Pragmatic Corrective Feedback in L2 Classrooms: Investigating EFL Teachers' Perceptions and Instructional Practices

Document Type : Original Article

Authors

1 Servat Shirkhani

2 Professor of Applied Linguistics, Allameh Tabataba'i University, Iran

Abstract

Corrective feedback (CF) has shown to be an effective way of developing learners' pragmatic awareness and subsequently pragmatic competence. However, one of the influential factors in the effectiveness of CF is teacher perceptions. On the other hand, teachers' perceptions are not always reflected in their teaching practices. Therefore, this study attempted to investigate the English as a Foreign Language (EFL) teachers' perceptions of pragmatic corrective feedback and to compare their perception with their practice of pragmatic corrective feedback. To achieve this goal, a 44-item questionnaire was developed, piloted, and administered to 300 teachers and analyzed quantitatively in terms of the five components of the questionnaire which dealt with the teachers' perceptions of pragmatic corrective feedback. Furthermore, class recordings of 40 of these teachers were analyzed in terms of their ways of treating pragmatic errors. The findings revealed that the teachers had positive attitudes especially toward the significance of pragmatic corrective feedback, teachers' knowledge and agency about pragmatic corrective feedback and provider of CF, and the way pragmatic corrective feedback should be provided. The results, however, indicated that the teachers' instructional practice of pragmatic corrective feedback was not congruent with their perceptions toward it. The results of the study show the dissonance between teachers' perceptions and instructional practices and the need for teacher education courses for the enhancement of teachers' pragmatics instruction.

Keywords


1. Introduction

A large number of studies (e.g., Lee, 2013; Li, 2010; Lyster, Saito, & Sato, 2013) have provided evidence for the influential role of corrective feedback (CF) in promoting learners' linguistic competence. Likewise, CF can be a good practice for EFL teachers who want to help learners with pragmatic development. The necessity of applying CF to the pragmatic domain of language development has been stressed by Schmidt (1990) and House (1996). Moreover, a number of empirical studies (e.g., Fukuya & Zhang, 2002; Koike & Pearson, 2005; Nipaspong & Chinokul, 2010) have confirmed the effective role of CF in increasing learners' pragmatic awareness. However, teachers' successful implementation of pragmatic corrective feedback is relevant to their attitudes toward the importance of pragmatic corrective feedback and ways of employing it. Thus, this study sought to examine the English as a Foreign Language (EFL) teachers' perceptions about the significance of pragmatic corrective feedback and effective ways of providing such feedback. Furthermore, in order to see how these perceptions are reflected in the teachers' classroom practices, the study examined the extent to which the teachers corrected pragmatic errors and the ways in which they provided the feedback.

2. Literature Review

2.1 Pragmatic Corrective Feedback

The importance of CF, also referred to as negative evidence, in second language acquisition has been stressed by Long (1996) and White (1990). Furthermore, different experimental studies (e.g., Bitchener, Young, & Cameron, 2005; Carroll & Swain, 1993; Li, 2010; Lyster & Saito, 2010; Russell & Spada, 2006; Sato & Lyster, 2012; Trahey & White, 1993) have revealed that CF can have positive effects on learners' language acquisition. Despite the plethora of research on the role of CF in the development of formal aspects of language, little research has addressed the influence of CF in the pragmatic development of language learners. However, emphasizing the necessity of noticing, Schmidt (1990) maintains that the "requirement of noticing is meant to apply equally to all aspects of language" (p. 149). Moreover, House (1996) refers to some evidence showing that the consciousness-raising techniques helpful in developing formal aspects of language may also facilitate learners' pragmatic development. In addition, some experimental studies (e.g., Fukuya & Zhang, 2002; Koike & Pearson, 2005) have revealed that CF can positively influence learners' pragmatic development.

A number of studies (e.g., Alcon Soler, 2000; Hajmalek, 2011; Martinez-Flor & Uso-Juan, 2010; Omar, 1992; Takimoto, 2006; Washburn, 2001) have centered on the role of CF in the development of language learners' pragmatic ability. Nonetheless, none of these studies has used the term pragmatic corrective feedback. Based on the definitions of pragmatics and CF, pragmatic corrective feedback can be defined as any reaction to a learner's utterance which aims to help the learner notice their pragmatic failure and understand what the true form is with regard to the social context in which it is used. Emphasizing the role of CF in learners' development of pragmatic competence, Martinez-Flor and Uso-Juan (2010) state that this CF should address both form and meaning. In fact, one of the concerns of pragmatic corrective feedback is learners' sociopragmatic failure where they interpret a situation differently from a native speaker (NS). Another concern of pragmatic corrective feedback is learners' pragmalinguistic failure. In this case, pragmatic corrective feedback deals with situations when learners have the same understanding of a given context as that of the NSs but do not have enough knowledge of linguistic means to enable them communicate appropriately in that particular context.

2.2 Teachers' Perceptions and Practice of Pragmatic Corrective Feedback

Language teachers' attitudes are of great significance because, as Cook (2002) asserts, these attitudes influence their teaching practices, which, in turn, influence the learning process and outcomes. Moreover, Kennedy and Kennedy (1996) believe that teachers' attitudes are one of the important factors in the implementation of change in language teaching. The study of attitudes is important as attitudes can change through indirect learning or direct influence (Zimbardo & Lieppe, 1991). Obviously, the presence or absence of CF in language classes and the way it is practiced depend on the attitudes of the teachers toward both the necessity of CF and the way it should be carried out. Moreover, provision of CF is not a straightforward activity, but a complex issue with a number of important questions the answers to which vary widely. Questions related to the timing of CF, the type of CF, the types of errors to be corrected, and the degree of explicitness of CF are among the ones which complicate the issue. Teachers' attitudes toward any of these issues can influence the effectiveness of CF.

To our knowledge, no study has directly addressed the teachers' perceptions and practice of pragmatic corrective feedback. Some studies have focused on teachers' attitudes toward CF in general. Schulz (1996), for example, compared the attitudes toward explicit grammar instruction and CF of 92 teachers and 824 students. The results indicated that there were major disagreements among teachers regarding their attitudes toward the significance of explicit grammar instruction and CF though the participating students held positive attitudes toward explicit grammar instruction and error correction during instruction.

Regarding teachers' attitudes toward pragmatics, as Ishihara (2011) states, research into the area of instructional pragmatics does not have a long history. Among the small number of studies on the issue, a few have investigated teachers' attitudes toward pragmatics in teacher education programs. For instance, Vásquez and Sharpless (2009), as part of a larger study, investigated language teachers' attitudes and beliefs toward the necessity of pragmatic courses in MA level TESOL courses. Almost all of the participants in the programs chose to attend applied pragmatics elective courses, while 50 percent or less showed readiness for the theoretical equivalent course. As Vásquez and Sharpless conclude, this is an interesting result indicating that the teachers feel the need to take pragmatic-focused courses and that they prefer applied courses to mere theoretical ones. Moreover, Vellenga (2011) studied teachers' views toward teaching pragmatics. According to Vellenga, instructor responses to demographic questionnaires, comments on lesson checklists, and responses to midteaching and postteaching interviews showed that teachers believed there was value to continuing professional development on the topic of how to teach pragmatics for teachers with a range of previous experience in a variety of contexts. All the six teachers participating in the teaching phase of the study had positive attitudes toward teaching pragmatics and believed that teaching the lessons increased their own knowledge of pragmatics. The results of the study further showed that most teachers were interested in teaching pragmatics and believed that they needed more information about ways of teaching pragmatics in EFL contexts (Vellenga, 2011).

A few studies have compared the teachers' perceptions about CF and their actual practices in treating errors. For instance, Basturkmen et al. (2004) compared three teachers' stated beliefs about focus on form with their practices of focus on form in intermediate ESL classes. The results indicated some mismatches between the teachers' stated beliefs and their classroom practices, especially about the time of correction and correction techniques. Moreover, Méndez and Cruz (2012) examined Mexican teachers' ideas about CF and provision of CF through implementing a semi-structured interview with 5 teachers and administering a questionnaire to 15 teachers. The results were indicative of the teachers' positive perception of oral corrective feedback although some of them viewed CF as optional due to their being concerned with students' feelings. They thought they were mainly using unfocused oral corrective feedback and implicit strategies. In addition, they preferred teacher feedback to peer feedback and considered self-correction as the least popular type of correction.

In addition, Kamiya (2016) studied the relationship between stated beliefs of four English as a second language (ESL) teachers about oral CF and their actual classroom practices. The results showed a congruence between their stated beliefs about CF and their practice of oral CF. According to the findings, oral CF was not of particular concern to the participating teachers because they believed that a main component of success in teaching was creating a comfortable environment for learners. Therefore, they tended to avoid giving CF and where they provided CF, they preferred to use implicit CF. Furthermore, Roothooft (2014) investigated the corrective behavior of 10 EFL teachers in Spain through observing one or two sessions of their classes. Additionally, Roothooft used a questionnaire to examine the 10 teachers’ perceptions about oral CF. The results showed a general mismatch between the teachers' views toward CF and their way of practicing CF. It was shown that the teachers did not have full awareness of their correction practices. They considered CF as an important aspect of teaching; however, they were concerned more with the learners’ emotions and fluency. In another study, Dilāns (2016) compared perceptions of 66 L2 Latvian teachers about providing oral CF to L2 Latvian learners to the actual provision of CF in Latvian classrooms by 13 teachers of Latvian. In this study, four types of oral CF were studied: explicit correction, recasts, elicitation, and repetition. The results of this survey showed that the teachers believed that the distribution of the four types of CF they provided was quite even. However, the study of the classroom practice of the 13 teachers indicated the predominant employment of explicit, isolated recasts over the use of explicit correction, elicitation, integrated recasts, and repetition; 60 percent of the corrections were explicit, isolated recasts.

3. Purpose of the Study

In their intention to increase learners' pragmatic ability, language teachers should incorporate the instruction of pragmatics in their teaching and giving CF can act as one of the useful ways for increasing learners' pragmatic awareness and for teaching pragmatics. However, teachers' perceptions may influence their success in effectively employing CF in pragmatics instruction. On the other hand, the teachers' instructional practices may not necessarily match their perceptions about the how of treating pragmatic errors. For this reason, Basturkmen et al. (2004) argue that investigations of teachers' beliefs should include studying both their stated beliefs and their observed behaviors. As such, this study sought to investigate the EFL teachers' perceptions about different aspects of pragmatic corrective feedback provision and their actual practice of providing such feedback. To achieve this goal, the following two research questions were proposed:

  1.  What are the perceptions of EFL teachers about pragmatic corrective feedback?
  2.  How do the EFL teachers treat pragmatic errors in their classrooms?

4. Method

4.1 Participants

Initially, 345 teachers teaching in different language institutes were selected through convenience sampling as the participants of the first part of the study. Some of these teachers were the second author's MA students majoring in Teaching English as a Foreign Language (TEFL). Many of them were the institute teachers whom the first author and her colleagues, classmates, and students had invited to attend the study. The participants were asked to complete the attitude questionnaire. Of the 345 teachers, 300 returned the completed questionnaire. Of the 300 participants, 284 had filled out the first part of the questionnaire which was related to the teachers' background information. According to the responses, the teachers were between 18 and 44 years old with the mean of 27.5, and their teaching experience ranged from one to 23 years with a mean of six years. As many as 191 (67.3%) of the participants were female and 93 (32.7%) were male. Two hundred and forty-four teachers were English majors, 108 (38%) with a BA degree and 136 (47.89%) with an MA degree. Thirty-six were nonEnglish majors, 18 (6.34%) with a bachelor's degree, and 18 (6.34%) with a master's degree.

Participants of the second part of the study were 40 of the 300 teachers attending the first part. These teachers were selected based on their own and their institute directors' agreements to have their classes be recorded. They were 19 to 35 years of age and had on the average four years of teaching experience. Thirty-two of them were female and eight were male. Also, 19 teachers held a bachelor's degree and 21 held a masters' degree.

4.2 Instrumentation

Two instruments were used to gather the data needed for this study: a pragmatic corrective feedback attitude questionnaire (PCFAQ) and classroom recordings.

Pragmatic corrective feedback attitude questionnaire. The pragmatic corrective feedback attitude questionnaire (PCFAQ) was prepared for the purpose of this study. It was on a five-point Likert scale, ranging from 'strongly disagree' to 'strongly agree.' The questionnaire encompassed a part at the beginning focused on demographic information about the teachers, including gender, age, education, and years of experience. In addition, the questionnaire included 44 items concerned with teachers' attitudes toward the necessity of CF, in general, and pragmatic corrective feedback, in particular, and with whether the teachers think they provide pragmatic corrective feedback in their classrooms and, if they do, how they think they do it.

In order to prepare the questionnaire, at first a pool of items was created in accordance with the main issues in the literature regarding both the provision of CF and pragmatic competence. In relation to CF, mainly the questions posed by Hendrickson (1978) were taken into consideration. Hendrickson’s first question is concerned with whether errors should be corrected or not and if they should be corrected, the next questions deal with when, which errors, how, and who. These five points, that is, efficacy of CF, choice of errors to correct, choice of the feedback provider, choice of CF strategy, and timing of CF are the most controversial issues about CF (Ellis, 2009) which have formed the basis of item construction on CF in this study. For items related to the components of pragmatic competence and sources of pragmatic failure, item development was based on seminal sources, such as Leech (1983) and Thomas (1983). As Taguchi (2009; 2012) states, pragmatic development requires gaining both the knowledge of language-specific linguistic behaviors and that of the sociocultural norms behind those behaviors. Thus, as mentioned by Leech (1983) and Thomas (1983), the learners' pragmatic failure may have pragmalinguistic or sociopragmatic sources.

In accordance with the reviewed literature, the pool of items was based on different aspects of pragmatic corrective feedback, including (a) teachers' attitudes toward the necessity of CF, in general, and pragmatic corrective feedback, in particular; (b) the focus, time (i.e., immediate vs. delayed pragmatic corrective feedback), and degree of explicitness of CF; (c) reasons for failure of pragmatic corrective feedback; (d) CF provider (i.e., learner, peers, or teacher); (e) teachers' sources of knowledge about pragmatic corrective feedback; and (f) their current practice of pragmatic corrective feedback in their classes. The items were then discussed with a professor of applied linguistics with expertise in both pragmatics and corrective feedback. Following expert judgment, a total of 61 items were chosen to go through further steps of the ongoing piloting process, as Dornyei (2003) refers to. The items of the questionnaire were discussed with five English teachers who were teaching for 10 years on the average; two of these teachers held MA degrees and three PhD degrees. The feedback of these teachers was considered in rewording a few of the items. Then, the questionnaire was given to eight language teachers teaching at language institutes at the intermediate level and representing the target population. They were asked to fill out the questionnaire in the presence of one of the researchers and demonstrate any probable ambiguities they would encounter. They were also asked to give their ideas for the improvement of the questionnaire. After revising the questionnaire based on the feedback received from these eight people, the questionnaire went through the piloting process.

The aim of the pilot stage was to investigate both the total questionnaire and item characteristics. At this stage, the questionnaire was administered to 69 EFL teachers representing the target population and the elicited data were analyzed using SPSS 20. For the purpose of examining questionnaire characteristics, the reliability of the questionnaire (Cronbach's alpha) and item-to-test correlation were calculated. Based on the results, Cronbach's alpha was 0.77. Furthermore, item-to-test correlation did not suggest the necessity of omitting any of the items. So, none of the items was omitted based on the result of the total questionnaire analysis.

For individual items, descriptive statistics including mean, standard deviation, range, skewness, and kurtosis were calculated. In addition, the number of missing data for each of the items was examined for any probable problem such as ambiguity of the items. The examination of mean, standard deviation, range, skewness, and kurtosis suggested no specific problem except for item 5 and item 28. These two items were thus omitted, resulting in a 59-item questionnaire.

To examine the construct validity of the questionnaire, a factor analysis was carried out. Data from the 300 respondents were fed into SPSS (the total number was 307; seven questionnaires were omitted because of the missing data). First, descriptive statistics and KMO and Bartlett's Test of Spherity were calculated to see whether the data suit factor analysis. The Determinant= 1.109E-010 and results of descriptive statistics and KMO and Bartlett's Test of Spherity suggested that the data suited factor analysis. The scree plot suggested five factors. The five factors constituted 34.1% of the total variance. On the whole, of 59 items, 44 loaded on the five factors and 15 (i.e., 25% of the original items) were omitted; items with loadings lower than 0.4 were not included in any of the factors. Based on the content of the factors, they were named as follows but they were presented randomly through the questionnaire to get more reliable responses:

Component 1: Significance of pragmatics and pragmatic corrective feedback (12 items);

Component 2: Teacher's knowledge and agency about pragmatic corrective feedback and provider of CF (9 items);

Component 3: Ways to practice pragmatic corrective feedback (8 items);

Component 4: Reasons for failure of pragmatic corrective feedback (8 items);

Component 5: Time and degree of explicitness of pragmatic corrective feedback (7 items).

Observation recordings. A transcript of 128.12 hours of recorded classroom sessions consisting of recordings of two class sessions for each of the 40 teachers participating in the second part of the study was used to show the instructional practices of the teachers in giving pragmatic correct feedback in EFL classes. A detailed transcription of pragmatic corrective feedback moves in the audio-recorded materials for each teacher during the two sessions was prepared. The transcription was concerned with the frequencies of treated pragmatic errors, frequencies and types of CF, the number of moves in treating the errors, and the possible uptake following the feedback.

4.3 Data Collection and Analysis

The first step in this study was to develop and pilot the PCFAQ. The next step was to select the participants of the study. Then, the selected questionnaire was given to all the participants to examine their attitudes toward the necessity and ways of practicing pragmatic corrective feedback in EFL classes. To answer the first question of the study, which addressed the perceptions of teachers about pragmatic corrective feedback, the frequencies and percentages of the teachers selecting each of the five scales for each of the 44 items of the questionnaire were calculated. Then, the results were interpreted based on the five components of the PCFAQ.

To answer the second question, 40 of the teachers attending the first phase of the study were selected and their classes were recorded for two sessions each. Then, transcriptions of the recordings were prepared and analyzed. The analysis included computing the frequencies of the types of CF, the number of moves in treating pragmatic errors, and the frequencies of each uptake condition. The defined uptake conditions included: (a) Uptake: where the CF move was followed by the correct form by the learner; (b) No evidence of uptake: where neither there was immediate uptake nor was the error repeated by the same learner; and (c) No: where the moves following the CF move contained either the same error or similar errors showing that the corrected feature has not been understood by the learner.

5. Results

5.1 Teachers' Perceptions of Pragmatic Corrective Feedback

To examine EFL teachers' attitudes toward pragmatic corrective feedback, the frequency with which each of the five components of the questionnaire was chosen by the participating teachers was calculated for all the 44 items of the questionnaire. The scale frequencies were then interpreted according to the five components of the PCFAQ. The results for each of the components are described as follows:

Significance of pragmatics and pragmatic corrective feedback.The first component of the PCFAQ, which was related to the significance of pragmatics and pragmatic corrective feedback, entailed 12 items of the questionnaire. The number of the items and the frequencies and percentages of the teachers selecting any of the five points on the Likert scale for these items are presented in Table 1.

Table 1

Frequencies and Percentages of the Items in the First Component

                                
  

Items

  
  

Strongly    disagree

  
  

Disagree

  
  

Undecided

  
  

Agree

  
  

Strongly agree

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  1. Teachers should   pay attention to pragmatic errors in their classes.

0

(0)

9

(3)

11

(3.7)

165

(55)

115

(38.3)

  1. Pragmatic errors   should be corrected in EFL classes.

0

(0)

9

(3)

31

(10.3)

178

(59.3)

82

(27.3)

  1. Pragmatic errors   are important in communication.

0

(0)

12

(4)

23

(7.7)

156

(52)

109

(36.3)

  1. Pragmatic errors   should be incorporated in assessment of EFL learners.

5

(1.7)

19

(6.3)

66

(22)

161

(53.7)

49

(16.3)

  1. Learners will   learn to use the language appropriately if their pragmatic errors are   corrected.

0

(0)

17

(5.7)

45

(15)

166

(55.3)

72

(24)

  1. Pragmatic   competence helps learners communicate more effectively in the second   language.

0

(0)

8

(2.7)

25

(8.3)

157

(52.3)

110

(36.7)

  1. There should be   more pragmatic corrective feedback than is presently the case in EFL classes.

0

(0)

20

(6.7)

79

(26.3)

154

(51.3)

47

(15.7)

  1. Pragmatic   corrective feedback is important for language learners.

0

(0)

11

(3.7)

29

(9.7)

175

(58.3)

85

(28.3)

  1. Pragmatic   corrective feedback should deal with errors related to the violation of   sociocultural norms of English such as politeness.

3

(1)

20

(6.7)

57

(19)

168

(56)

52

(17.3)

  1. Pragmatic   corrective feedback should be provided at intermediate levels.

6

(2)

53

(17.7)

40

(13.3)

168(56)

33(11)

  1. Pragmatic   corrective feedback should be provided at advanced levels.

16

(5.3)

47

(15.7)

25

(8.3)

144

(48)

68

(22.7)

  1. I give pragmatic   corrective feedback at advanced levels.

4

(1.3)

23

(7.7)

70

(23.3)

114

(38)

89

(29.7)

Note.   No.= Frequency, %= percent

Based on Table 1, more than 60 percent of all participating EFL teachers agreed or strongly agreed with these statements. Therefore, a large percentage of teachers agreed that pragmatic competence and pragmatic corrective feedback were important (Item 18). Eighty-nine percent agreed that pragmatic competence helped learners communicate more effectively in the second language (L2) (Item 6), 88 percent thought pragmatic errors were important in communication (Item 3), and 79 percent considered pragmatic corrective feedback helpful to learners in learning to use language appropriately (Item 5). Moreover, 93 percent agreed that teachers should pay more attention to pragmatic errors (Item 1), 87 percent thought pragmatic errors should be corrected (Item 2), and 67 percent believed that there should be more pragmatic corrective feedback in EFL classes than is presently the case (Item 9). Furthermore, 70 percent believed that pragmatic corrective feedback should be incorporated into the assessment of EFL learners (Item 4) and 73 percent thought that pragmatic corrective feedback should deal with sociopragmatic errors (Item 20). In addition, 67 and 71 percent of the teachers agreed on the necessity of pragmatic corrective feedback in intermediate (Item 24) and advanced (Item 25) classes, respectively, and 68 percent claimed that they were providing pragmatic corrective feedback at advanced levels (Item 43).

Teacher's knowledge and agency about pragmatic corrective feedback and provider of CF.The second component, which dealt with teachers' knowledge and agency about pragmatic corrective feedback and provider of CF, contained nine items of the questionnaire. Table 2 shows the number of these items and the frequencies and percentages with which the teachers chose any of the scales of these items.

Table 2

Frequencies and Percentages of the Items in the Second Component

                                
  

Items

  
  

Strongly    disagree

  
  

Disagree

  
  

Undecided

  
  

Agree

  
  

Strongly agree

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  1. When learners make errors in   speaking, they should be corrected.

5

(1.7)

44

(14.7)

43

(14.3)

159

(53)

49

(16.3)

  1. Teachers should correct   learners' pragmatic errors.

14

(4.7)

48

(16)

28

(9.3)

161

(53.7)

49

(16.3)

  1. Teachers themselves should   provide the pragmatic corrective feedback.

3

(14)

56

(46.3)

60

(20)

139

(18.7)

42

(1)

  1. Most learners dislike it when   their peers correct them in class.

10

(13.7)

48

(41.7)

76

(25.3)

125

(16)

41

(3.3)

  1. Teachers' experience as language   learners can be a source of their knowledge about pragmatic corrective   feedback.

6

(2)

25

(8.3)

68

(22.7)

151

(50.3)

50

(16.7)

  1. Teachers' experience as language   teachers can be a source of their knowledge about pragmatic corrective   feedback.

4

(1.3)

17

(5.7)

49

(16.3)

174

(58)

56

(18.7)

  1. Teacher training courses can be   a source of teachers' knowledge about pragmatic corrective feedback.

10

(3.3)

34

(11.3)

36

(12)

175

(58.3)

45

(15)

  1. Relevant books and articles can   be a source of teachers' knowledge about pragmatic corrective feedback.

3

(1)

13

(4.3)

50

(16.7)

194

(64.7)

40

(13.3)

  1. I tell my learners about the   importance of pragmatic corrective feedback.

13

(4.3)

58

(19.3)

99

(33)

86

(28.7)

44

(14.7)

Note.   No.= Frequency, %= percent

Analysis of results in Table 2 showed that more than 60 percent of all teachers agreed with all items in this component except item 37. This item read as "I tell my learners about the importance of pragmatic corrective feedback." Of all the teachers, 23.7 percent disagreed or strongly disagreed with this statement, 33 percent could not decide, and 43.4 agreed or strongly agreed with this statement.

Regarding teachers' knowledge, most of the teachers thought that teachers' experience as language teachers (Item 28), their experience as language learners (Item 27), teacher education courses (Item 29), and relevant books and articles (Item 30) could be sources of teachers' knowledge about pragmatic corrective feedback. Based on the results of other items in this component, a large percentage of the teachers agreed that learners' errors in speaking (Item 7) and learners' pragmatic errors (Item 8) should be corrected. However, they were in agreement that most learners disliked their peers' correction of errors (Item 16) and, thus, believed that teachers themselves should provide pragmatic corrective feedback (Item 10).

Ways to practice pragmatic corrective feedback.The third component was mostly related to the ways of providing pragmatic corrective feedback and consisted of items 35, 41, 36, 44, 42, 32, 23, and 33 for which the frequencies and percentages are given in Table 3.

Table 3

Frequencies and Percentages of the Items in the Third Component

                                
  

Items

  
  

Strongly    disagree

  
  

Disagree

  
  

Undecided

  
  

Agree

  
  

Strongly agree

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  1. Pragmatic corrective feedback   should be provided at elementary levels.

15

(5)

70

(23.3)

56

(18.7)

131

(43.7)

28

(9.3)

  1. I provide my learners with   opportunities to notice their pragmatic errors themselves.

5

(1.7)

28

(9.3)

94

(31.3)

126

(42)

47

(15.7)

  1. I pay attention to learners’   pragmatic errors.

1

(0.3)

17

(5.7)

52

(17.3)

125

(41.7)

105

(35)

  1. I provide my learners with   opportunities to correct their pragmatic errors themselves.

12

(4)

29

(9.7)

102

(34)

103

(34.3)

54

(18)

  1. I provide my learners with   opportunities to correct each other’s pragmatic errors.

20

(6.7)

76

(25.3)

106

(35.3)

65

(21.7)

33

(11)

  1. I give pragmatic corrective   feedback at elementary levels.

28

(9.3)

76

(25.3)

89

(29.7)

75

(25)

32

(10.7)

  1. I give pragmatic corrective   feedback at intermediate levels.

7

(2.3)

29

(9.7)

94

(31.3)

122

(40.7)

48

(16)

  1. I incorporate aspects of   pragmatics in my classroom assessment.

12

(4)

43

(14.3)

108

(36)

112

(37.3)

25

(8.3)

Note.   No.= Frequency, %= percent

As shown in Table 3, 77 percent of the teachers agreed or strongly agreed that they paid attention to learners' pragmatic errors (item 33). Between 50 and 60 percent of the teachers agreed with four items in this component (35, 42, 32, & 23), indicating that they believed pragmatic corrective feedback should be provided at elementary levels (Item 23), that they provided pragmatic corrective feedback at intermediate levels (Item 42), and that they provided their learners with opportunities to notice (Item 32) and correct their pragmatic errors (Item 35). Less than 50 percent agreed that they gave pragmatic corrective feedback at elementary levels (Item 41), that they
incorporated aspects of pragmatics in their assessment (Item 44), and that they provided their learners with opportunities to correct each other's pragmatic errors (Item 36).

Reasons for failure of pragmatic corrective feedback.Reasons for the failure of pragmatic corrective feedback were the focus of the fourth component of the attitude questionnaire, which comprised eight items of the questionnaire. The frequencies and percentages of these items are given in Table 4.

 

Table 4

Frequencies and Percentages of the Items in the Fourth Component

                                
  

Items

  
  

Strongly disagree

  
  

Disagree

  
  

Undecided

  
  

Agree

  
  

Strongly agree

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  1. When provided with pragmatic   corrective feedback, most learners do not notice that they are being   corrected.

16

(5.3)

111

(37)

71

(23.7)

87

(29)

15

(5)

  1. When provided with pragmatic   corrective feedback, most learners notice that they are being corrected but   do not get the pragmatic point in the feedback.

6

(2)

69

(23)

99

(33)

108

(36)

18

(6)

  1. Most nonnative English speaker   teachers cannot provide pragmatic corrective feedback.

18

(6)

113

(37.7)

72

(24)

86

(28.7)

11

(3.7)

  1. Most nonnative English teachers   do not have enough competence about the right linguistic forms for performing   different speech acts (e.g., apology, request, complaint, etc.) to give   learners pragmatic corrective feedback.

18

(6)

120

(40)

73

(24.3)

79

(26.3)

10

(3.3)

  1. Most nonnative English teachers   do not have enough competence about sociocultural norms of English to give   learners pragmatic corrective feedback.

6

(8)

54

(44.7)

82

(27.3)

134

(18)

24

(2)

  1. Because of my insufficient   knowledge of sociocultural norms of English, I face problems in correcting my   learners' pragmatic errors.

35

(6.3)

110

(12.7)

98

(32.7)

38

(36.7)

19

(11.7)

  1. Because of my insufficient   knowledge of correct linguistic forms for performing different speech acts   (e.g., apology, request, complaint, etc.), I face problems in correcting my   learners' pragmatic errors.

9

(3)

33

(11)

69

(23)

138

(46)

51

(17)

  1. Because of my learners'   inability to get the points in the corrective feedback I give them, I face   problems in correcting my learners' pragmatic errors.

31

(1.3)

99

(14.7)

122

(40.7)

44

(33)

4

(10.3)

Note. No. = Frequency, %=   percent.

As displayed in Table 4, more than 60 percent of the teachers admitted that they faced problems in correcting their learners' pragmatic errors due to their insufficient pragmalinguistic (Item 39) and sociopragmatic (Item 38) knowledge. Regarding other nonnative English teachers, however, less than 50 percent agreed that they could not provide pragmatic corrective feedback (Item 21) and that the reason was their insufficient pragmalinguistic knowledge (Item 22). On the other hand, more than 60 percent agreed with the statement that most nonnative English teachers did not have enough sociopragmatic knowledge to correct learners' pragmatic errors (Item 26). Furthermore, less than 45 percent agreed that when provided with pragmatic corrective feedback, learners did not notice they were being corrected (Item 14) or did not get the pragmatic point in the feedback (Item 17) and that learners' inability to get the points in the CF faced teachers with problems in correcting their pragmatic errors (Item 40).

 Time and degree of explicitness of pragmatic corrective feedback. The last component of the questionnaire was concerned with the time of providing pragmatic corrective feedback and the degree of explicitness of CF. The seven items of this component and their frequencies are displayed in Table 5.

Table 5

Frequencies and Percentages of the Items in the Fifth Component

                                
  

Items

  
  

Strongly    disagree

  
  

Disagree

  
  

Undecided

  
  

Agree

  
  

Strongly agree

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  

No.

  

(%)

  
  1. Teachers should not disrupt the   flow of communication to correct learners’ pragmatic errors.

1

(0.3)

33

(11)

51

(17)

156

(52)

59

(19.7)

  1. Teachers should provide   pragmatic corrective feedback immediately after the error has been made.

22

(7.3)

76

(25.3)

60

(20)

116

(38.7)

26

(8.7)

  1. The pragmatic corrective   feedback should be implicit and indirect.

7

(2.3)

75

(25)

65

(21.7)

115

(38.3)

38

(12.7)

  1. Teachers should provide learners   with explicit pragmatic corrective feedback.

10

(7.7)

53

(53.7)

53

(17.7)

161

(17.7)

23

(3.3)

  1. Teachers should provide learners   first with implicit and indirect pragmatic corrective feedback and then if necessary   provide more direct and explicit pragmatic corrective feedback.

1

(0.3)

39

(13)

37

(12.3)

169

(56.3)

54

(18)

  1. I give explicit pragmatic   corrective feedback.

27

(9)

111

(37)

109

(36.3)

49

(16.3)

4

(1.3)

  1. I provide my learners with   immediate pragmatic corrective feedback.

22

(7.3)

69

(23)

125

(41.7)

71

(23.7)

13

(4.3)

Note.   No.= Frequency, %= percent

As to the time of giving pragmatic corrective feedback, Table 5 shows that more than 60 percent of the teachers agreed that the flow of communication should not be disrupted for correcting learners' pragmatic errors (Item 11). Therefore, less than 50 percent of them agreed that teachers should provide immediate CF (Item 12) and 30 percent stated that they provided immediate CF in their classes (Item 34). Regarding the degree of explicitness of CF, the results were divergent. Fifty-one percent agreed that pragmatic corrective feedback should be implicit (Item 13). On the other hand, more than 60 percent thought teachers should give explicit pragmatic corrective feedback (Item 15) and admitted that they provided explicit pragmatic corrective feedback (Item 31). More than 60 percent believed that teachers should provide learners first with implicit and indirect pragmatic corrective feedback and then, if necessary, provide more direct and explicit pragmatic corrective feedback (Item 19).

5.2 Teachers' Practices of Pragmatic Corrective Feedback

To see how teachers provided pragmatic corrective feedback in their classrooms, the frequencies of both pragmatic and linguistic errors treated by the teachers were computed. The results, displayed in Table 6, indicated that of the 1898 errors treated, only one percent (19 errors) were related to pragmatics (i.e., 15 pragmalinguistic & 4 sociopragmatic errors) and 99 percent (1879 errors) were linguistic.

Table 6

Descriptive Statistics for Types of Errors Addressed by Teachers     in Each Session

              
  

 

  
  

Minimum

  
  

Maximum

  
  

Sum

  
  

Percent

  

Grammar

0

33

668

35.1

Vocabulary

0

15

331

17.2

Pronunciation

0

50

880

46.7

Total LCF

2

93

1879

99

Pragmalinguistic

0

3

15

.8

Sociopragmatic

0

2

4

.2

Total PCF

0

3

19

1

At the next step, the types of CF used to address the pragmatic errors, the numbers of moves in treating these errors, and the conditions of uptake were determined. The results are depicted in Table 7, in which the teacher column shows the number used to code the teacher providing the pragmatic corrective feedback and the session column indicates the first or the second session of recordings for each teacher.

Table 7

Descriptive Statistics for Pragmatic Error Treatment

                 
  

Error

  
  

CF

  
  

Number    of Moves

  
  

Uptake

  
  

Teacher

  
  

Session

  

PS

EC

3

Yes

4

1

PS

EC

3

Yes

4

1

PP

EC

3

Yes

4

1

PP

EC

3

Yes

4

1

PP

EC

2

No evidence

4

2

PP

EC

2

No evidence

4

2

PP

EC

3

Yes

4

2

PP

ML

2

No evidence

7

1

PP

EC

2

No evidence

7

1

PS

EC

2

No evidence

7

2

PP

EC

6

Yes

14

1

PP

EC

5

Yes

15

2

PP

EC

5

Yes

17

2

PP

EC

2

No evidence

17

2

PP

EC

2

No evidence

17

2

PS

ML

3

Yes

30

1

PP

EC

2

No evidence

31

1

PP

EC

3

Yes

36

1

PP

EC

5

Yes

40

2

Note. EC= explicit correction, ML= metalinguistic feedback; PP =   pragmalinguistic, PS = sociopragmatic.    

As displayed in Table 7, only 9 of the 40 teachers (coded as 4, 7, 14, 15, 17, 30, 31, 36, & 40) provided their learners with pragmatic corrective feedback. The next point is that of the 19 pragmatic errors, 2 (10.53%) were corrected through metalinguistic feedback and the other 17 (89.47%) were addressed through explicit correction. In addition, the sum of the number of moves was 58, with a mean of three. Of the 19 pragmatic errors, eight (42.1%) were treated through two moves, seven (36.8) through three moves, and only four (21.1%) through five and six moves. Furthermore, for 11 (57.9% of the) cases, there was evidence of immediate uptake while for eight (42.1%) there was no evidence of uptake.

6. Discussion

The findings indicated that the 300 participating teachers had positive attitudes toward pragmatic corrective feedback, especially toward three of its components, that is, significance of pragmatic corrective feedback, teachers' knowledge and agency about pragmatic corrective feedback and provider of CF, and the way pragmatic corrective feedback should be provided. Analysis of individual items within each of the five components of the questionnaire indicated that EFL teachers believed that pragmatic competence is helpful to learners’ successful communication in the L2 and pragmatic corrective feedback is necessary for pragmatic development. The teachers thought that pragmatic corrective feedback should deal with sociopragmatic errors and should be provided to learners at all levels of proficiency. Nonetheless, they agreed that the flow of communication should not be disrupted for providing pragmatic corrective feedback and, thus, maintained that pragmatic corrective feedback should not be immediate. In addition, they thought that only if implicit CF is not effective, explicit correction should be provided. With regard to their practice of pragmatic corrective feedback, they stated that they paid attention to their learners' pragmatic errors and provided them with pragmatic corrective feedback. They believed that in treating pragmatic errors, they provided their learners with opportunities to notice and correct their own errors. However, they admitted that they faced problems in giving pragmatic corrective feedback due to their insufficient pragmalinguistic and sociopragmatic knowledge. Therefore, they expressed the need for teacher education programs on teaching pragmatics and agreed that teacher education courses can be sources of knowledge about pragmatic corrective feedback.

On the other hand, the analyses of the teachers' practice of pragmatic corrective feedback revealed that the teachers were not practicing what they thought should be done or believed they were doing. First, they were not providing pragmatic corrective feedback sufficiently; only one percent of the corrections were directed at pragmatic errors. Second, they used explicit techniques of correction to correct all the pragmatic errors though they thought explicit correction should be the last resort. Third, in 42 percent of the cases, the number of moves in treating the pragmatic errors was two, consisting of just the error by the learner and the correction by the teacher, with no evidence of uptake. This indicates that the teachers were not concerned with the provision of opportunities for self-correction while in expressing their attitudes, the teachers  stated they should help learners with noting and self-correcting the errors. It further means that the teachers did not attend to the uptake of the correction.

No study on the attitudes of teachers toward pragmatic corrective feedback was found in the literature. However, Vásquez and Sharpless (2009), in their analysis of teachers' attitudes toward the necessity of pragmatic-focused courses in MA level TESOL programs, concluded that the teachers think attending pragmatic courses, especially applied courses in contrast to mere theoretical ones, is essential for them. Moreover, studying teachers' views about teaching pragmatics, Vellenga (2011) found that the participating teachers believed in the necessity and usefulness of teacher development in teaching pragmatics. In this way, the results of the present study are in agreement with Vásquez and Sharpless (2009) and Vellenga’s (2011) findings that teachers have positive attitudes toward programs aimed at increasing their knowledge of pragmatics instruction.

As to teachers' attitudes toward the significance of CF, the results of the present study do not confirm Kamiya's (2016) finding that teachers are not concerned with providing CF. The findings, however, conform to Lee's (2013) conclusion that teachers are aware of the importance of providing CF though they believe they should not correct all learners’ errors. The results are also congruent with the finding by Méndez and Cruz (2012). They showed that teachers held positive perceptions about oral CF; however, they thought teachers should not prefer giving feedback to caring about students’ feelings. 

Concerning teachers' perceptions about the implicitness/explicitness of CF, the findings support Kamiya's (2016) report that teachers believe CF should be implicit where it must be provided because teachers must be more concerned with creating a comfortable learning environment. The present study shows that a large percentage of teachers think they should first give implicit CF and then, if it is not effective, switch to more explicit types of CF. Teachers' preference for implicit over explicit CF was further supported by Dilāns (2016). Dilāns found that teachers believed they used all the four types of CF under study with almost the same frequency. The CF types studied were explicit correction, which is considered an explicit type of CF, and recasts, elicitation, and repetition, which are classified as implicit CF by Ellis (2009). The even distribution of the four types of CF in this study indicates that the teachers believed they employed implicit CF three times as much as they used explicit CF. Lee's (2013) study provides additional support for teachers’ preference of implicit CF. Lee reported that teachers believed they preferred to give implicit CF even though they stated that in practice they used explicit CF more. Méndez and Cruz (2012), too, found that implicit CF strategies were more favored than explicit ones by the participating teachers.

In relation to the immediacy of CF, in this study teachers did not agree with immediate feedback as they believed they should avoid the interruption of communication in the classroom. The results in this case are in line with Han and Jung's (2007) and Roothooft's (2014) findings. Han and Jung found that teachers believed that correcting errors might result in the interruption of learner speech and hinder learners' confidence in speaking freely. In the same vein, examining the stated beliefs of ten teachers about CF, Roothooft reported that in spite of admitting the importance of CF in language teaching, the teachers expressed worries about providing immediate feedback for two reasons. They thought immediate CF could disturb the flow of communication and that it could lead to negative emotional reactions in students. Lee (2013), on the other hand, reported that teachers thought students' repair of the error and practice of the correct form occurred mostly after immediate CF and that immediate correction could enhance learners’ oral proficiency. With regard to the provider of CF, Méndez and Cruz (2012) found that the teachers considered teacher feedback more appropriate than peer feedback and that for them self-correction was the least popular CF type. The results of the present study contrast with this as only 20 percent of the teachers were in favor of teacher correction and that 19 percent of the teachers thought that learners disliked peer correction.

The incongruence found in this study between the teachers' perceptions and their CF practice is supported by Basturkmen et al. (2004). They found that the teachers' time of correction and use of CF types were not in accordance with their stated beliefs about the time of correction and types of correction techniques. The results further corroborate Dilāns' (2016) finding that there are discrepancies between the CF types the L2 Latvian teachers think they use and those the teachers actually use in practice. Nonetheless, the results contravene Kamiya's (2016) finding about the positive relationship between the stated beliefs of ESL teachers about oral CF and their practice of such CF.

The results of this study and the few other studies with similar foci (e.g., Basturkmen et al., 2004; Dilāns, 2016) demonstrate a mismatch between teachers’ perceptions of CF and their CF practice.  Various factors may lead to this mismatch. One reason for the mismatch could be the ignorance of second language acquisition research to take into account the reality of the classroom context (Mori, 2011). Another reason might be related to Ellis' (2013) argument that teacher guides do not take into account the findings of CF research. The results of this study suggest that teachers' perceptions are largely consistent with recent research findings; however, these perceptions are not fully reflected in their practice of CF. Other possible reasons, as Roothooft (2014) has suggested, could be the complexity of perceptions, on the one hand, and unplanned nature of CF, on the other. Although the reasons for the incongruence between the perceptions and practices of teachers are not the focus of this study, consideration of the possible reasons has implications for researchers and language teachers. It can suggest, as Mori (2011) havs pointed out, that the the teachers' practice of CF should be considered within the teaching context. This calls for teachers’ reflections on what other factors can contribute to their ways of implementing CF and leads to more teacher awareness concerning the relationship between their perceptions and teaching practices. Teachers' reflection and awareness can convince them to regard the incongruence as an opportunity rather than a problem as argued by Dilāns (2016). 

7. Conclusion

In accordance with the results of the studies by Vásquez and Sharpless (2009) and Vellenga (2011), the findings of this study demonstrate that teachers have positive attitudes toward the incorporation of a pragmatics section in teacher education programs. The positive attitudes of teachers, who are regarded as the primary agents in developing learners' pragmatic competence (Ishihara, 2011), indicate both the necessity of pragmatics-focused teacher education programs and the high probability of success of such programs.

According to the results of the study, the EFL teachers teaching in language institutes believe that they cannot successfully teach pragmatics and cannot identify or correct pragmatic errors because they do not have sufficient pragmatic competence. The first suggestion of this finding might be the teachers' need for courses targeting the development of their pragmatic competence. The second, and more important, implication of these attitudes, however, is that teachers need teacher training courses aimed at increasing teacher awareness about pragmatics and pragmatic instruction. Higher awareness about pragmatic development encourages teachers to embark on providing pragmatic corrective feedback. The reason is that this awareness can give them the insight that they are not supposed to be fully competent in L2 pragmatics to deal with pragmatic errors; rather, they can focus on pragmatic awareness-raising in their learners. Therefore, courses are needed to remind teachers that they can help their learners with developing the pragmatic aspects of language even when they are not certain of the inappropriateness of the language their learners use in the classrooms.

Analysis of the responses to the PCFAQ suggests that the EFL teachers believe that they cannot teach pragmatics and that they are not successful pragmatic corrective feedback providers because of their pragmatic incompetence. Thus, teachers' levels of pragmatic competence and pragmatic awareness can direct future studies. In addition, more studies are required to look at the relationship between teachers' pragmatic competence and their way of treating pragmatic errors in their classes. Furthermore, comparison of teachers' perceptions to their practice of pragmatic corrective feedback shows that the teachers are not providing pragmatic corrective feedback as much as they think they do or should do. Neither are the teachers providing pragmatic corrective feedback in the ways they think increase the effectiveness of CF. More studies centering on the teachers' treatment of pragmatic errors in other and similar situations can be helpful in raising the teachers' awareness of their practice and the discrepancies between their perceptions and practices of pragmatic corrective feedback. Furthermore, as the incongruence in teachers’ stated beliefs and practices has been attributed to factors such as contextual factors (e.g., Borg, 2003) or teacher experience (as shown by Basturkmen et al. 2004), future studies on the issue can shed light on the factors contributing to the relationship between teachers' perceptions and practices of pragmatic corrective feedback.

It should be noted that this study has a few limitations. First, because it was a large-scale study, it was not feasible for the researchers to control the possibly influential individual factors. However, studies (e.g., Junqueira & Kim, 2013) have indicated that individual factors, such as teaching experience and previous trainings, influence the way teachers practice and consider CF. Therefore, further studies comparing attitudes of different groups of language teachers, for example, experienced versus inexperienced teachers, toward pragmatics and pragmatic corrective feedback would be insightful. The second limitation concerns the issues covered in the pragmatic corrective feedback questionnaire. Due to the variety of issues related to pragmatic corrective feedback, a small number of items were related to each issue, like significance, timing, and explicitness of CF. In view of this shortcoming, future studies should address each of these issues with narrower and deeper scopes.