The Effect of Pragmalinguistic Focus on Form on Iranian EFL Learners ’ Production of Refusal Forms : with a Focus on Marked Recasts and Elicitation

The role of instruction to develop learners’ pragmatic competence in both second and foreign language contexts has recently motivated a great deal of research. However, most of this research has adopted an explicit instructional approach with only a few studies attempting to operationalize a more implicit condition for pragmatic learning. This study was set up to reexamine the effects of pragmalinguistic focus on form on Iranian EFL learners' pragmatic development and to compare the effects of two feedback types (marked recasts and elicitations) in raising the learners' awareness in producing appropriate English refusals. Adopting a pretest-posttest design, the study included 21 participants with two experimental groups (marked recast group versus elicitation group) but no control group, adopting a pretest-posttest design. Both of the groups performed role plays. In marked recast group, the researcher provided the learners with marked recasts while in elicitation group, he used elicitation strategies to deal with erroneous utterances. The results from a written Discourse Completion Test (DCT) showed the effectiveness of pragmalinguistic focus on form on the participants’ pragmatic development and revealed that although both of the techniques were effective, there were no significant differences between them in raising the learners’ pragmatic ability to produce appropriate refusal forms. The findings of the study suggest that pragmatic instruction, which seems necessary, can be operationalized through both recast and elicitation, as two techniques of Focus on Form.


Introduction
Nowadays, learning a foreign language is regarded as an essential component in the curricula at different educational levels.In particular, learning English has become necessary given its widespread use throughout the world (House & Kasper, 2000).Considering the worldwide importance of the use of English as a means of international communication, it seems that learning and teaching English is a necessity in our society.However, in order to make learners become communicatively competent in English, there was a need for a shift from previous theoretical frameworks, which consider language as a formal system based on grammatical rules, towards a more communicative perspective.This change became possible with the introduction of pragmatics as a specific area of study within linguistics that had a focus on interactional and contextual factors of the target language (Alcaraz, 1990, 1996, as cited in Martinez Flor, 2004).In this way, different models of communicative competence were developed in the field of second language acquisition, which included not only grammatical, but also pragmatic competence as one of its main constituent (Bachman, 1990).Thus, the growing importance and attention paid to examining learners' pragmatic knowledge has given rise to a new area of research known as Interlanguage pragmatics.The findings from this line of research suggest that learners' pragmatic competence is incomplete despite having a high level of grammatical knowledge or having spent time in the target language community (Bardovi-Harlig, 2001).Therefore, it has been argued that teaching pragmatic conventions is necessary to develop learners' ability to communicate appropriately in the target language, particularly in the foreign language context (Bardovi-Harlig, 2001;Rose & Kasper, 2001).
Based on the fact that learners of English as a foreign language have little, if any, opportunity of being exposed to authentic language outside the classroom, even grammatically proficient ones lack pragmatic competence which is considered as a vital component of language competence and as important as grammatical competence.According to Bachman (1990), language competence is comprised of organizational competence and pragmatic competence.Research findings assert that grammatical development does not guarantee a corresponding level of pragmatic development and that even advanced learners may fail to comprehend or to convey the intended meaning and politeness values.Thus, for a learner to be competent enough, developing both of these competencies is necessary.
It is claimed that learners have more or less some pragmatic knowledge in their first language that can be positively transferred into L2 provided that learners become aware of it, for example, turn taking (Yule, 2006).However, there are other aspects of pragmatics that are specific and exclusive for a language which may not be present in one's L1.It may be thought that perhaps pragmatic knowledge simply develops alongside grammatical knowledge, without requiring pedagogical intervention.However, research into the pragmatic competence of adult foreign and L2 learners have demonstrated that pragmatic knowledge of learners and native speakers are quite different (Bardovi-Harlig & Mahan-Taylor, 2003).On the other hand, the consequences of pragmatic failures, unlike grammatical errors, are often interpreted at social or personal level; they may even hinder a good communication or make the speaker appear rude and uncaring in social interactions (Bardovi-Harlig & Mahan-Taylor, 2003).Therefore, as Bardovi-Harlig (2001) demonstrates, there is a need for instructional intervention to focus on the pragmatic knowledge of a language, which can have positive impact on raising learners' pragmatic awareness and their abilities to produce appropriate language.
Moreover, feedback provision is considered a key factor in developing L2 learners' interlanguage and approaching them as much closer as possible to the target language (Martinez Flor, 2004).Nevertheless, its effectiveness for raising learners' pragmatic knowledge in order to produce accurate and appropriate utterances in a given context has not been fully addressed and is under investigation at the moment.
Hence, the purpose of this study is to investigate the effect of pragmalinguistic focus on form through two feedback types, namely, marked recasts and elicitations, on learners' production of English refusal forms.
To this end, the following research questions were posed: 1-Does marked recast, as a FonF technique, have any effect on the Iranian EFL learners' production of English refusal forms?2-Does elicitation, as a FonF technique, have any effect on the Iranian EFL learners' production of English refusal forms?
3-Are there significant differences between the effects of marked recasts and elicitation on the Iranian EFL learners' production of refusal forms?And accordingly, the following null hypotheses were formulated: 1-Marked recast, as a FonF technique, has no effect on the Iranian EFL learners' production of refusal forms.
2-Elicitation, as a FonF technique, has no effect on the Iranian EFL learners' production of refusal forms.
3-There are no significant differences between the effects of marked recast and elicitation on the Iranian EFL learners' production of refusal forms.

Pragmatics
Pragmatics is defined as "the study of the ability of language users to pair sentences with the contexts in which they would be appropriate" by Levinson (1983, p.24).In other words, as stated by Yule (1996), pragmatics deals with the relationship between linguistic forms and the human beings who use those forms.In Leech's (1983) words, pragmatics could be defined as the study of the use and meaning of utterances to their situations.Leech (1983) suggests that language involves a pragmalinguistic perspective and a sociopragmatic perspective.Pragmalinguistics refers to the grammatical side of pragmatics and addresses the resources for conveying particular communicative acts.Such resources include pragmatic strategies like directness and indirectness, pragmatic routines, and a range of modification devices which can intensify or soften the communicative act.On the other hand, as Leech (1983) puts it, sociopragmatics deals with the relationship between linguistic actions and social structure since it refers to the social factors, such as status, social distance, and degree of imposition that influence what kinds of linguistic acts are performed and how they are performed.

The Speech Act of Refusal
Refusals, as all other speech acts, occur in all languages.The act of refusing is a complex issue, as the speaker directly or indirectly says "no" to his/her interlocutor's request, invitation, suggestion, and/or offer.According to Al-Kahtani (2005), how one says no is probably more important than the answer itself.Therefore, sending and receiving a message of "no" is a task that needs special skill.The interlocutor must know when to use the appropriate form and function depending on the community and its cultural-linguistic values (Al-Kahtani, 2005).It seems that refusals are often realized through indirect strategies which require a high level of pragmatic competence.If refusals are challenging for native speakers, the situations become even more complex for non-native speakers.That is why refusals are known as "striking points" for many non-native speakers in cross-cultural communications (Beebe, Takahashi, & Ulitz-Weltz, 1990).A failure to refuse appropriately can risk the interpersonal relations of the speaker.
Since a refusal is a face threatening act, it causes damage to the face of both interlocutors; It may threaten the negative face of the hearer and also be a threat to the positive face of the person who refuses (Beebe, et al., 1990).In order to preserve face in this situation, people use a lot of face-saving strategies to maintain harmony in their relations.Therefore, the interlocutors must be sufficiently competent in order to be able to handle the situation through using appropriate pragmatic strategies.Bardovi-Harlig (1999) noted that research into pragmatic competence has repeatedly asserted the fact that even those proficient L2 learners of English lack the necessary pragmatic competence.That is, they are not aware of the social, cultural and discourse conventions that have to be followed in various situations.This may be due to several factors among which the most important one is as fallow: teachers often lack adequate textbook materials that focus on pragmatic functions (Bardovi-Harlig, Hartford, Mahan-Taylor, Morgan, & Reynold, 1991).They state that as language classroom is regarded as an ideal opportunity for foreign language learners to develop their pragmatic ability alongside their grammatical proficiency, such inadequacy of pragmatic-focused material, complicates the prearranged objectives, i.e., learners do not acquire the necessary pragmatic conventions.According to Blum-Kulka (1991), L2 learners can get a considerable amount of L2 pragmatic knowledge for free provided that they become aware of the fact that some pragmatic conventions are universal (such as turn taking and politeness values) and thus transferable from L1 to L2.This phenomenon is called "positive transfer" in linguistics.The role of pedagogical intervention is not only to provide learners with new information, but also to make them aware of what they know already and encourage them to use that universal or transferable L1 pragmatic knowledge in L2 contexts (Rose & Kasper, 2001).On the other hand, L2 learners may not be aware of pragmatic discrepancies between the two cultures and simply transfer their L1 sociopragmatic norms to the L2 context which leads to pragmalinguistic and/or sociopragmatic failure in their encounters with native speakers (Blum-Kulka, 1991).Therefore, as Bardovi-Harlig (et al. 1991) state, leaving to their own devices such as contact with the target language in and out of classroom, the majority of learners apparently do not acquire the secret pragmatic knowledge on their own and lack of this knowledge may have some bad consequences, e.g., it may hinder good communication between speakers, or make the speaker appear rude and uncaring in social interactions.As Bardovi-Harlig and Mahan-Taylor (2003) put it "Unintentional insult to interlocutors has been identified as another potential pragmatic hazard"(p.38).There is, thus, a strong indication that instructional intervention may be facilitative to, or even necessary for, the acquisition of L2 pragmatic ability specifically in foreign language context.

Focus on Form
Conceptually, focus on form involves "an occasional shift in attention to linguistic code features-by the teacher and/ or one or more students-triggered by perceived problems with comprehension or production" (Long & Robinson, 1998, p.23).In other words, it is characterized as learners' engagement in meaning with brief interventions and brief explicit instruction of linguistic codes as needed (Doughty & Williams, 1998).Thus, FonF approach would allow for L2 learners to concentrate on the grammatical rules and constructs of the language.A focus on meaning (FonM) approach, on the other hand, would be concerned with getting L2 learners to concentrate solely on understanding the message being conveyed.FonF is, also, distinguished from Focus on forms (FonFs) by Long (1991Long ( , as cited in muranoi, 2000) ) in that in a FonF approach the primary focus is on meaning and communication, with the learners' attention being drawn to linguistic elements only as they arise incidentally while in FonFs the focus is on forms themselves in isolation.In other words, while FonF integrates attention to form, meaning, and use, FonFs entails isolation of linguistic features from context or from communicative activity (Doughty & Williams, 1998).

Corrective Feedback
Regarding the acquisition of pragmatic competence, research on language socialization has shown that parents instruct their children in pragmatics by providing them with negative feedback (Kasper & Schmidt, 1996); thus, corrective feedback on both meaning and form plays an important role in developing learners' pragmatic ability in the classroom.Lyster and Ranta (1997) distinguished six types of corrective feedback in teacher-student interaction in French immersion classrooms and described them as fallow: 1-Explicit correction (i.e., the teacher supplies the correct form and clearly indicates that what the student said was incorrect).
2-Recasts (i.e., the teacher implicitly reformulates all or part of the student's utterance, minus the error).
3-Clarification requests (i.e., the teacher indicates to students that their utterance has been misunderstood by the teacher and a repetition or reformulation is needed).4-Metalinguistic feedback (i.e., the teacher provides comments or questions related to the well-formedness of the student's utterance).5-Elicitation (i.e., the teacher directly elicits a reformulation from the students).
6-Repetition (i.e., the teacher repeats the student's ill-formed utterance, adjusting intonation to highlight the error).Lyster and Ranta (1997) found that among these feedback types, recast was the most frequent but the least effective one in their study.However, they found metalinguistic feedback, elicitation, and clarification requests as the most effective ones.They maintained that ineffectiveness of recasts comes from their ambiguous nature in that learners hardly ever notice their corrective function.Some researchers, however, have argued that the juxtaposition of the learner's illformed utterance with the teacher's reformulation provides the learner with an ideal opportunity to make a cognitive comparison and notice the gap between the target-like and non-target forms, especially because meaning is held constant and so the learner's processing resources are freed up to focus on form (Doughty, 2001;Long & Robinson, 1998).As mentioned earlier, recast as a form of negative feedback, involves" the teacher's reformulation of all or part of a student's utterance, minus the error" (Lyster & Ranta, 1997, p.46).They draw learners' attention to discrepancies between erroneous and target forms which enhance the salience of the target form (Doughty & Varela, 1998).Lyster and Ranta (1997) described recasts as implicit focus on form and believe that they are effective because they add attention to form in a primarily communicative task rather than to depart attention from it.Lyster (2002), also, asserts that as positive evidence, recasts may facilitate the internalization of new forms and as negative evidence draw attention to the mismatch between the interlanguage and the target form.
Researchers (e.g., Farrokhi, 2005;Doughty, 2001) have identified several advantages for recasts as follow: they are1) less time-consuming; 2) less disruptive of the ongoing flow of the talk while the problem is dealt with; 3) less face threatening; and 4) easy and quick.However, they suffer from a main shortcoming which causes researchers and teachers to doubt their effectiveness, i.e., their ambiguity.Some researchers (Lyster & Ranta, 1997;Lyster, 1998bLyster, , 2004) ) argue that because of their ambiguous nature they may not be perceived as corrective feedback and not be effective to draw attention to form.They are ambiguous because they have other discourse functions such as confirmation check and they often co-occur with signs of approval (Lyster, 2004).To overcome such problem some researchers (Doughty & Varela, 1998;Fukuya & Zhang, 2002;Farrokhi, 2005) employed various techniques to enhance the amount of their perceptibility as corrective feedback which were, then, called as marked/focused recasts.Farrokhi (2005) categorized recasts as unmarked recasts and marked recasts in his study.He characterized marked recasts as " a teacher's implicit corrective reformulation of a learner's non-target like production while preserving the meaning of the student's utterance and highlighting or marking the reformulation in a number of ways" (p.109).He differentiated them from unmarked recasts in a way that the latter lacks the additional elements to highlight the reformulation.The following examples (Farrokhi, 2005, p.109, 111) illustrate unmarked and marked recasts respectively:

Extract 1
T: what are you going to do during the holiday?S: I'll go to movies, and I'll study some books.
T: excellent, so you will go to the movies and study some books.What else?
In this extract, the teacher reformulates the learner's non-target-like utterance without highlighting it (movies is changed into the movies).
Extract 2 S: I was with my family at home because we had some guest.
T: Guests, yeah S: Guests and we had to stay home without being able to go anywhere.
In this extract, the teacher highlights the reformulation with rising intonation to signal the error to the learner.Farrokhi (2005) argues that marked recasts not only do not interrupt the flow of talk, but also are perceived as corrections by learners, which help them to notice the gap between target-like and non-target-like forms.Marked recasts encompass all of the advantages attributed to unmarked recasts.Besides, as Farrokhi states, they can be optimal way to integrate focus on form with focus on meaning at the level of error correction.

Elicitation
Another kind of negative feedback is elicitation, a specific type of prompts where the teacher pushes the learner to formulate their own erroneous utterances without stating that an error has occurred or suggesting where the error may be (Lyster, 1998a).According to Lyster and Ranta (1997), elicitation refers to at least three techniques that teachers use to elicit the correct form from the student.Through elicitation moves, the teacher does one of the following techniques: a) the teacher strategically pauses to allow students to fill in the blanks; b) the teacher uses questions to elicit correct forms (e.g. , How do we say that in English?); and c)the teacher asks the students to reformulate their utterances.Elicitation moves provide learners with opportunities for self-repair by withholding correct forms which help learners to enhance the retrieval of the forms learners already know (Lyster & Ranta, 1997;Lyster, 2004).They have, also, been argued to provide learners with opportunities for pushed output which cause them to attend to form.

Empirical Research
Nipaspong and Chinokul ( 2010) conducted an experimental study on the effectiveness of explicit feedback and prompts in raising English as a foreign language learners' pragmatic awareness of the use of appropriate refusals.After a 10week treatment, the results from a pragmatic awareness multiple-choice test and qualitative data revealed a significant improvement in the pragmatic awareness of the prompts group over the explicit feedback and control groups.They argued that the key factor leading to the advantages of prompts may result from their demand for learners to generate repairs and their provision of more opportunities for learners' uptake.Some other interactional studies on pragmatic development have employed FonF techniques.For instance, Fukuya and Clark (2001) conducted a study in which they examined whether FonF was an effective pedagogical framework for teaching pragmalinguistics; the results of the study proved to be inconclusive due to several reasons such as brevity of IJALEL 3(2):53-63, 2014 57 treatment and weak statistical power.Emphecising the need for further research, they concluded that the inconclusive findings should not be seen as evidence of the failure of FonF in the realm of L2 pragmatic instruction.Fukuya and Zhang's (2002) study was another interventional study that tried to operationalize the treatment group by employing a FonF technique, namely that of recasts.In this way, they were interested in examining whether pragmalinguistic recasts were effective for teaching pragmatically appropriate and linguistically correct requests.The results from a written Discourse Completion Test showed that the instructed group who received recasts outperformed the control group in their use of the target forms addressed in the study.They claimed that the improvement observed in both groups might have been the result of their performance in the role-plays.
Considering the points mentioned above and the gap in the related literature, the present study attempted to investigate the effect of pragmalinguistic focus on form through two feedback types, namely, marked recasts and elicitations, on learners' production of English refusal forms.

Participants
The participants of this study were 21 female English language learners from a private language institute.According to the registration department, they were all intermediate learners who had been studying English as a foreign language at least for 9 terms.In this language institute, prior to take any course, the students have to take a proficiency exam (TOEFL).That is why the researcher found it unnecessary to give any language proficiency test initially.The participants' first language was Azari and they were either high school or university students.Their age ranged between16 to 25 and none of them had the experience of living in an English speaking country.

Instruments and Materials
The materials used in this study included a number of conversations containing the speech act of refusal which had been selected from "Spectrum" series.They had refusing of high, equal and low status person's request, invitation, offer, and suggestion.Some video movies from "Interchange" series, which contained refusal speech act, were also used.
The instrument used to collect data was a discourse completion test (DCT), adopted from Beebe et al. (1990), which was used as both the pretest and posttest (with altered order of items).It included 12 items in which the participants were asked to refuse a request, invitation, offer, or suggestion.Apart from these 12 items, used in the pretest and posttest, 16 other situations were, also, used as a follow-up practice.Furthermore, a four-point scale, developed by Jernigan (2007), was used to score the participants' responses to the written DCT items.

Design
This study was a qusai-experimental research with a pretest-posttest design including two experimental groups but no control group.One of the experimental groups was called marked recast group (MRG) and the other group Elicitation group (EG).Their written production of refusal forms was the dependent variable of the study while the independent variables were: marked recasts and elicitations.

Procedure
At first, the researcher asked 23 language learners to take a short course for free and they all accepted to participate except two of them.Then, some demographic information about the participants was gathered via a demographic survey.In this study, the researcher himself taught both groups six sessions in two weeks.In the first session, the participants in both groups were given the pretest which was a discourse completion test including 12 items adopted from Beebe et al. (1990).For each item, they had to provide a refusal form to a request, an invitation, an offer, or a suggestion posed by a high, equal, or low-status speaker.Then, the two groups received their particular treatment (two different techniques of pragmalinguistic Focus on Form) in four 90-minute sessions.The instruction for both of the groups was the same but they differed in the feedback they received.That is, in both groups the participants, first, listened to three dialogues in every session which included a speech act of refusal and practiced them with their partners.Then, two pairs were randomly asked to role-play them.In the second phase, the teacher provided them with two situations, similar to the dialogue, and asked them to prepare a role-play to act out.At this time, the teacher would walk around the class to assist their interactions.After they became ready, two pairs were asked to perform their roleplays and then they were asked to switch their roles.The last 30 minutes of each session were devoted to the participants' role-plays with the teacher himself.It was the third phase of the treatment in which all the participants in both groups individually role-played with the teacher in the class and it was the only time they received pragmalinguistic feedback.That is, the errors of the participants in the MRG group were treated via pragmalinguistic marked recasts while the errors of the participants in the EG group were treated through pragmalinguistic elicitations.The type of feedback they received was the only operational difference between the two groups.During each session every participant had a chance to role play with the teacher once and receive pragmalinguistic feedback.Finally, the posttest was administered to both groups which included the same twelve items of the pretest but in an altered order.

Inter-rater Reliability
As mentioned before, the participants' answers to the pre-and post-test written DCT were rated by two raters, including the researcher himself and one of his colleagues who has been teaching English for several years.Before the data analysis, the reliability of the rating procedure in all given tests throughout the present study was determined by measuring inter-rater reliability.Tables 4.1 and 4.2 show the results of the first and second ratings of the pre-and posttests of MRG group respectively.As illustrated in Tables 4.5 and 4.6, since both of the correlations in the pre-and post-tests are more than 0.70 (.958 > .70 and .958>.70),thus, it can be claimed that the rating procedure has been highly reliable.Therefore, the average scores of the two raters were considered as the final scores of the participants.The maximum score of the questionnaire with 12 items was 48.

The Pre-and Post-tests analyses
To analyze the pre-and post-test scores, the researcher first conducted a Kolmogorov-Smirnov test to ensure that there was no violation of the assumption of normality.Table 4.7 illustrates the results.The descriptive statistics of the pretest show the invariability of the scores between the two groups.
To find out if there was any progress made because of the treatments within the individual groups, the researcher conducted Paired T-test for both of the groups.Table 4.10 shows the results for MRG group.4.9, the treatment for MRG group was significant (P= .000<.05), i.e., providing the learners of the MRG group with marked recasts was effective for raising their awareness to produce appropriate refusals.Thus, the first null hypothesis, stating that marked recast has no effect on the Iranian EFL learners' production of refusal forms, was rejected.
Paired Samples Test was also run for the EG group.Table 4.11 shows the results.4.10, the treatment for the EG group was also significant (P= .000<.05).This means that providing the learners in the EG group with elicitations was also effective for raising their awareness to produce appropriate refusals.Thus, the second null hypothesis, stating that elicitation has no effect on the Iranian EFL learners' production of refusal forms, was, also, rejected.In other words, both of the treatments were significantly effective to raise the learners' awareness to produce appropriate refusals.Yet, to compare the scores of the MRG and EG groups for any possible differences between them, ANCOVA was carried out with an independent variable (including two levels of marked recast and elicitation), a dependent variable (i.e., the written production scores of the participants on the posttest) and the pretest scores as a covariate.Table 4.11 shows the results of this analysis.As illustrated in Table 4.11, the results of ANCOVA show that there is no significant difference between the means of the two groups on the posttest scores [F (1.27) = .893,P= .715> .05].This means that the effects of both treatments were the same.Thus, the null hypothesis, stating that there are no differences between MRG and EG in terms of their production of refusal forms, was not rejected.

Discussion
The first research question examined the effects of pragmalinguistic marked recasts on learning refusal forms.The corresponding hypothesis predicted that marked recasts have positive effects on the Iranian EFL learners' production of refusal forms, i.e., learners who received marked recasts would benefit them in producing refusal forms.Comparing the learners' posttest and pretest scores in MRG group, the researcher found that providing learners with pragmalinguistic marked recasts can assist them to develop their pragmatic competence, which is delimited to the production of refusal forms in this study.This finding is in accordance with the findings of previous studies.For instance, Fukuya and Zhang (2002) attempted to raise Chinese learners' awareness in producing English requests through "focused recasts".The findings of their study showed that learners receiving focused recasts outperformed those learners who did not receive any kind of feedback.Although their study was one of the pioneers in investigating the role of FonF paradigm in promoting learners' pragmatic awareness , it was unidimensional in that they only examined the effects of one of the FonF techniques in their study and other studies were necessary to shed more light on the ground.Martinez Flor ( 2004) conducted another study to compare the effects of two instructional approaches (i.e., Focus on FormS vs. Focus on Form) on EFL learners' pragmatic development with a focus on suggestions.In this study, FonFs was operationalized through a sequential method including the presentation of videotaped situations that involved native speakers interacting in different computer-related situations, the video scripts from these situations, and a sequence of activities from awareness raising tasks to production tasks.On the other hand, FonF instruction was operationalized through a parallel method with the combination of two implicit techniques, i.e., input enhancement through the video presentation and recasts during the role-play practice.The results of the study showed that although learners receiving either explicit or implicit instruction significantly increased their use of suggestions after the study, there were no significant differences between the two experimental groups; that is, both treatments were equally effective in the development of learners' production of pragmatically appropriate suggestions.Even though a combination of tow FonF techniques was employed to operationalize the FonF paradigm in this study, the efficacy of FonF approach on pragmatic development was verified.
The second research question was concerned with whether pragmalinguistic elicitations have any positive effect on Iranian EFL learners' production of refusal forms.The corresponding hypothesis predicted that pragmalinguistic elicitations have positive effect on developing Iranian learners' pragmatic knowledge in terms of refusal forms.The analysis of the EG group's posttest scores showed that, similar to marked recasts, elicitations had positive effects on developing Iranian EFL learners' production of refusal forms; thus, the predicted hypothesis was supported.Accordingly, it seems that providing learners with either marked recasts or elicitations can be beneficial and effective in developing their pragmatic production.This finding supports the findings of the only conducted study in this respect (Nipaspong and Chinokul, 2010) which compared the effects of three types of feedback, prompts, explicit feedback, and delayed explicit feedback used as the control group treatment, in promoting learners' pragmatic awareness.They reported that even though learners of both experimental groups outperformed the learners in the control group, prompts including elicitations were more effective.They concluded that the key factors influencing the effectiveness of prompts were the provision of multiple opportunities for uptake and its demand for learners' generated repair.In other words, the type of corrective feedback that leads to greater opportunities for learners' uptake and immediate repair benefits learners' pragmatic competence in terms of their awareness and production.This is so because when teacher prompts a cue or a question to push learner to do self-repair, it would activate learners' pragmatic awareness as they have more opportunities to perform mental processes like rethinking, retrieving, and reformulating their utterance (Nipaspong & Chinokul, 2010).
Finally, the third research question seeks to find out whether there was any difference between the effects of two FonF techniques (i.e., marked recasts vs. elicitations) on developing the Iranian EFL learners' refusal production.The corresponding hypothesis predicted that there were significant differences between the effects of marked recasts and elicitations on the production of the learners' refusal forms.The comparison of the post-test scores of the two groups revealed that there were no significant differences between the performances of the two groups, holding that there were no significant differences between the performances of the two groups in the pretest.That is, although the two techniques were equally effective in developing the Iranian EFL learners' pragmatic production, no significant differences were found between them.To the knowledge of the researcher, no study has been conducted to compare the effects of marked recasts and elicitations on the level of pragmatic development.However, some studies have attempted to compare their effects on different aspects of second/foreign language development other than pragmatic aspects.For instance, Nassaji (2009) compared the effects of both explicit and implicit forms of elicitations and recasts on targeted forms.He found that in both cases the more explicit forms of each feedback type resulted in higher rates of immediate and delayed postinteraction correction than their implicit forms.However, the effects of explicitness were found to be greater for recasts than for elicitations.Farrokhi (2005) also compared the effects of marked recasts vs. elicitations across fluency, accuracy and personal contexts.The findings of his study revealed that both of the feedback types were effective.However, he suggested that marked recasts were more in harmony with fluency contexts whereas negotiated feedback (elicitations) was more congruent with accuracy contexts.
The findings of the present study can be summarized as fallows: first, both marked recasts and elicitations, as two kinds of pragmalinguistic FonF techniques, were effective in developing Iranian EFL learners' production of refusal forms.The consistency of this finding with the findings of previous studies may vanish the doubts about the efficacy of FonF paradigm in pragmatic development.That is, language teachers can draw on such easy and economical techniques to develop their learners' pragmatic ability in producing appropriate and accurate target forms in the given context.
Second, no significant differences were found between the effects of marked recasts and elicitations on pragmatic development of Iranian EFL learners as far as refusals are concerned.Although the learners who received elicitations gained slightly higher mean size on the post test than the learners who received marked recasts, this difference was not statistically significant.The second finding does not reflect the findings of previous studies in that they have reported the efficacy of elicitations over marked recasts (Lyster & Ranta, 1997;Lyster, 2004).However, these studies have been conducted on the level of grammar.Based on the results of these studies, recasts were less effective than elicitations due to two reasons: 1) recasts were ambiguous which decreases learners' possibility of receiving them as corrective devices; and 2) elicitations push learners to self-correct their erroneous utterances by providing them with opportunities to produce repaired output.
The inconsistency of the findings of this study with the findings of previous study with respect to the third research question can be explained from two perspectives: first, the ambiguity of recasts have been compensated with adding an attention-getting element (i.e., the teacher's repetition of the erroneous part of the learners' utterances with rising intonation) which highlighted their corrective functions, i. e., in so doing, the teacher's rising intonation signals the learners that something went wrong in their output.Second, in the present study, after the provision of marked reformulations, the teacher paused about three to five seconds to provide the learners with an opportunity to juxtapose their erroneous output and the teacher's highlighted reformulation to notice the gap between them.

Limitations
This study suffers from two limitations.First, due to time and institutional constraints, the size of the sample was small.In fact, they were the only available participants for this study at the moment.Second, as there was no native speaker available, the written DCTs were rated by two non-native speakers although a reliable and valid scale was used to score them.

Conclusion
Based on the results of the present study, it can be concluded that both treatments were effective in raising the learners' pragmatic knowledge to produce appropriate and accurate refusal forms.When comparing the two groups, no significant differences were found between them; that is, similar to elicitation moves, in which the learners had an opportunity to repair their erroneous output, marked recasts were, also, effective to draw the learners' attention to the teacher's reformulated utterance and the discrepancies between their and the teacher's utterances.In other words, while elicitations pushed learners toward self-correction, marked recasts provided learners with both positive and negative evidence of target language and they had opportunity to match and notice the gap between them.The results can have pedagogical implications for language teachers.

Table 4 .
1 Results of the first and second ratings and their average(MRG, Pretest)

2):53-63, 2014 59
Moreover, the correlation between the results of first and second ratings for the pre-and post-tests of both groups was conducted.

Table 4
b. Calculated from dataAs illustrated in Table4.7, the test distribution is normal ( P= .83>0.05).Having made sure about the normality of the distribution, the researcher conducted T-Tests and ANCOVA.Table4.8presents the descriptive statistics of the written DCT as the pretest and posttest scores for both groups of the participants (MRG and EG groups).

Table 4 .
10 Paired Samples Test for EG group