EFL Teachers’ Cognition of Teaching English Pronunciation Techniques: A Mixed-Method Approach

In recent years, a great number of attempts have been made on teachers’ cognition with the aim of understanding the complications reinforcing the teachers’ cognitions and their classroom practices. Such studies shed light on how teachers’ cognitions expand over time and how they are reflected in their classroom practices. The aim of the present study was to investigate Iranian EFL teachers’ cognition particularly in terms of the pronunciation techniques they apply in the oral communication classrooms and their knowledge about their language learners’ characteristics. To achieve the goals of the study, the cognitions of five English teachers in the oral communication classrooms were explored. The teachers were requested to answer two semi-structured interviews to obtain the data about their cognitions regarding the pronunciation techniques. Furthermore, their students were asked to fill out a questionnaire to express their opinions about the techniques applied by their teachers during instruction of English pronunciation. The qualitative and quantitative results showed that there was an intricate relationship between language teachers’ experience with their cognitions about their language learners. Moreover, those teachers who were in higher level language courses showed to have broader cognitions about both the techniques they used in classrooms and the language learners’ characteristics as well.


Introduction
In recent years, a great number of attempts have been made on teachers' cognition with the aim of understanding the complications reinforcing the teachers' cognitions and their classroom practices (Baker, 2014). Based on Borg's (2006) definition, second language teacher's cognition (SLTC) is "an often tacit, personally-held practical system of mental constructs held by teachers and which are dynamic-i.e. defined and refined on the basis of educational and professional experiences throughout teachers' lives" (p. 35). The perception of teacher cognition includes a variety of notions like teachers' knowledge, perceptions, beliefs and attitudes towards their actual performances and practices in a specific context.
The researchers and experts of teacher education have paid a large amount of attention to SLTC. The data reported in the literature shed more light on the modules and components of teachers' beliefs and knowledge, the way their cognitions have extended and how they are expressed in their classroom practices (Borg & Burns, 2008;Borg, 2006).
Taking previous studies in the field as a starting point, the present study was set out to analyze Iranian EFL teachers' cognition (knowledge, beliefs, perception, and attitude) about techniques for teaching English pronunciation in their oral communication classrooms. Furthermore, the study aimed to explore language teachers' cognitions about the characteristics of their language learners. This is the time to take a sober look at the way pronunciation is taught in English classrooms in Iran.

Literature Review
This part establishes a theoretical framework of the study by providing a theoretical background in depicting the historical legacy of the concepts and theories underlying the pillars of teachers' cognitions in one phase, and the related studies which draws a picture of influential studies done by Iranian and non-Iranian researchers in the other phase. What follows is a review of empirical studies done in the area of teachers' cognition.
Chart 1. The key themes in the educational literature regarding the elements of teachers' cognition Bailey (1996) reported a research project in which seven MA candidates and a teacher educator investigated the role of their language learning experiences in shaping their current teaching practices and philosophies through autobiographical writing and reflection on it. As a result, the writers recognized several factors related to teaching and learning contexts which had made their own language learning experiences positive: "(1) teacher personality and style mattered more than methodology; (2) teachers were caring and committed, and had clear expectations of their students; (3) teachers respected, and were respected by the students; (4) as students, their motivation to learn enabled them to overcome inadequacies in the teaching; and (5) learning was facilitated by a positive classroom environment" (p. 20).
The authors reworded the findings, quoting a comparable study by Freeman (2002, p. 11), that "the memories of instruction gained through their 'apprenticeship of observation' function as de facto guides for teachers as they approach what they do in the classroom". However, the analyses of actual practices were not reported in this report. Numrich (1996), who was working with beginner teachers, realized that teachers agreed to develop or avoid some specific instructional strategies based on their positive or negative experiences of these respective strategies as learners. For example, 27% of the teachers stated in their diaries reflections that they tried to incorporate a cultural element into their teaching process because they had figured out that learning about the L2 culture was a pleasurable part of their L2 own learning experiences. In contrast, the teachers notified that they avoided to teach grammar or to correct errors because these domains of L2 instruction had been considered as being negative experiences for them. Regarding the latter negative experience, Numrich reported that: "Error correction was most often cited as a technique that had been used by their language teachers and that had inhibited them from speaking. In some cases it had even turned them off to [sic] language learning because they had felt so humiliated and uncomfortable being corrected. Because of the negative experiences of being corrected, several teachers chose not to interrupt their students' flow of speech in the classroom to correct errors" (p. 139).
In investigating teachers' beliefs about teaching grammar, Eisenstein-Ebsworth and Schweers (1997) study also resulted in the fact that teachers' experiences as language learners were of great importance. One teacher, for instance, described his/her experience as "my own education included very formal language study including memorization, reading, writing, and grammar. Now I'm using a communicative approach, but I won't completely abandon the teaching that worked for me" (p. 252). The overall image to appear here, then, is that teachers' previous language learning experiences establish and develop cognitions about language learning and language in general, which shape the foundation of their initial conceptualizations of L2 teaching through teacher education, and which may continue to be significant during their professional lives.
Majority of the educational studies have revealed that, at the beginning of teacher education programs, students might have unrealistic, improper or immature understandings and images of teaching and learning (Brookhart & Freeman 1992).
Studies by Cumming (1989) and Brown and McGannon (1998) clarify this point regarding language teaching. Cumming studied student teachers' concepts of curriculum and came to this conclusion that these were insufficient and inadequate as the base of principled and effectual program design in ESL. Students were required to produce "a schematic chart outlining the curriculum decisions they would consider to be most important in teaching an ESL course" (p. 35). The author stated that the charts created by the student teachers were mostly inadequate and inappropriate regarding the relationships they postulated between theoretical and practical issues, the way various constituents of the curriculum were connected and sequenced, and the relative emphasis they put on specific constituents. Brown and McGannon (1998) distributed a questionnaire about L2 acquisition (taken from Lightbown & Spada 1993) among a total number of 35 TESL and LOTE method students in the early phases of their program. There were two beliefs upon which both groups agreed: languages were acquired chiefly by imitation and that mistakes were mostly because of L1 interference. These beliefs and attitudes were obviously insufficient as the foundation of efficient L2 pedagogy.
The impact of teacher education on trainees was indicated in Almarza (1996), who pursued the learning of four student teachers on a PGCE9 course. Especially, the results underlined the discrepancy between cognitive and behavioral changes which teacher education programs might induce. Behaviorally, all four students accepted the particular teaching system they were taught on their program, and applied this in their classrooms during practice teaching. This behavior, however, was no less than a result of the need felt by the student teachers to follow to specific standards (they were, in the end, being assessed). Cognitively, nevertheless, the student teachers differed in their agreement of the proposed method to teaching. These differences appeared when the students talked about their work, rather than through their practice, and were largely rooted in the different cognitions about language, learning, and teaching they held prior to their training. For instance, on accomplishment of her teaching exercise, one of the student teachers "saw herself free from the constraints imposed by the context of www.ccsenet.org/elt English Language Teaching Vol. 9, No. 2;2016 the classroom, she was back in a position in which she could continue to explore the ideas she had about language prior to the beginning of the course" (p. 69). This study concluded that although teacher education played an influential role in forming the student teachers' behavior during teaching practice, it did not modify meaningfully the cognitions the students brought to the course. Cabaroglu and Roberts (2000) utilized an order of three in-depth interviews to analyze the procedures, instead of the content, of belief improvement in 20 PGCE Modern Languages students. They observed that just one participant's beliefs remained unchanged during the program. In this study, evidence of change appeared from the analysis of interview data, from which the authors established categories of belief development processes (listed with definitions and examples on p. 393). One category, for instance, was called "re-ordering", defined as the "rearrangement of beliefs regarding their importance"; a second was "re-labeling", which involved the renaming of a construct; a third example was "reversal", the "adoption of opposite of previous belief".
The researchers concluded that in contrast to viewpoints about conceptual inflexibility in student teachers' professional growth, the procedures they depicted in their study were a more realistic image of the modifications that could happen during teacher education in student teachers' belief systems.
Peacock (2001) (1999), and were compared with the beliefs of experienced ESL teachers. Three major dissimilarities between trainees' beliefs and those of experienced teachers were recognized regarding the following statements:  Learning a foreign language is mostly a matter of learning a lot of new vocabulary words.  Learning a foreign language is mostly a matter of learning a lot of grammar rules.  People who speak more than one language well are very intelligent.
In each item, the percentage of the first year trainees agreeing to these statements was much higher than that for experienced teachers (e.g., only 7% of experienced teachers agreed to the second statement, compared to 52% of trainees). Such beliefs were realized by the author as "detrimental to their [trainees'] own language learning or to their future students' learning" (Horwitz, 1999, p. 183) and he hoped that they would be removed in the course of the teacher education program (where explicit attention was given to the nature of L2 learning). To observe this, he requested the trainees to finalize the BALLI at two further points in their course. Based on the findings, Peacock deduced that "there was surprisingly little change over the three years on Horwitz's two core beliefs about vocabulary and grammar, or ... about the role of intelligence in language learning" (p. 184) and that the "data did not support the belief that trainees' beliefs are formed by their pre-service methodology courses" (p. 187). These conclusions conflicted with those largely appearing from studies of teacher cognition in language teacher education.
Numerous studies in mainstream educational research have shown that teacher cognition and classroom practice exist in "symbiotic relationships" (Foss & Kleinsasser 1996, p. 441). Several studies have also studied these relationships in the field of language teaching (Bailey, 1996;Bartels, 1999;Breen, 1991;Breen, Hird, Milton, Oliver, & Thwaite, 2001;Burns, 1996;Gatbonton, 1999;Golombek, 1998;Johnson, 1992a;Lam, 2000;Nunan, 1992;Richards, 1996Richards, , 1998aRichards, , 1998bRichards, Li, & Tang, 1998;Smith, 1996;Ulichny, 1996;Woods, 1991Woods, , 1996. Reflecting findings from the mainstream literature, these studies collectively showed that language teachers' classroom practices were shaped by a wide range of interacting and often conflicting factors. Teachers' cognitions, though, emerge consistently as a powerful influence on their practices, though, and these do not eventually always echo teachers' uttered beliefs, personal theories, and pedagogical principles. The thoughts forming language teachers' classroom practices were depicted in several ways in the studies listed above. These attempts have been explained in terms of instructional affairs or considerations teachers had, principles or maxims they were trying to apply, their thinking about different levels of context, and the pedagogical knowledge they held. Furthermore, the bases of teachers' instructional practices have been clarified in terms of their personal practical knowledge (Golombek, 1998), beliefs (Smith, 1996;Woods 1991), and, as indicated by Lam's (2000) study of L2 teachers' use of technology, teachers' personal convictions.
To acknowledge the different research traditions on which these studies extracted is of importance. While a more meticulous analysis would detect a range of positions, two opposing viewpoints can be underlined here. One originates from the educational literature on decision-making (see, for instance, Shavelson, & Stern 1981), the second from that on teachers' personal practical knowledge (Elbaz 1981;Clandinin & Connelly 1987). Though both viewpoints identify the role of teachers' mental lives in forming classroom events, the work on decision-making adopts a somewhat technical view of teaching which concentrates on recognizing the antecedents for teachers' interactive decisions and describing effectual decision-making processes.
The focal point of most studies of SLTC and pronunciation pedagogy is chiefly on teachers' beliefs regarding which features of pronunciation to teach and which teaching techniques to use, and, with few exceptions, the data collection relies on surveys or questionnaires only. In EFL contexts, several survey studies have been carried out. Sifakis and Sougari (2005) studied teachers in Greece and found that many teachers deeply valued native-speaker (NS) norms and inclined to follow NS-oriented approaches (e.g., role-plays emphasizing NS roles). Hismanoglu and Hismanoglu's (2010) questionnaire study in North Cyprus verified that reading aloud, dictionaries, and dialogues constituted the top three favorite techniques of teachers. Saito's (2011) survey found that Japanese English teachers detected eight segments believed to have a negative influence on learner comprehensibility, and Saito and van Poeteren's (2012) questionnaire study showed that the teachers reported using pronunciation-related adjustment strategies (e.g., speech rate modification, word-level enunciation, segmental-level enunciation) to modify their pronunciation in the classroom.
Finally, two studies by Jenkins (2005Jenkins ( , 2007 and one by Timmis (2002) concentrated on the subject of teachers' perceptions regarding accents and L2 instruction. In Jenkins's (2007) study, for example, interviews indicated that teachers seemed to favor teaching English as a lingua franca accents in theory, but in practice considered such teaching as impractical in the classroom.
In ESL contexts, research into teachers' beliefs about pronunciation instruction has received less attention to date and has relied almost solely on interview data. Macdonald (2002) interviewed teachers who had reported an unwillingness to teach pronunciation and realized that lack of institutional resources as well as inadequate knowledge of how to assess student pronunciation contributed to their avoidance to teaching pronunciation.
Baker's (2011) interview study indicated that teachers whose TESOL training comprised a course in pronunciation pedagogy reported prioritizing the teaching of supra-segmental features of pronunciation in their classes; at the same time, however, many of these teachers still seemed to lack confidence in teaching some components of English pronunciation.
Finally, Cathcart and Olsen (1976) investigated teachers' and students' beliefs about grammar and pronunciation correction and found that students asked teachers to correct them more frequently than the teachers actually did and both teachers and students favored the "correct" (e.g., NS) model approach. This final study alone included an examination of actual classroom practices; however, the article did not clearly connect the observed lessons with the teachers who taught those lessons.
What is missing from all these studies, hence, is an in-depth exploration of how teachers apply their beliefs effectively and fruitfully in the classroom to help students attain comprehensible pronunciation.
Aside from these studies, the teaching and learning of pronunciation in typical ESL or EFL classrooms have remained mostly unexplored, signifying that research into current pronunciation-oriented teaching practices of L2 instructors is long overdue. In a similar vein, relatively few teacher education programs provide courses on how to teach L2 pronunciation.
The studies conducted on the issue of cognition among the Iranian teachers and language learners of English are included, but not limited to, the following issues. Rokhsari (2012)  comprehension and reading cognitive, metacognitive, and test-taking strategies. The results showed that the successful readers used reading strategies more frequent than their unsuccessful peers.
In line with others, Sadeghi and Bidel Nikou (2012) investigated the perceptions of Iranian EFL high school teachers and students on teaching and learning reading skill by using questionnaires. The results revealed that both groups reached to a consensus that there should be more attention to reading cognition.
Alijanian (2012) investigated whether variables such as work environment, teachers' experience, and their genders could affect beliefs about grammar and its teaching. The researcher used quantitative techniques of data collection (questionnaire) to answer the research question. As the results showed, the aforementioned variables could have a significant effect on the teachers' beliefs and cognitions about teaching grammar.
Investigating the influence of EFL teachers cognition on teaching English language grammar, Moini (2009), addressed the differences in Iranian EFL teachers' beliefs about grammar and teaching grammar. The intended differences lied on variables like work environment (private school vs. state schools), degree, gender, and teaching experience. The results of a grammar belief questionnaire indicated a significant difference between school teachers and private language institute teachers in their cognition and practice. The observed differences were also significant for teachers' degree and experience. However, the results indicated no significant difference regarding teachers' gender. In general, the results revealed that teachers' characteristics affected some aspects of their cognition and teaching practice in terms of teaching grammar.
Despite its significance as a basic constituent of oral communication, pronunciation has been under-researched and neglected among language researchers in Iran (Behzadi & Fahimniya, 2014). They stated that Iranian researchers have mostly concentrated on skills and sub-skills much more than pronunciation in language teaching. That is why they investigated the influence of two different approaches of teaching pronunciation: intuitive-imitative and analytic-linguistic on students' speaking fluency among a group of Iranian non-native EFL learners. Age was considered as a moderating variable in applying the two approaches. The findings indicated that the intuitive-imitation approach was more fruitful and efficient for the younger participants while the analytic-linguistic approach was more effectual for the older ones.

Method
In this section, the methodology used to elicit data was explained.

Participants
Five experienced EFL teachers were invited to take part in this study. The teachers were randomly selected based on their current placement as an oral communication instructor, their teaching experience, and their willingness to participate in a research study. All of the teachers had taught their oral communication courses at least once in a previous semester and each of them had 5-10 years of teaching experience. All of these teachers were working in private English language institutes in Najafabad, Isfahan, Iran. They had graduated from the field of TEFL (Teaching English as a Foreign Language), or held a master's degree in a TEFL-related fields. They voluntarily agreed to participate in the present study and gave the researcher oral consent. Regarding their language proficiency, the teachers varied: two of them had passed their TOFEL or IELTS in recent years, two of them were enrolling in pre-IELTS and pre-TOFEL courses in other language institutes, and one of them was recently graduated from his/her university; therefore, his/her proficiency in English was not as high as others. In addition to the English teachers, 54 students with the range of 14-17 years who were enrolling in these classes participated in the study. The students whose parents permitted them to take part in the study were asked to fill up an opinion questionnaire which helped the researcher gather more data on the teachers' cognitions about pronunciation techniques and language learners' characteristics.

Semi-Structured Interviews (SSIs)
Contrary to previous studies conducted on SLTC which collected a considerable amount of data from a small number of participants (Farrell & Lim, 2005), the present study conducted three SSIs with each of the teachers.
The interviews were carried out in three phases: one at the beginning of the semester, one in the middle of the semester, and one at the end to triangulate the obtained data.

Questionnaires
As was mentioned, in addition to the teachers' participants, students were also required to take part in this study.
A brief questionnaire about students' opinions and beliefs about the pronunciation learning and teaching was administered among the students. As stated by Baker (2014), only a small number of studies benefited from the

Procedure
In order to investigate the current issue from different angels, the current study applied a sequential mixed-method approach of data collection. In the qualitative phase of the study, the researcher applied interviews to collect in-depth data about the teachers' cognitions (knowledge, beliefs, perception, and attitude) towards the pronunciation techniques they used in their classrooms. In doing so, five EFL teachers were invited to participate in this phase of the study. The teachers were chosen based on two criteria: their current placement as an oral communication instructor and their teaching experience being more than five years. As was already mentioned, the interviews were conducted in three phases: beginning, middle, and end of the course. In addition, a quantitative research approach, i.e. questionnaire, was also used in order to gather students' perceptions and opinions about the techniques their teachers apply in the oral communication classrooms. To this end, a number of 54 students from the above teachers' classes were asked to fill out an opinion questionnaire in order to gather more data about the teachers' pronunciation techniques. This questionnaire was administered among the students at the last week of the semester.

Data Analysis
Interview data was analyzed based on the following stages (Baker, 2014;Woods & Fassnacht, 2009): (1) transcription of interview data, (2) data segmentation and coding of the interview data. Analysis of quantitative data, i.e. data obtained from questionnaire, was done by using Statistical Package for the Social Sciences (SPSS).

Results and Findings
This part is divided into two sections; in the first section, the breadth of teachers' knowledge about the types of techniques they use, their beliefs about these techniques, and finally, the degree to which each of these techniques were used in the classrooms were dealt with. The second section was devoted to the findings of teachers' cognitions about their learners, specifically their cognitions about learner confidence, desire for instruction, and desire for feedback. The results of the study were obtained from the triangulation involved the interviews with the teachers and students questionnaires. Both quantitative and qualitative approaches were applied to evaluate research questions.
In order to answer the first research question, the teachers were invited to an interview session in which the interviewer asked them to express some of techniques they frequently use in their teaching pronunciation, particularly those which enhance language learners' pronunciation skill. A summary of these techniques are presented in Table 1. As shown in the table, teachers almost utilized all kinds of techniques in their classrooms. Generally, teachers' interview proposed that controlled activities were dominant in OC classrooms comparing to other kinds of activities. All teachers began their teaching by supplying goals for their lessons in which pronunciation instruction was included in this preparation section (plan and purpose). As expected, explanations and examples, activity set-up, and checking activities were the central part of teachers' instruction. At some points of the instruction, teachers spent some moments for explaining English pronunciation features, followed by giving instruction of an activity which involved listening to or producing that feature. Students were also faced with feedbacks regarding their pronunciation by their teachers (checking activities). Other controlled activities applied by the teachers were "production and practice", where students were requested to practice the new feature they have just acquired, "repetition drills", which were mostly preferred by the institutes where teachers were working, and "reviewing" and "testing activities".
Other controlled activities were used in various degrees by teachers. T1 and T4 showed approval of using "kinesthetic/tactile practice", i.e. using whole body movement when teaching different aspects of English pronunciation. "Display questions", which were aimed either to assess students' prior knowledge of a pronunciation feature (knowledge exploration activities) or to decide if students had acquired previously-taught information (knowledge verification activities), served a role in T2, T3 and T5 classrooms. The two sets of "identification" (visual and audio identification) and "recognition" (visual and audio) activities were only implemented in T2 and T5 classrooms. Based on the bio-data questionnaire, these two teachers were taking upper graduate courses in English and that is why they employed a wider repertoire of teaching pronunciation techniques than three other teachers.
The variety of guided techniques, notwithstanding, was considerably more limited comparing to controlled activities. Except for "question and answer referential" and "preparation" techniques, which were used by all five teachers, "production" techniques (both audio identification and audio recognition) seemed to be the central part of T2 and T5 classroom instruction. "Mutual exchange", which was the type of information gap technique, was only applied in T2 teaching pronunciation.
As far as free techniques are concerned, the results of teachers' interview showed that, in general, those teachers who were in lower level classrooms used techniques such as "game" (T3 and T4) and "drama" (T1 and T4), while T2 and T5, who were both teaching in higher level classrooms and were majoring in upper-graduate courses in TEFL, adopted techniques such as "presentation" (T2 and T5) and "discussion" (T2 and T5). Figure  4.1 represents the distribution of techniques applied by five teachers in their OC classrooms.

Figure 1. Distribution of techniques used by the teachers
On the whole, the teachers participated in this study seemed to apply a variety of controlled, guided, and free pronunciation-oriented techniques in their classroom instruction, while the controlled techniques played a leading role in their teaching in all the classes. The next section summarized the results of students' questionnaire about the type of techniques their teachers applied in their classroom.
In order to answer the second research question, the cognitions of teachers were compared with the students' beliefs as reported from the questionnaire completed by students following the second round of teachers' interview. The questionnaire dealt with subjects like learners' confidence, desire for feedback, and desire for instruction, the results of which are exhibited in the following parts.
The answer the teachers provided for the question "how confident do you feel your students are in their English pronunciation techniques?" showed that nearly all teachers believed their students' to be confident in this regard. The results correctly agree to the findings gained from students' questionnaire. As shown in Table 2, students marked "strongly agree" or "agree" in their response to the statement "I am confident about my English pronunciation skills".  As shown in the following figure, students who were learning English in the T2 and T5 classes appeared to be more confident in their pronunciation skill. In the interview, T2 stated that some of his students aimed to pursue their education overseas; consequently, they were working on their pronunciation skills long before enrolling in this class. A number of students in T2 and T5 classes, however, were uncertain about their pronunciation skills. Only one student in T4 class appeared to strongly disagree with his/her pronunciation skills. When inviting T4 to have comments on students' response, she seemed to agree with student comments about his pronunciation capabilities.

Figure 2. Students' response to question 1
While the results gained from students' questionnaires indicated that students' response to the first statement were different, all of the students answered "strongly agree" to the statement "I want to improve my pronunciation skills in English". Along with the students' response, teachers believed that the majority of the students were enthusiastic to enhance their pronunciation abilities. The exception to this was T4 who believed that the students in her class were reluctant to practice their pronunciation skills. She expressed the attitude that "students in my class believe that they know everything about English and they do not care about my recommendations of practicing English more". The distribution of student's answers to this statement is demonstrated in Figure 3. Although a number of students fairly agree about their desire to improve their pronunciation abilities, generally, students intended to work on their skills. It was proven that all teachers have cognitions about students' desire of their teachers to teach pronunciation in the classrooms. In answering the research questions about the reason why they thought so, T5 mentioned: "I think students like to improve their pronunciation abilities. I think what they need is an extra class for teaching pronunciation features widely and effectively. In that course, students would be familiar with even tiny differences between different English accents, such as British and American accents. They should be provided with some pronunciation-oriented courses which, subsequently, lead to improvement of their listening and speaking skills".
As clarified in the following table, students agreed with their teachers and believed there should be instruction on English pronunciation techniques. Table 4 and Figure 4 demonstrate the results of students' answers to the statement "I want my teacher to teach pronunciation".  These results clearly showed the agreement between teachers' cognitions and students' beliefs about pronunciation instruction.
When the teachers were demanded to give comments on the time they usually spent on teaching pronunciation in their classrooms, they complained about the limitation of the time. As stated by the T2 and T5, pronunciation was not properly addressed in these courses. The reason was that students were presented with a variety of communicative activities and tasks; hence, no time would be left for teachers to concentrate on teaching pronunciation. T1 and T3 also believed that they did not spend sufficient time on teaching pronunciation in their classes, whereas their views opposed their students' views. On the word of the T3, "Quite surprisingly, my students imagine I spend more time than needed on teaching, practicing and checking their pronunciation abilities".
With the next three questions which were in the form of questionnaire and interview, students and teachers were asked to explore teachers' cognition about students' desire for feedback. As the results showed, almost all students in these five classes expressed desire for receiving feedback from their teachers. To determine the ways of giving feedbacks to students, the researcher requested them to select whether they wanted to give feedback 1) in front of the class, 2) privately, or as a group in the class. T2 and T3 stated that their students would be more enthusiastic to get the feedback in the group. To the researchers' surprise, students in these classes preferred to be presented with feedback privately. Students in these classes were proficient learners; therefor, they would not like to be corrected in front of other students; this can be a justification for this contradiction.
In contrast to these two teachers, T5 supposed that it would be more embarrassing to correct students in front of the class; thus, private or in-group correction would be more beneficial. As commented by T5,

"Correcting the students on the spot and in front of the class give them bad feeling, I think. When I was student, I didn't really like to be corrected by my teachers. Sometimes if he/she remind me of my pronunciation mistake after the class would produce more fruitful results!"
The results of students' questionnaire also approved teacher knowledge of students' inclinations.
T1 and T4 were two teachers who believed that students would not feel embarrassed about correcting their mistakes in front of the class. For example T1 stated: "If they (the students) know me well, they will understand that I correct students' pronunciation mistakes without feeling as if they are disappointed or shy! This feedback would give other students benefit as well! So why not correcting them in front of others? This approach is more beneficial than the other two!" The students' answers to this statement in the questionnaire differed, though.
Students in T1 class liked to be provided with feedback mostly in the form of private or in-group correction, while students in T4 agreed with their teacher's comments. T4 commented that, "students are more comfortable in the classroom environment and therefore they are not afraid of error correction". The general outcomes showed that students did not like to receive feedback in front of the class. The following tables clearly demonstrate the results obtained from question 4 to 6 of students' questionnaire.

Discussion and Conclusion
In spite of its influential and significant role in communication, pronunciation has been regarded as one of the isolated areas of language teaching. According to Harmer (2001), English teachers have more focus on areas such as grammar, vocabulary, and oral skills, without being much concerned about teaching pronunciation to students. Similarly, Fraser (2002) believed that most teachers do not provide students with proper instruction on pronunciation and if they teach students, they mostly adopt a hit-or-miss method which would not lead to desired results. Iranian EFL teachers are not exception to this situation (Gooniband, Mehrabi, & Mousavinia, 2013). As stressed by Gordani and Khajavi (2012), it is necessary for Iranian EFL teachers to have a comprehensive awareness of learners' needs and desires in order to deal with learners' problems in the process of language learning. SLTC, as an established tradition of research in language teaching, lies within these issues. As defined by Borg (2006), SLTC is the relationship between the beliefs, perceptions, and knowledge that the teacher brings to the classroom which can provide insight into how teachers' knowledge and opinions interact with his/her teaching practice and pedagogical behaviors and activities. SLTC stressed on the fact that teachers' cognitions should be investigated since it will help them improve their theoretical knowledge of teaching practice about English skills and sub-skills (Bartels, 2005).
With respect to this line of inquiry, the current study tried to provide a description of Iranian EFL teachers' cognition about their actual teaching of pronunciation to a group of language learners. The study examined the teachers' knowledge about the techniques they use for teaching English pronunciation in their OC classrooms and teachers' cognitions about Iranian EFL learners, more specifically those who were their students in these classes. For these purposes, a triangulation of data collection method (questionnaire and two rounds of interviews) was used to collect appropriate data. A summary of main findings was dealt with in the previous section and the discussion which has arisen from the findings will be provided in the following sections.
With the analysis of teachers' interview comments it was realized that while teachers had cognitions about all three types of techniques, "controlled techniques" were their knowledge base in teaching pronunciation.
Remarkably, teachers desired to use some limited number of "guided techniques", mainly those which can be called "semi-controlled techniques". The results, moreover, indicated that the teachers in lower level classrooms utilized some "free techniques" (such as drama and game), while other types of these techniques (such as discussion and presentation) were only used for upper level students. These findings directly mirrored the status of pronunciation pedagogy in language teaching. As declared by Baker (2014), for decades, pronunciation instruction "was considered synonymous with imitative-intuitive and analytic-linguistic approaches in which controlled techniques formed the foundational core of teaching" (p. 153). While traditional approaches were criticized for their "learning-that" attitude they had towards language teaching, and more recent communicative approaches adopt "know-how" attitude instead, still, controlled techniques and activities are evaluated (Kumaravadivelu, 2012), as expressed by teachers in this study. Conspicuously, "controlled techniques" are more preferred by students since, as some studies show, these techniques have marked influence on the development of students' intelligibility (Reed & Michaud, 2011) and their phonological enhancement (Saito, 2007). Yet, the controlled activities have meaningful inadequacies as well. Studies have shown that the use of semi-controlled and free techniques, comparing to focus-on-form and controlled techniques, have substantial impacts on learners' uptake and consequently, automatic application of targeted features of pronunciation (Saito & Lyster, 2012). In the same way, Khatib and Nikouee (2012) proposed that the combination of controlled and communicative techniques could have obvious influence on Iranian EFL learners' retention and automatization of some grammatical features than the application of controlled and mechanical drills alone.
The second phase of the present study investigated the general cognitions of teachers about their language learners, in terms of three factors of learners' confidence, learners' desire for feedback, and learners' desire for instruction. The main finding of this part of the study was that teachers' education played a fundamental role in forming their cognitions about usual practices of pronunciation pedagogy. Without such education, neither their knowledge about pronunciation instruction was in-depth and up-to-date nor was their confidence enormous sufficient to direct classroom practices. In this study, for instance, those teachers who were taking upper graduate courses in English teaching seemed to be more aware of their students' different characteristics which, accordingly, resulted in their showing confidence in teaching these skills. This finding is in line with Fraser (2000) and Macdonald (2002) who had found that teachers without enough guidance and training in instructing pronunciation appeared to be less confident in teaching this sub-skill.
The results, moreover, showed that teachers had different cognitions about evaluating their students' pronunciation either in front of the class, in-groups correction or privately. A group of teachers believed that it would be embarrassing for language learners to be assessed in front of their peers. In fact, they were uncertain www.ccsenet.org/elt Vol. 9, No. 2;2016 about this fact that whether they correct their students in front of class might be damaging to the students' emerging identities. This finding is in accordance to the ongoing discussion about pronunciation instruction and feedback. As Golombek and Jordan (2005) stated, pronunciation instruction can even be considered a threat to language learners' identity, especially when non-native English teachers instruct it. As pronunciation instruction and feedback are concerned, therefore, language teachers are left with their own capabilities in order to choose if they should ever correct their students' errors or give them instructions. The imperative concern here is that Iranian EFL teachers need to recognize the critical nature of pronunciation instruction, on one hand, and the cognitions about their learners' particular characteristics, on the other hand, in order to help them better improve their communication skills.
Due to some methodological limitations, which were imposed on this study, generalization cannot be made. However, the present study revealed that there was a close relationship between language teachers' experience and education with their cognitions about their language learners. In the present study, for example, those teachers who were taking upper graduate courses in English teaching seemed to be more aware of their students' different characteristics which, accordingly, resulted in their showing confidence in teaching these skills. The more experienced the teachers, the broader knowledge they have about their students.
The findings of this study can affect the way language teachers are prepared in teacher training courses (TTC). Unlike grammar, vocabulary, and communication skills, which are in focal point of TTCs, pronunciation seems to be rarely taught. Thus, without any preparation, teachers are left on their own to decide how to teach pronunciation and how to deal with their language learners' characteristics. Indeed, language teachers might be unaware of the complication of English pronunciation and different techniques and approaches for teaching this sub-skill. In this regard, one of the important implications of this study was that TTCs in Iran must be equipped with better pronunciation pedagogy.
The results of the present study can also enrich the current literature of SLTC. While the study suffers from some data collection methodologies, including data collection methods such as observation of teachers' actual practices, it supplied effective results about the representation of a group of teachers' cognition.
If possible, further research in the area of language teachers' cognitions should be conducted, viewing the current study's findings, in order to further improve the teachers, practitioners, and specialists' knowledge about teachers' beliefs and perceptions in teaching pronunciation.

Limitations and Suggestions
Although the researcher made his best to avoid research design imperfections of prior research studies, this study cannot claim to be totally free from limitations. This section foregrounds a few of these limitations.
Delving into the cognitions of teachers needs more than one technique and research methodologies, and if appropriate methodologies were used to study this issue, one cannot claim that the insight gained is complete for sure. The present study mixed data collection methods like interview from teachers and questionnaire from both teachers and students to create a widespread picture of teachers' cognitions in pronunciation pedagogy. However, we were only able to explore those procedures and areas which teachers articulated into words. Teachers could not always outline their reasons for what they do in the classrooms even when students claim their teachers behave in a particular way.
Another limitation is placed on the cognition components this study tried to investigate. Shulman's (1987) category of knowledge base, which were applied in this study, contains seven elements, including content, general pedagogical, curriculum, pedagogical content, language learners', educational materials and techniques, and finally, educational purposes and ends, where only knowledge base related to language learners and educational materials were the focus in this study. This limitation challenges the generalizability of the results since teachers' knowledge only was explored regarding their cognitions about materials/techniques and language learners' characteristics and preferences.
This study also suffers from limited number of teacher participants. While the study presented a clear picture of Iranian EFL teachers' cognitions in their pronunciation instruction, the limited number of teachers who agreed to participate in the data collection process of this study imposed further limitation to this research project.
Further research in teachings' cognitions in pronunciation pedagogy will ideally combine a triangulation of different data collection methods to provide a more clear portrait of teachers' knowledge than any conclusions obtained from single method alone. Classroom observation, for instance, can best be performed to examine what www.ccsenet.org/elt Vol. 9, No. 2;2016 teachers believe to be the most efficient techniques they can apply in pronunciation instruction and the situations under which language learners can best learn the pronunciation characteristics.
An investigation of other components of Shulman's (1987) model can be worthwhile, too. As mentioned before, this model contains seven different categories of knowledge, which only the two of them had been worked on in this study.
The present study investigated teachers' cognition regarding techniques they apply in language classrooms learners' characteristics and preferences of these techniques. Other studies also presented more data about teachers' attitude, beliefs, and perceptions, as other constituents of teachers' cognitions (Baker, 2014), about pronunciation pedagogy.
Data collection can be performed with more number of teachers. Greater and more active teachers more likely result in better understanding of their attitude, knowledge, perceptions, and beliefs about their pedagogical practices.