(Im)politeness strategies and use of discourse markers

Abstract: This study aimed to investigate the L2 learners’, EFL teachers’, and American native speakers’ use of discourse markers as hedging devices to mitigate face-threatening acts considering gender, proficiency level, and control–experimental variables. It used open discourse role-play tasks, a self-assessment report of English competence, as well as a seven-scenario questionnaire with a five-point Likert scale and without it for L2 learners to translate into Persian. To this end, three groups of participants took part in the current study: (a) 8 groups of 20 L2 learners; (b) 90 participants (i.e. 30 L2 learners, 30 EFL teachers, and 30 native speakers); and (c) 150 Iranian advanced L2 learners. The results revealed that native speakers significantly surpassed EFL teachers and L2 learners in employing DMs and that instruction and proficiency level played a significant role in L2 learners’ use of DMs. The findings also substantiated that female L2 learners outperformed their male counterparts in using approximators, modals, and passives significantly. Furthermore, based on MAXQDA software, two areas of discrepancies, namely “precision” and “direct reasoning” in Persian versus “approximators” and “indefinites” in English, were found to delve into the subtleties between the two cultures. *Corresponding author: Hamid Allami, Department of Foreign Languages, Yazd University, Yazd, Iran E-mail: hallami@yazd.ac.ir


PUBLIC INTEREST STATEMENT
Since the world is significantly growing smaller, learners of English encounter more pragmatic breakdowns which hinder intercultural communication. As an aspect of pragmatic competence, discourse markers can be used to mark politeness to soften the force of commands. To investigate the pragmatic competence of Iranian learners of English, their teachers, American native speakers, and Iranian native speakers, this study used open discourse role-play tasks and a questionnaire. This perspective article found that women present a higher pragmatic competence in terms of discourse markers to mitigate commands than men do. Also, it indicated that the American native speakers tend to use more discourse markers followed by L2 teachers and their students. In addition, two areas of discrepancies, namely "precision" and "direct reasoning" in Persian versus "approximators" and "indefinites" in English, are found to delve into the subtleties between the two cultures. Such findings offer implications for EFL practitioners and material designers.

Introduction
Not only is defining politeness an elusive phenomenon, but also interpreting impoliteness is problematic since (im)politeness phenomena are not considered as isolated phrases or sentences and are not inherent in the words used (Bargiela-Chiappini, 2003). Since the world is significantly growing smaller, second Language (L2) learners encounter more pragmatic breakdowns which hinder intercultural communication (Bardovi-Harlig & Dörnyei, 1998;Byram, 1997;Kádár & Bargiela-Chiappini, 2010;Meier, 1995;Mohammadi & Tamimi Sa'd, 2014;Mugford, 2008;White, 1993). L2 learners then should be instructed to sound appropriate in different speech events with varying contextual features based on sociocultural assumptions. In this regard, they should be empowered to decide how to react in different (im)polite everyday realities and to understand the consequences of their being (im)polite (Mugford, 2008). Politeness and impoliteness, as suggested by Mills (2009), are linguistically and culturally relative to L2 learners to sound appropriate in different language encounters.
Politeness approaches which emphasize the role of context are called "discursive" or "postmodern" (Culpeper, 2010;Locher, 2006;Mills, 2003). Such approaches are against universalizing generalizations focusing on "the participants' situated and dynamic evaluations of politeness, not shared conventionalised politeness forms or strategies" (Culpeper, 2010, p. 3235). As Haugh (2007) aptly puts it, "the discursive approach abandons the pursuit of not only an a priori predictive theory of politeness, but also any attempts to develop a universal, cross-culturally valid theory of politeness altogether" (2007, p. 297). Therefore, there is an urgent need to analyze the unspoken rules of discourse which imply that certain utterances are manifested as appropriate to show reality and social norms (Mills, 2011). In order to adopt a discursive approach toward analyzing politeness, researchers should delve into longer stretches of interaction which focus on the judgment issues of (im)politeness realized as a resource by the participants.
As an aspect of pragmatic competence, vague language may serve as a politeness strategy to soften the force of commands with the purpose of saving face (Boncea, 2014). Discourse markers (DMs) can be used as hedging devices in a certain context to add to the vagueness of an utterance which can be used to mark politeness and mitigate face-threatening acts (FTAs) (Boncea, 2014;Yates, 2010). In this regard, the appropriate use of DMs as hedging devices is regarded as a considerable challenge for L2 learners at different proficiency levels since they may not be familiar with how to make their language fuzzier to achieve communicative goals (Fraser, 2010).

Research aims
In an attempt to allow a deeper insight into DMs as hedging devices, the present study aims to contribute the research on hedge use by Iranian L2 learners and to fulfill several gaps in the previous studies. This study delves into investigating the DMs as hedging mechanisms by intermediate and advanced Iranian L2 learners in experimental and control groups to find the frequency of DMs in their speech. It, also, includes comparison data from American and Iranian Native Speakers (NSs). This study attempts to build on previous research in finding the nuances between Iranian L2 learners', EFL teachers', and American NSs' discourse in terms of DMs to soften the force of commands. Moreover, it examines whether Iranian NSs use different types of hedges compared to Iranian L2 learners. Last but not least, the current research tries to scrutinize gender differences among Iranian L2 learners in using DMs as hedging devices.

Review of related literature
Earlier studies on politeness have mostly focused on the traditional Brown and Levinson's (1987) framework to measure politeness using three factors of social distance, relative power, and absolute ranking of impositions as perceived by the interlocutors. Nevertheless, (im)politeness conventions vary from one culture to another leaving one-theory-fits-all inapplicable to all situations in which language is implicated (Mills, 2009). Thus, as Hsieh (2009) shrewdly observes, "[c]onfined politeness theory cannot adequately explain the various kinds of human interaction" (2009, p. 56). He, also, argues that a "more contextualized investigation is needed in order to gain a more comprehensive understanding of what (im)politeness is" (Hsieh, 2009, p. 56).
(Im)politeness often varies across people with different cultural backgrounds. Yule (1999) believes that people can be polite through "being tactful, modest and nice to other people" (1999, p. 134). Also, Lakoff (1989) defines politeness as "a means of minimizing confrontation in discourse-both the possibility of confrontation occurring at all, and the possibility that a confrontation will be perceived as threatening" (1989, p. 102). Another definition of politeness was put forward by Leech (1983) as "maintain[ing] the social equilibrium and the friendly relations which enable us to assume that our interlocutors are being cooperative in the first place (1983, p. 82). Watts (2003) holds that politeness should be defined through discursive approach. He believes that such struggle defines "the ways in which (im)polite behaviour is evaluated and commented on by lay members and not with ways in which social scientists lift the term '(im)politeness' out of the realm of everyday discourse and evaluate it to the status of a theoretical concept" (Watts, 2003, p. 9). Although discursive practice seems messier than Brown and Levinson's (1987) universal framework, the analysis has proven to probe into the subtleties of culturally situated communicative behavior (Watts, 2003).
To Fraser (2010), hedging is an aspect of pragmatic competence. He defines it as "a rhetorical strategy, by which a speaker, using a linguistic device, can signal a lack of commitment to either the full semantic membership of an expression, … or the full commitment to the force of the speech act being conveyed" (Fraser, 2010, p. 22). Also, he holds that since hedging is drawn from every syntactic category, there exists no grammatical class of hedges. Fraser (2010) provides the following list of hedging devices: Adverbs, adjectives, impersonal pronouns, concessive conjunctions, indirect speech acts, introductory phrases, modal adverbs, modal adjectives, hedged performatives, modal nouns, modal verbs, epistemic verbs, negation, tag questions, agentless passives, parenthetic constructions, if clauses, progressive forms, tentative inference, hypothetical past, metalinguistic comments, etc. (Fraser, 2010, p. 22) Nikola (1997) discusses hedging as "a strategy which renders speakers' messages more tentative and vague, and thus reduces the force of what they are saying" (1997, p. 190). She holds that hedging achieves consequential interpersonal functions for L2 learners who might deliberately be considered as impolite or rude due to insufficient pragmatic skills. She claims that such lack of sufficient pragmatic knowledge may culminate in pragmatic failure. In this respect, hedges may compensate for the speaker to minimize the imposition or directness of his or her utterance (Wilamová, 2005).
A large number of studies have been carried out on the interlanguage development of DMs among L2 learners (Aşik & Cephe, 2013;Jalilifar et al., 2011;Neary-Sundquist, 2013;Siu, 2014;Yu, 2009). The types and functions of hedging devices were examined in these research studies to demonstrate whether the L2 learners used DMs appropriately. Aijmer (2002) suggests that a description of DMs be provided at various interlanguage stages of L2 learners of English. She holds that DMs carry interpersonal functions in everyday conversation; hence, face-saving, politeness, as well as indirectness are relevant in the usage of DMs. Aijmer (2002) believes that DMs such as tag questions and approximators cause the reduction of social distance between the speaker and the hearer adding to the politeness load of a given utterance. The following studies underscore the language learners' use of DMs in EFL classes and among NSs of English.
Aşik and Cephe (2013) based their study on the production of DMs by Non-Native (NN) speakers of English as compared with those used in NSs' discourse on two separate corpora: a corpus of 20 NN learners in an English language teaching program in Turkey in comparison with the research corpus of MICASE for the NSs' presentations. By running frequency analysis, their study delved into the occurrences of DMs in both corpora. The findings of their research demonstrated that NN English speakers made use of a finite "number and less variety of DMs in their spoken English" (Aşik & Cephe, 2013, p. 144). Fung and Carter (2007) drew their data from a secondary Hong Kong classroom discourse corpus, a corpus of spoken British English, as well as a pedagogic sub-corpus from CANCODE to compare and contrast the production of DMs by both British NSs and L2 learners both qualitatively and quantitatively. Their findings demonstrated that both groups of NSs and L2 learners used DMs for interactional maneuvers on referential, interpersonal, cognitive, and structural levels. L2 learners showed a more frequent use of referentially functional DMs such as "but", "OK", "because", "so", as well as "and" than the NSs. However, the L2 learners illustrated a limited use of DMs such as "yeah", "really", "say", "sort of", "I see", "you see", "well", "right", "actually", "cos", and "you know". In addition, NSs regarded DMs for a broader range of pragmatic functions in comparison to their counterparts.
Neary-Sundquist (2013) investigated the data from 37 NN examinees at different proficiency levels and 10 NSs based on four tasks (news, personal, passing information, as well as telephone tasks) on the test of oral proficiency. She analyzed the range and the rate of hedges used to lessen the force of an utterance or the certainty of its content among L2 learners and NSs. The coded hedges in her study were "like", "I think", "just", "sort/kind of", "a bit", "or/and whatever", "or something", "everything/that/stuff/things", and "not really". She concluded that the L2 learners at elementary and intermediate levels significantly underused hedges in comparison with NSs and advanced L2 learners. The study revealed that the L2 learners at the advanced level surpassed NSs in using hedges while performing various monologic testing tasks. The most frequently used hedges among both NSs and L2 learners were "I think" and "just" in her study. In addition, she delineated that the two groups of L2 learners' and NSs' range of hedges on different monologic tasks was significantly higher than the rate of hedges.
Carrying out her research on 211 Chinese L2 learners from junior high school-, high school-, and university-level English courses in China on the pragmatic development of hedging, Yu (2009) studied simulated debates, written questionnaires, and oral interviews of the L2 learners. She found that intermediate L2 learners only showed an awareness of mitigators, the performative "I think", and intensifiers; while, the advanced university L2 learners significantly manifested an awareness of all categories of hedges at a rate higher than their intermediate counterparts. The findings of her study, also, indicated that there were significant differences between the oral interviews and the debate task with the L2 teacher according to the frequency and range of hedging clusters. Jalilifar et al. (2011) developed and distributed a reading comprehension test among 100 undergraduate L2 learners examining the influence of explicit instruction of DMs as hedging devices through a pre-test-post-test research design. Attending 10 sessions of awareness-raising treatment with regard to the appropriate use of hedges in English, the L2 learners in the experimental group and the control group took the posttest of the same reading comprehension test as it was employed for the pre-test. It is worth mentioning that the treatment included three sessions of awarenessraising for the types and functions of English hedging clusters. The remaining seven sessions instructed the learners on the practical use of DMs in academic texts. Their findings substantiated the fact that the L2 learners in the experimental group significantly surpassed their control group counterparts in employing DMs as hedging devices in the post-test phase. Also, the results revealed that explicit instruction in perceiving hedging exerts a facilitative influence on improving university learners' reading comprehension and language proficiency.
In another study, Siu (2014) investigated L2 learners' use of hedging in academic writing among 136 N Cantonese-speaking L2 learners who enrolled in the course of English for Academic Purposes (EAP) in China. The participants of her study were divided into control (i.e. at the beginning of a oneyear EAP course) and experimental (i.e. at the mid-course of a one-year EAP course) groups to delineate whether the learners used DMs as hedges to mitigate the force of an utterance appropriately. The DMs examined in her study consisted of nouns (e.g. probability and possibility), adverbs/ adverbial phrases (e.g. always and possibly), adjectives (e.g. probable and likely), epistemic modals (e.g. will and shall), non-verbal hedging which included quantifiers (e.g. some of and a few of), verbal hedging involving main verbs (e.g. assert and argue), as well as passive voice (e.g. be regarded as and be viewed as). The results of her study illustrated that the L2 learners who were explicitly instructed on the appropriate use of DMs as hedges significantly outperformed the language learners in the control group in academic writing. Furthermore, the study revealed that the L2 learners in the experimental group made use of the modal verb "may" more frequently than their counterparts in the control group since they were more conscious of the importance to minimize the force of an utterance while writing an academic essay.
By and large, DMs as hedging devices can add to the politeness load of an utterance and soften an FTA. Brown and Levinson (1987) argued that although a plethora of studies have been carried out on gender differences in using language in context, such discrepancies are "epiphenomenal-neither the social underpinnings nor the linguistic manifestations are specific to gender" (1987, p. 31). Nonetheless, previous research has indicated that gender significantly influences notions of politeness (Gilligan, 1982;Holmes, 1995;Hsieh, 2009;Mills, 2003;Tannen, 1990). In this regard, Mills (2003) contends that there is an urgent need for politeness research to emphasize gender through a contextualized analysis. She opposes the perspective that "politeness or gender consists of a range of stable predictable attributes" (Mills, 2003, p. 1). Moreover, Holmes (1995) discusses that female interaction is ordinarily less aggressive, argumentative, and competitive than that of males. Women, also, try to prevent disagreement and to provide supportive feedback to reinforce relationships (Gilligan, 1982).

Research questions
Given the above, the current study aimed to provide answers to the following questions: Q1: What types of DMs do Iranian L2 learners most frequently use?

Q2:
Are there any significant differences among control, experimental, male, and female L2 learners regarding different types of DMs and proficiency level? Q3: Are there any significant differences among the perception of native speakers, advanced Iranian L2 learners, and Iranian EFL teachers regarding the use of DMs presenting (im)politeness? Q4: How do L1 (im)politeness strategies influence the utilization of L2 (im)politeness DMs?

Participants
Three groups of participants took part in the current study. None of the participants had any experience living in an English-speaking country except the American NSs who lived in New York. It is worth mentioning that the researchers were in touch with the American NSs through email and Telegram since they did not have an opportunity to visit the NSs through face-to-face interaction. However, the researchers of the present study could easily contact the other participants through different direct meetings and frequent visits.
The first group consisting of 160 L2 learners (i.e. 80 males and 80 females) was divided into 8 groups of 20 L2 learners according to their proficiency levels (i.e. intermediate and advanced), gender as well as experimental-control variables. The female EFL learners comprised 42 learners of 15-20 years of age, 21 learners of 20-25 years old, 13 learners of 25-30 years of age, and 4 learners of 31 years and above. The male group included 47 learners of 15-20 years of age, 18 learners of 20-25 years old, 12 learners of 25-30 years of age, and 3 learners of 31 years above (See Table 1).
The second group included 90 participants among whom 30 were advanced Iranian L2 learners, 30 were American NSs, and the other 30 were Iranian EFL teachers. The L2 learners, EFL teachers, and NSs included 12, 9, and 7 participants of 15-20 years of age; 11, 14, 16 participants of 20-25 years old; as well as 7, 6, 9 participants of 25-30 years of age, respectively (See Table 2).  The third group consisted of 164 Iranian advanced L2 learners. Nevertheless, 14 L2 learners did not completely fill out the background information part and the scenarios; hence, they were omitted from the present study. The remaining 150 participants consisted of 59 learners of 15-20 years of age, 62 learners of 20-25 years old, 20 learners of 25-30 years of age, and 9 learners of 31 years above (See Table 3).
The L2 learners were asked to assess their proficiency level (Dewi, 2011) based on International Testing System or IELTS score band (IELTS, 2011, See Appendix A). According to Common European Reference Framework (CERF), the L2 learners who scored themselves 5-6 were reported to be at intermediate level, and those who marked themselves 7-9 were recognized to be at advanced level of proficiency.

Instruments
In order to assess the frequency and differences among the Iranian male and female L2 learners regarding their awareness of DMs as indicators of (im)politeness with reference to proficiency level and control-experimental variables, open Discourse Role-Play Tasks (DRPTs) were used (See Appendix B). An open DRPT is widely used for assessing language learners' pragmatic competence "without determining the course and outcome of the situations" (Eslami & Mirzaei, 2012, p. 203). To this end, the L2 learners looked into 21 role-play cards in which the researchers developed describing some situations without interfering with their course and outcome.
For evaluating the advanced L2 learners', EFL teachers', and NSs' awareness of DMs as indicators of (im)politeness, a questionnaire with seven scenarios was developed (See Appendix C). Each scenario consisted of two sentences: one including a hedging device (i.e. a DM) and one without it. A five-point Likert scale (i.e. 1 as very polite, 2 as polite, 3 as neutral, 4 as impolite, and 5 as very impolite) was established for each sentence for the participants to check. The first sentences of the scenarios 1, 2, and 3 included the approximator "or so", the modal verb "may", as well as the indefinite "anyone" as hedging devices, respectively; while, their second sentences did not incorporate any DMs. Nevertheless, the first sentences of scenarios 4, 5, 6, and 7 did not include any DMs; whereas, their second sentences deliberated the passive structure "I have been told", the tag question "won't you", the conditional "if" clause along with its modal "will", and the impersonal "It". The questionnaire was distributed among 10 advanced L2 learners, 10 EFL teachers, as well as 10 NSs to pilot the test. The reliability index, assessed by applying the Cronbach alpha (a) formula for the questionnaire was found to be .79.
To provide an adequate answer for whether there existed any influence on the utilization of L2 (im)politeness DMs by L1 (im)politeness strategies, the questionnaire with seven scenarios was used (See Appendix C). However, the researchers omitted the Likert Scale and instead told the advanced L2 learners to write the Persian translation for the two responses which were given to each scenario. After doing so, they answered to "Why do you think one or both of the responses are (im)polite in both English and Persian translation?"

Procedure
The data for this study were gathered in three phases which are discussed below:

Phase 1:
In order to answer the first and second questions of the current study, the background information along with the self-assessment of English competence were administered to four male and four female EFL classes in the Iran Language School (ILI) among 160 adult (i.e. 80 males and 80 females) L2 learners during four days in May 2017. The EFL classes consisted of 20 L2 learners each. Four of the classes were categorized into intermediate L2 learners and the other four were identified as advanced L2 learners based on CERF. Then, two of the male EFL classes at the intermediate and advanced levels were put into experimental groups. Also, two of the female EFL classes at the intermediate and advanced levels were considered as experimental groups. The remaining four classes of male and female L2 learners were regarded as control groups.
The four EFL teachers of the experimental groups were then trained for one hour discussing the importance of DMs as hedging devices to minimize the force of commands through different examples. These teachers provided the L2 learners of four classes with an instruction of one hour on the importance of DMs which may minimize the imposition and might compensate for the directness of an utterance. The instruction mainly included a limited number of DMs (i.e. approximators, modals, indefinites, passives, tag questions, conditionals, and impersonal "It") with different examples used as hedging to delve into the notion of (im)politeness phenomenon. The purpose was to make the learners aware of DMs as hedging devices and not to restrict them with these seven DMs. The control groups at the intermediate and advanced levels of language proficiency had no such instruction.
Afterward, the researchers and the eight EFL teachers of the current research provided the learners with 21 role-play cards each consisting of a situation with two characters that the L2 learners were expected to role-play. Using the role-play cards, the EFL teachers along with the researchers recorded the conversation between two L2 learners interacting with each other regarding the 21 situations. Each recording of the two learners role playing the 21 situations lasted approximately 60 min. This procedure occurred for the 160 male and female participants at the intermediate and advanced levels in both experimental and control groups during11 days in the ILI. Subsequently, the data were transcribed by the researchers to calculate the number of the DMs used as hedging devices to scrutinize the (im)politeness phenomenon. The DMs were computed through SPSS 22 using frequency distribution, measures of central tendency, and standard deviation for the eight groups of L2 learners. Also, three independent t-tests were conducted for comparing the use of DMs among the male and female L2 learners, the intermediate and advanced L2 learners, as well as the L2 learners at the control and experimental groups.
Phase 2: A questionnaire with seven scenarios was developed to assess the awareness of the advanced L2 learners, the EFL teachers, and the NSs considering the use of DMs as hedging devices as indicators of (im)politeness. Each scenario included two sentences: one consisting of a DM and one without it. Then, in order to pilot the test, 10 advanced L2 learners, 10 EFL teachers, as well as 10 NSs filled out the five-point Likert questionnaire. Through applying the Cronbach alpha (a) formula for the questionnaire, the reliability index was calculated to be .79. Using the background information along with the self-assessment of English competence, 30 L2 learners identified themselves at the advanced level of language proficiency according to CERF. After pilot testing, the questionnaire was distributed among 30 advanced L2 learners and 30 EFL teachers in two different sessions of 45 minutes. As for gathering of the data from American NSs who lived in New York, the researchers sent them the questionnaires through Telegram and email. They filled out the questionnaire and responded to the researchers' emails and telegram. The DMs were computed through SPSS 22 using frequency distribution, measures of central tendency, and standard deviation for both the EFL learners, EFL teachers, as well as the NSs. Moreover, a one-way between-groups analysis and post hoc comparisons using the Tukey HSD test were conducted to compare and to contrast the use of DMs among the L2 learners, the EFL teachers, and the NSs.

Phase 3:
In order to investigate whether Persian (im)politeness strategies influence the utilization of L2 (im)politeness DMs, the seven-scenario questionnaire was used (See Appendix C). Nevertheless, the researchers decided to omit the five-point Likert scale so that the L2 learners translate the two responses given to each scenario into Persian language. Afterward, the learners were asked to write in both Persian and English why or why not they think the responses were polite. To this end, the questionnaire was distributed among 164 advanced L2 learners; nevertheless, 14 of them were discarded from the present study since they did not fill out the questionnaire or fill it out incompletely. For classifying the DMs in both languages, the data were entered into MAXQDA Software, Version 11 (Verbi, 1989). Then, they were codified and described qualitatively.

Inferential statistics
Frequency and mean scores of seven DMs, namely approximators, modals, indefinites, passives, tag questions, conditionals, and Impersonal "It" based on control-experimental groups, intermediate and advanced levels of language proficiency levels, and gender are presented in Table 4.
For the male and the female L2 learners at the intermediate group in the experimental group, tag questions made up the highest frequency percent of 15.9 and 23.6 with the means of 2.90 and 4.30, respectively. However, for the language learners at the advanced level in the same group, the DMs of indefinites and impersonal "It" had the highest frequency percent of 29.5 and 37.1 with the means of 5.80 and 4.55, respectively.
Moreover, the highest frequency percent for the male and female L2 learners at the intermediate group in the control group was assigned to conditionals and modals with 10.4% and 7.2% and the means of 2.25 and 4.45, respectively. Nevertheless, for the L2 learners at the advanced level in the same group, modals accounted for the highest frequency of 11.3 and 12.2 with the means of 7 and 7.60.
An independent-samples t-test was conducted to compare the DM scores for advanced and intermediate L2 learners regarding approximators, modals, indefinites, passives, tag questions, conditionals, and impersonal "it". The results are presented in Table 5. In order to determine the difference in the males' and females' use of DMs presenting (im)politeness, an independent-samples t-test was calculated. The findings are shown in Table 6.     To assess whether there were significant differences among the language learners with reference to DMs in the control and experimental groups, an independent-samples t-test was computed. The results are illustrated in Table 7. With respect to all the DMs, significant differences were found in the scores of control and experimental L2 learners revealing that language learners in the experimental groups used DMs more frequently that their counterparts in the control groups.
In order to determine the difference among advanced L2 learners, EFL teachers, and American NSs in the use of DMs presenting (im)politeness, a one-way ANOVA was run through SPSS. The results are presented in Tables 8-10. To scrutinize how L1 (im)politeness strategies influenced the utilization of L2 (im)politeness DMs, 150 advanced L2 learners filled out the questionnaire from Appendix C. The analysis was conducted with the software MAXQDA, Version 11 (Verbi, 1989). Being regarded as a professional software for qualitative data analysis, MAXQDA is allows coding of a large amount of research materials such as interviews, written questionnaires, and so forth. These data were then sorted and retrieved from the MAXQDA software according to both the Persian and the English categories established. The findings are shown in Table 11.
As can be seen in Table 11, the Persian and the English shared the classifications of "modals", "impersonal "It", "tag questions", "conditionals", and "passives"; nevertheless, the categories differed in "precision" and "direct reasoning" in Persian versus "Approximators" and "Indefinites" in English.

Discussion
This study confirmed that the L2 learners at the intermediate group in the experimental group employed tag questions more frequently than their advanced counterparts who significantly used indefinites and impersonal "It". Also, the findings showed that the language learners at the intermediate level in the control group had the highest frequency of conditionals and modals in comparison to the Table 9. One-way between-groups analysis and post hoc comparisons using the Tukey HSD for the comparison of DMs among advanced L2 learners, EFL teachers, and American NSs *The mean difference is significant at the .05 level.

(I) Native speakersadvanced L2 learners and EFL teachers (J) Native speakersadvanced L2 learners and EFL teachers
Mean difference (I-J) Std. error Sig.

Native speakers
Advanced L2   advanced L2 learners in the same group who frequently used modals. In addition, the results of the current study demonstrated that the advanced L2 learners employed DMs more frequently than intermediate ones and that the L2 learners in the experimental groups outperformed their counterparts in the control groups in using DMs.
Considering gender differences the male L2 learners underused the DMs of approximators, modals, and passives in comparison to the female ones demonstrating that the female L2 learners soften the force of commands more than male language learners. Nonetheless, regarding the DMs of indefinites, tag questions, conditionals, and impersonal "It", the male and female L2 learners did not significantly differ in employing them. Furthermore, a one-way between-groups analysis of variance indicated that the EFL teachers surpassed the L2 learners in using DMs and that the American NSs employed DMs more frequently than both EFL teachers and L2 learners.
The classification of the Persian and the English DMs based on MAXQDA software, also, revealed that the shared areas between the two languages were "modals", "impersonal 'It', 'tag questions', 'conditionals', and 'passives'". However, the Persian categories were "precision" and "direct reasoning" in comparison to the English classifications of "approximators" and "indefinites". Such areas of difference between the two languages are illustrated below.
In English, most of the L2 learners consent to the fact that using approximators such as "or so", "almost", and "kind of" make the sentence fuzzier and more polite. Participant 21 writes, "The mechanic does not have to lie. He uses the approximator "or so" to add to the politeness of the sentence". Another participant also believes, "Since 'or so' is used in the first sentence, the utterance is more polite than the second sentence". Also, participant 78 holds, "Using approximators such as 'almost', 'or so', 'kind of' and so forth make the English sentences more polite". Nevertheless, in Persian, the participants mostly mention that giving an exact time demonstrates politeness. They assert that using approximators make them unsure about the certain time of doing some tasks which is impolite. Participant 16 believes that the sentence "It takes twenty minutes or so". is impolite. She claims, "It's impolite because the mechanic does not take account of my precious time". Another participant, also, believes that this sentence is rude to him. He asserts, "The mechanic does not specify the exact time of repair; hence, the customers' time is not important for him". Moreover, participant 47 believes, "It's always great to be exact and tell the exact time. It shows how punctual the mechanic is. I think being punctual shows politeness". Participant 88 writes, "When approximators are used in Persian, I think they make the sentences blunt and impolite". In addition, another participant holds, "Because of the word 'or so', the sentence becomes less polite than the sentence 'It takes twenty minutes'". Participant 113 claims, "Using an approximator is an indicatation of unfaithfulness and disloyalty which is impolite. It shows imprecision and inaccuaracy. On the other hand, telling the exact time is a sign for punctuality and faithfulness adding to the politeness of the sentence". Furthermore, participant 134 holds, "It's rather impolite because he wants to reassure if it takes longer he won't accept any complaints. It's polite because he reassures me that there won't be any delay".
A large number of the advanced L2 learners agreed to the fact that using indefinites such as "anyone", "anybody", "anywhere", "people say that", "somewhere" and so on make the utterance vague and more polite. Participant 67 holds, "The police officer does not mention a specific person in the first sentence. Thus, the sentence is more polite because the word 'anyone' is employed". Another participant writes, "The sentence which employed 'anyone' is more polite than the second one because it's a clear statement of law". In addition, participant 135 holds, "In the first sentence, the police officer is neutral in giving a ticket to the faulty person since he used the word 'anyone'. So, it is more polite than the second sentence which is much more direct". Nonetheless, in Persian, participants mainly write that giving reasons show politeness. Also, they delineate that employing indifinites such as "anyone" and "people say that" make the sentence impolite. Participant 93 holds, "The first sentence is impolite because any driver knows this. The second sentence is more polite because he restates what he has done wrong". Also, participant 89 believes, "I think the translation of the first sentence is impolite in Persian because the officer did not give any reasons to the driver. However, in the second sentence, the officer gives a reason and that's what makes the sentence more polite. Besides, participant 10 writes, "The first sentence is less polite than the second one since the police officer used the word 'anyone'. On the other hand, the second sentence mentions the specific law for giving a ticket. So, sentences become more polite by giving reasons". Last but not least, participant 147 mentions, "Giving reasons make the utterance more polite".
In the same vein with the results of the current study considering gender differences in using approximators, modals, and passives as hedging devices, Holmes (1995) argues that female's discourse is typically less aggressive than male's. Moreover, Gilligan (1982) argues that females typically avoid disagreement and provide supportive feedback more frequently than males in order to establish and to reinforce relationships. Pettersson Granqvist (2013) maintains that "women tend to use more hedges, boosters and facilitative tag questions: they might merely be evidence of a female conversational style" (2013, p. 28). However, the findings of this research are not in line with Brown and Levinson's (1987) claim that the discrepancies found in males' and females' discourse are "epiphenomenal-neither the social underpinnings nor the linguistic manifestations are specific to gender" (1987, p. 31). This study did not find any significant difference between males and females in using the DMs of indefinites, tag questions, conditionals, and impersonal "It".
In line with this study, Neary-Sundquist (2013) indicated that advanced L2 learners and NSs surpass the L2 learners at both elementary and intermediate levels in using DMs to mitigate the tone of an utterance or the certainty of its content in English. Nevertheless, with regard to various monologic tesing tasks, she confirmed that the advanced learners are significantly better than both L2 learners and NSs in using "I think" and "just" as the two most frequently used hedges in her study. The current research has found empirical evidence that the American NSs surpass not only the L2 learners at the intermediate and advanced levels but also the EFL teachers in using DMs to moderate the force of an utterance. Drawing their data from three different corpora, Fung and Carter (2007), also, confirmed that British NSs use DMs for a broader range of pragmatic functions in comparison to L2 learners. In addition, Aşik and Cephe (2013) carried out their study on the occurrences of DMs by both L2 learners and NSs demonstrating that L2 learners underuse DMs in their spoken English compared to NSs.
Likewise, Yu (2009) confirmed that advanced L2 learners show a higher awareness of hedging clusters than the intermediate ones. He found that the advanced L2 learners significantly use modal shields, performative shields, quantificational approximators, pragmatic-marker hedges, and other strategies for discoursal and syntactic hedges; while, the intermediate L2 learners only employ mitigators, intensifiers, and the performative "I think". Although results of the present study reveal that the advanced L2 learners outperform their intermediate counterparts in using DMs as hedging devices, intermediate L2 learners use tag questions, conditionals, as well as modals more frequently than the advanced ones who significantly make use of indefinites, impersonal "It", and modals.
In correspondence with the present research results, Jalilifar et al. (2011) found empirical evidence of explicit instruction of hedging for specific academic purposes. They maintain that the undergraduate L2 learners' reading comprehension performance in using hedging devices to mitigate the force of commands in appropriate contexts significantly heightens their language and pragmatic proficiency level. Similarly, Siu (2014) examined the L2 learners' use of hedging in academic writing and illustrated that the learners who have hedging instruction outperform their counterparts in the control group in employing hedging devices appropriately in academic writing. She holds that the L2 learners in the experimental group are more familiar with using hedges in appropriate contexts to soften the force of an utterance when they write an academic essay; therefore, their writing piece becomes more polite with reference to the academic context. Also, in the present study, two groups of L2 learners at both the intermediate and advanced language proficiency levels had an explicit instruction regarding the use of hedges in different contexts which culminated in their more significant use of DMs as hedging devices than their counterparts in the control groups.
As Alijanian and Vahid Dastjerdi (2012) aptly puts it, "indirectness is considered a universal discoursal strategy but the extent to which it is applied varies from culture to culture" (2012, p. 60). Scollon (1997) argues that indirectness is communicated with the collectivist values and cultural notions which may differ in Eastern and Western societies influencing the way individuals interact with each other. Regarding Persian language and culture, Eslami-Rasekh (2004) delineates, "Iranian society, being a more group-oriented society, … puts more emphasis on the importance of society, family, solidarity, and common ground as opposed to individual privacy … and autonomy of individuals" (2004, p. 189). As the results of the present research for the classification of the Persian and the English DMs based on MAXQDA software suggest, the two areas of controversy between Persian and English cultures are "precision" and "direct reasoning" versus "approximators" and "indefinites", respectively. Regarding (im)politeness phenomenon in Iranian culture, interaction with other interlocutors of a community is a way to build relationship and harmony which makes them cautious and precise in communicating their own statements and goals in relation to other individuals. In this respect, Iranian L2 learners aim to arrive at reciprocal harmony and consensus through employing DMs as hedging devices to underscore priority for partnership and other people's face through "precision" and "direct reasoning" rather than "approximators" and "indefinites". Also, it can be mentioned that dispute and controversy are viewed as FTAs that may cause the L2 learners to lose face.

Conclusions
The current research investigated the interlanguage development of Iranian L2 learners and EFL teachers as well as the pragmatic knowledge of American NSs. It investigated the DMs as hedging devices as indicators of politeness with regard to gender, proficiency level, as well as control-experimental variables through open DRPTs and a seven-scenario questionnaire with a five-point Likert scale and without it for L2 learners to translate into Persian. The results demonstrated that the female L2 learners outperformed their male counterparts in using the DMs of approximators, modals, and passives revealing that the female language learners minimize a face-threatening act and the tone of an utterance significantly. Hence, such results corroborate Gilligan's (1982), Holmes (1995), Pettersson Granqvist (2013) studies in that females tend to use more hedges than males as mitigating devices. However, the findings of this study indicated that both groups of L2 learners did not show any significant differences in using indefinites, tag questions, conditionals, and impersonal "It" in their speech.
The results, also, indicated that the American NSs surpassed both EFL teachers and L2 learners in using DMs as hedging devices corroborating Fung and Carter's (2007) as well as Aşik and Cephe's (2013) studies. Besides, the current research found that intermediate language learners significantly used DMs to soften the force of commands and minimize FTAs less frequently than advanced L2 learners which substantiated earlier research by Neary-Sundquist (2013) and Yu (2009). This study confirmed that EFL teachers outperformed both intermediate and advanced learners in using DMs. In addition, the L2 learners in the experimental groups surpassed their counterparts in the control groups in employing DMs in their speech.
In addition, the present research revealed two areas of discrepancies, namely "precision" and "direct reasoning" in Persian versus "approximators" and "indefinites" in English, according to MAXQDA software to find the nuances of (im)politeness theory, contextual features and cultural subtleties between the two languages. Such results added to Eslami-Rasekh's (2004) research that Iranian culture underscore common ground as opposed to individual privacy which makes Iranians cautious and precise in order to be considered as polite in everyday language encounters in relation to other individuals. Hence, being more precise and providing more direct reasoning are regarded as two polite strategies in the Iranian culture rather than making the utterance fuzzier and indirect through using the DMs of "approximators" and "indefinites". 6. You are in a park to buy an ice-cream and there are many people in line. It's your turn to buy an ice-cream. You search your pocket and you remember that you put the money in your bag at home. The clerk in charge of the ice-cream cones wants you to try to be faster.
9. Your friend has financial problems and asks whether you can help him with some cash.

You: …………………………………………………………………………………………….
10. You are a judge in a court of law. The criminal in the court wants you to do him a favor illegally and shorten his imprisonment sentence. He wants to tell you that when he gets out of jail, he'll make it up to you.

You: …………………………………………………………………………………………….
15. You want to send a letter to your best friend so you go to a post office. You forget to close the door behind you. The clerk in the post office asks you whether you could close the door.

You: …………………………………………………………………………………………….
17. You bought a T-shirt at a popular department store which the seller told you that it was original and you paid a lot of money. However, one of your friends, an expert in distinguishing between original and fake T-shirts, tells you that the T-shirt is fake. You go back to the department store to tell them that it is not original.
18. You are sitting in a park bench to have some fresh air. A stranger comes and sits close to you smoking a cigarette. You feel so embarrassed. The stranger understands it and wants to know what bothers you.

You: …………………………………………………………………………………………….
19. You are a prominent lawyer in your city and want to defend your client whom you believe that he is not a criminal. The attorney general talks too much and does not allow you to introduce and defend your client. After some time, the attorney general tells you what evidence you have t defend your client.