, the present research attempts to explore the issue in the Turkish context by investigating the perceptions of Turkish EFL teachers towards the use of DMs in EFL classrooms.The questions guiding the research were as follows: 1. How do Turkish EFL teachers perceive DMs in terms of their pedagogic and pragmatic value? 2. What do Turkish EFL teachers think about how DMs Aşik 943 should be represented in the classrooms? 3. How do Turkish EFL teachers perceive the current representation of DMs in language classrooms?

Participants
Participants were selected via convenience sampling (Dörnyei, 2007).They were 104 EFL instructors teaching students of different levels of language proficiency at English preparatory programs from seven different large state universities in four different cities of Turkey.Instructors were invited to fill in the online questionnaire by email.Thus, the teachers participated in the study on a voluntary basis.As for the demographics of teachers, 75% were females while 25% of the teachers were males.In terms of the qualifications, 79% of the teachers held a BA degree in English Language Teaching, 16 % of them a BA in English or American literature, 4% a BA in translating and interpreting, and only 1 % a BA in linguistics.56% of the instructors held MA while 10 % held PhD degree in TEFL.As far as the years of experience in teaching are concerned, 34% of the respondents had 1-4 years, 23% of them had five to nine years and 43% of them had over 10 years of teaching experience.Regarding the teaching hours per week, 32% of them give English courses between 12-15 h weekly, 12 %, between 15-18 h; 56%, over 18 teaching h.Briefly, the participants were predominantly females, the graduates of English Language Teaching, qualified and mostly experienced teachers of English.

Data collection
The data for the research objectives were collected through an online questionnaire adapted from Fung (2011).Permission to use the instrument was granted by the corresponding author via e-mail.
The questionnaire consists of 48 items including the linguistic, pedagogic, cultural and pragmatic aspects of the use of DMs.The questionnaire was adapted by the researcher along with the viewpoints of 5 experts with PhD holder in ELT.In line with their feedback, some of the statements were clarified while some were re-arranged according to Turkish EFL context.The questionnaire is a five-point Likert scale which is helpful in terms of revealing teachers' perceptions.Likert scales are used for measuring people's attitudes for social issues (Busch, 1993) and how respondents feel about a number of statements (Brown, 2001).The scales included have the continuum as as follows: 1: Strongly agree; 2: Agree; 3: Uncertain; 4: Disagree; 5: Strongly Disagree.Some of the statements in the questionnaire were positively worded while some were negatively worded to counter-check the reliability of the statements.

Data analysis
The study employed two methods of quantitative analysis on the questionnaire through SPSSv 15.0 for Windows: reliability analysis and (confirmatory) factor analysis.For the reliability analysis, Cronbach's Alpha was calculated as 0.789 which displays that the questionnaire has moderately acceptable reliability.Moreover, an item-total correlation test was performed to check if any item in the set of tests is inconsistent with the averaged behaviour of the others, and thus can be discarded.This analysis displayed that no item had value smaller than .30,which displays that the list of the items or items were reliable.
Factor analysis was done to identify the internal consistency of the questionnaire.Factor analysis is conducted to "reduce the number of variables submitted to the analysis to a few values that will still contain most of the information found in the original variables" (Dörnyei, 2007, p.233).After the first Component Factor Analysis, item 9 was excluded from the analysis as its factor load of .22 was below 0.40, the accepted value for correlation between the factor and item.The KMO (Kaiser-Meyer-Olkin) measure of sampling adequacy was .686,above the recommended value of .6 and Bartlett's test of sphericity was significant (p = 0,000 /  2 = 2464.716;p<0.01).
The results of the factor analysis revealed that seven factor loadings were rotated through the responses of the teachers.Appendix A shows the factor loading matrix including the factor loadings of all variables on each factor.The factor loading matrix displays similarities with Fung's (2011) results.Similar items were mostly represented under the same factor.Some of the factors were named similarly with Fung's (2011) study while some were renamed according to the adapted versions of the items.Each factor is presented with the relevant items in the corresponding tables.

RQ1. How do Turkish EFL teachers perceive DMs in terms of their pedagogic and pragmatic value?
Factor 1, 2 and 3 display how Turkish EFL teachers perceive DMs in terms of pedagogic and pragmatic value.The items under Factor 1, named as the pragmatic value of DMs, investigate the attitudes of the teachers towards the DMs in terms of their communicative and pragmatic competence aspect, stated in Table 2.
These items correspond with the multi-functionality of DMs in spoken discourse; such as turn-taking, responding, listening and cognitive processing in interactional exchanges.Moreover, the items (26, 28, 29) clearly display that how knowledge of DMs help and facilitate language learners in real life; such as in future workplaces, in future exams which include testing listening proficiency and following a university lecture.The results of the means (1.55-1.93)and the small magnitude of the standard deviations (0.71-0.83) indicate that the teachers have considerably positive perceptions about the pragmatic value of DMs.
Furthermore, Factor 2, in Table 3, deals with mainly the optionality and non-truth conditionality characteristic of DMs in spoken discourse.These items question whether DMs are optional and influence the truth conditions of the utterances or not.
There are different viewpoints about the dispensable value of DMs.As items 4, 5 and 8 reflect, the teachers accept that DMs are useful and necessary devices to assist listeners to comprehend the conversation, to have coherence and to relate the ideas in talk.On the other hand, they also accept the subsidiary role of DMs in spoken discourse.Since the results of the items 6, 10, 13 and 14 indicate, they have a tendency to accept that the conversation can still be understood by means of other linguistic clues such as grammar and vocabulary without DMs.Moreover, item 19 also reflects that the teachers accept that the use of DMs by the students in spoken language is fairly less when compared to the use of discourse connectives (namely conjunctions) in their written discourse.
Factor 3 in Table 4 presents the items related to the pedagogic value of DMs in terms of their significance to teach and learn during language teaching and learning process.Particularly, according to the items 23, 24 and 25 (mean: 4.19-4.25),the teachers in the study accept that DMs have teaching value.They disagree that DMs are small, redundant and meaningless words.
Moreover, they agree that linguistic awareness about DMs should be developed (item 20) and the language learners should have proficiency of DMs to improve their speaking and listening skills (items 22 and 27).These results imply that the teachers have positive attitudes towards teaching DMs in language classrooms.However, they have uncertain attitudes towards whether to leave the idea of learning DMs in the future on their own to the students (item 38, mean: 3.28).They are also not clear about the level that the students should be able to speak DMs as native speakers do.

RQ2. What do Turkish EFL teachers think about how DMs should be represented in the classrooms?
Factor 4, in Table 5, presents the items including the issues regarding the attitudes of the teachers towards the use of DMs by native speakers of English and its reflection in language classrooms.Although the teachers agree that DMs to be taught as how native speakers use (item 30, mean: 1.83), they are not certain that being competent speakers of language mean speaking like its natives.The teachers are also unsure about that their students will be able to use DMs as native speakers do (item 40, mean: 3.45).Moreover, items 42 and 46 (mean: 3.10-3.04)reveal that there is not a clear consensus among the teachers on the issue which of DMs should be a model; American or British way.Furthermore, Factor 5 which has loadings from the items 43-47 reflects the opinions about which norms of DMs should be accepted by Turkish non-native speakers of English.According to the results of the items 45-47, the teachers mostly agree that different varieties of DMs should be recognized and accepted rather than sticking only to the native speaker norm of using DMs.The findings suggest that there should be acceptance of the local usages of DMs.The teachers also agree that exposing students to different varieties of DMs may be necessary to improve their comprehension skills.However, regarding the acceptance of Turkish style of using DMs, they are uncertain about it as it is reflected in items 43 and 44 (Table 6).
Regarding the issue of how and when DMs should be highlighted in EFL classrooms, Factor 6 (items 32-37) displays the opinions about the proficiency levels and receptive or productive skills.According to the results of the item 32, the teachers agree that pre-intermediate and intermediate levels are appropriate time to highlight DMs in spoken text when compared to upper-intermediate and advanced levels (item 33, mean: 2.94).On the other hand, they were nearly moderate about expecting students at pre-intermediate and intermediate levels to learn DMs for both listening and speaking purposes (Table 7).
Moreover, the results of the items 35 and 37 reveal that they were also uncertain about prioritizing DMs for receptive purposes at pre-intermediate and intermediate levels.The results of the item 36 are consistent with the ones of 34, 35 and 37 by reflecting that the teachers agree on the promotion of DMs for both receptive and productive purposes from the pre-intermediate and intermediate levels.

RQ3. How do Turkish EFL teachers perceive the current representation of DMs in language classrooms?
Factor 7 investigates whether the teachers and the course materials present DMs in EFL classrooms.The results indicate that the teachers accept the representtation of DMs in the oral materials during speaking and listening skill activities.The teachers also accept that they highlight DMs in speaking and listening activities (items 17 and 18, mean: 2.13 and 2.17) (Table 8).
As an overall evaluation of the factors specified, Table 9 displays the summary of the seven factors with their overall mean and standard deviations.Table 9 shows that Turkish EFL teachers had positive attitudes towards the pragmatic competence, dispensable value of DMs and pedagogic value of DMs.However, there are relatively neutral attitudes regarding the norm of DMs to be taught or used and how DMs should be included in language classrooms.Moreover, the highest standard deviations in Factor 4 (SD: 0.76) and 6 (SD: 0.74) show the diverse opinions regarding the representation of DMs in classrooms and the norm issue of DMs to be used or learned.

DISCUSSION
Data analysis of the questionnaire displays a detailed identification of the perceptions of Turkish EFL teachers towards DMs in several aspects.In terms of the pragmatic value of DMs in discourse, the findings suggest that the teachers agree on the contribution of DMs to communicative and pragmatic competence of their students.Moreover, the findings reveal that DMs are found significant in language classrooms, e.g. for comprehension of the lectures.In this sense, Chaudron and Richards (1986) report that overusing micro-markers (so, actually, well, right) may be distracting within the overall coherence of the lecture while macro-markers (to begin with, this meant that and in this way) increase the level of comprehension of lectures.However, the teachers agree on the contribution of DMs to future interactions as well.The positive perceptions of the teachers are in consistent with the results of Fung (2011) and several studies which discussed the pragmatic value of DMs in spoken discourse such as Fung and Carter (2007), Lam (2009) and Trillo (2002).
Furthermore, the results indicate that there are inconsistent views on the issue of dispensable value of DMs in conversation.Although the teachers accept that DMs contribute a lot to overall conversation, they are uncertain about the dispensable component of DMs.The teachers also accept that students use discourse connectives in written discourse compared to the use of DMs in spoken language.Yet, DMs have a characteristic of orality.Louwerse and Mitchell's (2003) research supports this characteristic of DMs by concluding that the DMs in spoken discourse are 10 times more than that in written discourse.Moreover, the studies about the use of DMs in non-native spoken discourse (Aşık and Cephe, 2013;Buysse 2010;Müller, 2005;Fung and Carter, 2007) reveal that the types of the DMs used by non-native language learners are mostly textual featured DMs probably because of a kind of transfer from their experiences in written English discourse.For instance, Buysse (2010) reports that when compared with their native speaker peers foreign language learners use hardly interpersonal DMs such as you know, like, kind of/sort of and I mean.The reasons can be various in this respect, such as students' engagement mostly to the language of the written texts in English rather than spoken interactions in classrooms or lack of activities to raise awareness on the use of DMs.
In terms of the pedagogic value of DMs, there is a strong consensus stated by the teachers.The perceptions of the teachers towards both the pragmatic and pedagogic value of DMs are consistent.However, the teachers reflect a kind of ambivalent certainty of their implementation in the classroom.They are unsure about the implementation of DMs in teaching in terms of the model, local usage, levels of proficiency and receptive or productive skills.Thus, it can be concluded that there is a need to raise awareness of the EFL teachers about how DMs should be taught in language classrooms.Literature on DMs reports several implications.The pedagogical implications may be summarised as implicit or explicit teaching, inductive or deductive method, noticing application or output-based application (Jones, 2011).Jones's (2011) study points out that both noticing (language awareness) and output-based application (presentation and practice) have clear influence on teaching DMs.Similarly, Wichmann and Chane (2009) also support the idea that learners' awareness towards these pragmatic items should be increased as the learners may not understand the real functions of DMs in spoken interaction.
The findings also raise these questions: Should the native speaker model be the ideal one for language learners?or Can a non-native speaker become a native speaker?.In this aspect, the results are in consistent with Lee (2005) and Andreou and Galamantos (2009).They claim that during language teaching, it is necessary to be realistic about the expectations.They suggest that the learners should be treated as what they are rather than potential native speakers.Moreover, Fung and Carter (2007) suggest that in teaching DMs to non-native speakers, the aim should be to help them be competent speakers in the target language, feel secure in it similar to the easiness in their native language.Moreover, House's (2013) study points out that ELF (English as a lingua franca) users use DMs (yes/yeah, so and okay) effectively and differently from native speakers by reinterpreting them both for their own and their interlocutors.
Furthermore, the results show that there is a consistency between the perceived value of DMs and the actual representation by the teachers.The teachers accept that DMs are represented in language classrooms in coursebooks and their classrooms.However, in Fung's (2011) study, there is a large gap between two in Hong-Kong context.Thus, in Turkish EFL context, the findings yield that the teachers are aware of the significant values of DMs and they accept that they highlight them in speaking and listening classes.This result is hopeful in terms of the use of DMs in EFL classrooms.But it should also be noted that the results can be peculiar to the context of the study, which is university context.Moreover, these four items and the study itself cannot signify alone that DMs are represented in EFL classrooms by the teachers and course materials.Since this kind of investigation is not the main focus in the scope of the study, the findings can be interpreted as only guiding, rather than concrete outputs.

Conclusion
This study aimed at investigating the perceptions of Turkish EFL teachers towards the use of DMs in the EFL classrooms.Particularly, the research objectives were to search how Turkish EFL teachers perceive DMs in terms of its pedagogic, pragmatic value of DMs and their current representations in classrooms and to find out their opinions about how DMs should represented in the classrooms.With the help of a questionnaire adapted from Fung (2011), the study collected the data from 104 Turkish EFL instructors.The findings revealed seven particular components: pragmatic value of DMs; dispensable value of DMs; pedagogic value of DMs; identification with the native speaker norm; acceptance of the local usage; representation of DMs in EFL Classrooms and how DMs should be highlighted in EFL classrooms.
The findings revealed that Turkish EFL instructors had positive perceptions about the pragmatic and pedagogic value of DMs.They also accepted that DMs were represented in classrooms by themselves and materials.It is really hopeful that the suggestion and implications postulated by the researchers in the field are also supported by Turkish EFL teachers in terms of the pragmatic and pedagogic value of DMs.However, the study also concluded that the issue of norm of DMs, whether local usage, American or British usage or other regional usages should be promoted, and how and when DMs should be included in the curriculum, remained ambivalent.
Lastly, representation of DMs in the classroom was not the direct main focus of this study.But it is hoped that this study may contribute to the field by providing perceptions of Turkish EFL teachers, which can be guiding for future research.Further research can be done regarding the aspects which are investigating teachers' knowledge of DMs, actual classroom practices, teacher talk, methodologies on how DMs should be included and their representation in coursebooks.It should be noted that this study has some limitations.The discussion of the conclusions drawn from the findings were restricted due to the lack of replication in other EFL contexts.This study is limited in terms of its sampling, which is a university context.Thus, apart from this academic context, the perceptions of the teachers working at high schools or other institutions need to be studied.
Appendix A. Factor Loadings of the Questionnaire.

Variant
Factor list the most E-mail: asuman.asik@gmail.com.Authors agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License

Table 5 .
Factor 4: Identification with the native speaker norm.Students should be taught how native speakers use DMs and follow their way of using them.

Table 6 .
Factor 5: Acceptance of the local usage.It is not necessary to stick to the native speaker norm of using DMs because English language teaching should seek relevance to local culture while trying to enable global transaction.

Table 7 .
Factor 6: How DMs should be highlighted in EFL classrooms.

Table 8 .
Factor 7: Representation of DMs in EFL classrooms.

Table 9 .
Summary of the perceptions of EFL teachers towards DMs with seven factors.