A case study of EFL teachers’ practice of teaching speaking skills vis-à-vis the principles of Communicative Language Teaching (CLT)

Abstract Though revising the existing syllabus is part of the Ethiopian education system, the study of the implementations of the notions of the syllabus attracts a few researchers’ interests. Therefore, this study investigates teachers’ practices of teaching speaking and compares them with the principles of Communicative Language Teaching (CLT), the underlying approach of the new syllabus. It also explores teachers’ rationalizations of their practices and the factors that influence the teaching of speaking. To achieve the objectives, the researchers used a qualitative case study research design. The data from the four purposively selected English teachers was gathered through classroom observations and stimulated recall interviews. In analyzing the data, a qualitative content analysis method was used. The results of the study showed that 1) the major activities and teaching cycle the teachers employed were similar, 2) most of the actual classroom practices did not match with the principles of Communicative Language Teaching, and 3) there were learner-related, teaching-related, curricular, and environmental factors that influenced the teachers’ enactment of the principles of CLT and the teaching of speaking in general. The predominant influencing factors were the learner-related: Lack of basic background skills, lack of experience in practicing speaking lessons, the use of mother tongue in pair and group work activities, fear of making mistakes, lack of interest in speaking lessons, and excessive focus on standardized tests.


PUBLIC INTEREST STATEMENT
The trainers in teachers' education colleges and universities introduce the trainee teachers to diverse teaching methods and approaches and encourage them to employ the contemporary ones. In the Ethiopian educational context, the English language syllabuses are designed based on the principles of Communicative Language Teaching (CLT). The preparation of the syllabuses based on CLT's principles and instructing teachers to employ them in the classroom cannot be an assurance for its implementation. Thus, this study explored EFL teachers' practices of teaching speaking and matched the practices with the principles of CLT. It was a qualitative case study that was conducted on four EFL teachers. The findings revealed that most of the teachers' practices did not match the principles of CLT. The teachers' enactment of the principles of CLT was influenced by learner-related, teaching-related, curricular, and environmental factors. The findings of this study, therefore, show that it is not always enough to introduce teachers to a certain approach and instruct them to implement it. There should be mechanisms to check if the practices are in congruence with the theories the teachers study.

Background
In the study of teacher cognition, investigating the relationships between teachers' beliefs and their actual classroom practices is one of the key areas. Although the comparison of beliefs and practices seems to be a well-established area of study, the relationships are not as simple as the associations between cause and effect. Hence, the debate is yet unresolved as different teacher cognition researchers report three forms of relationships: congruent (Mansour, 2009;Pajares, 1992), incongruent (Hedrick et al., 2004;Khader, 2012;Waluyo & Apridayani, 2021), and mixed (Borg, 2003;Farrell & Lim, 2005). Like the comparisons between teachers' beliefs and classroom practices, it is also significant to examine the associations between the classroom practices and the notions behind the syllabus.
In Ethiopia, the syllabus is considered to be an important aspect of the education system as it is believed to guide the instructional process. Thus, revising the existing English syllabus has been part of educational reforms whenever there have been shifts in the theories and principles of second/foreign language teaching. For example, the English for New Ethiopia syllabus had functioned for two decades, from the mid-1970s to the mid-1990s. It consisted of ten books from grades 3 to 12. This syllabus was criticized more for its focus on the usage of the language than the use of language (Leta, 1990). Hence, it was replaced with English for Ethiopia in 1996, a syllabus that focused on developing learners' ability to communicate in the language. As indicated by the authors, the features of the 1996 syllabus were: student-oriented, task-based and meaningful, and realistic communicative activities.
The English for Ethiopia syllabus was revised in 2008/09. In the syllabus, it is stated that the revision was done "based on the new curriculum framework for Ethiopian schools and needs assessment conducted before the revision work, and consideration of international content standards for a similar age and grade level of learners" (Ministry of Education (MoE), 2009b p. 5). The new Curriculum Framework, which played a significant role in the revision, was developed based on the analysis conducted by the General Education Curriculum Framework Development Department (GECFDD) between 2003 and 2005. According to the documents of the syllabus, the current "English Syllabus, Grades 9 and 10, and Grades 11 and 12" present six main changes from its predecessor (Ministry of Education (MoE), 2009b). One of these changes is the approach to how the teachers should teach the language. It is indicated that the approach that underlies the new syllabus is communicative and skills-based. According to the syllabus, the language items the learners learn and practice are meaningful and purposeful. The classroom activities are also designed to enable learners to practice the language in pairs, groups, and plenary as language is about communicating with others. The four language skills (listening, speaking, reading, and writing), therefore, are the primary focuses of the syllabus while other language items or components (grammar and vocabulary) are integrated into the practices of these skills. In a nutshell, the syllabuses require the English language teachers to employ the Communicative Language Teaching (CLT) approach.
The shift of focus to CLT is not peculiar to the Ethiopian education system alone. Littlewood (2007) reported that ESL/EFL countries have shifted their language teaching policies to CLT since the 1990s aiming to help their learners to be able to communicate in English. The shift to CLT was initiated by the scholars' evaluation that the linguistic structure-centered methods were not enabling the learners to use the target language for communication (Desai, 2015). Therefore, the Ethiopia education and training policy of 1994 ratified the use of CLT in the teaching of English hoping that the instructional process would shift from the structure-oriented to communicationoriented approach (ICDR, 1994).
Generally, all efforts to improve the syllabus emerge from the assumption that the syllabus guides the instructional process. In other words, it is assumed that an improvement in the syllabus is an improvement in classroom practices. However, teachers as humans do not unthinkingly implement all plans they are required to do (Borg, 2009). Hence, it is not enough to design a syllabus and instruct teachers to use a certain approach, but it is also mandatory to assess the implementation of the chosen approach. In addition to what is stated in the syllabus, there are contextual factors that play intermediate roles between what is expected to be implemented and what is implemented. This is likely why Mansour (2009) reported that a full understanding of the contexts enables us to understand the teaching and learning process. Accordingly, this study is a comparison of the classroom practices and the principles of Communicative Language Teaching (CLT), an approach that underlies the syllabus, within the framework of the contextual factors.

Research questions
This study investigates teachers' practices when teaching speaking and compares them with the principles of the Communicative Language Teaching (CLT) approach. This study also investigates the contextual factors that influence the teaching of speaking and/or teachers' enactment of the principles of CLT. Hence, the following were the research questions this study tried to answer: (a) What are the teachers' observed classroom practices in teaching speaking lessons?
(b) What are the teachers' rationalizations of their classroom practices?
(c) What is the relationship between the teachers' practices of teaching speaking and the principles of CLT?
(d) What are the contextual factors that influence the teaching of speaking and/or their enactment of the principles of CLT?

Communicative Language Teaching (CLT)
Communicative Language Teaching (CLT), as an approach to language teaching, was introduced to the field of second/foreign language teaching in the 1970s due to the insufficiency of the existing methods to enable learners to communicate in the target language (Desai, 2015). The learners in linguistic structure-centered methods were found to be ineffective in communicating in the target language though they were able to read and write sentences (Larsen-Freeman & Anderson, 2011). In the Communicative Approach, language is viewed as a means of communication (Richards & Rodgers, 2014). Hence, the goal of language teaching is to achieve communicative competence. Communicative competence refers to the knowledge and skills of individuals to use the language for diverse purposes and functions (Richards & Rodgers, 2014) which requires know-how in four areas: "grammatical competence", "sociolinguistic competence", "discourse competence", and "strategic competence" (Canale & Swain, 1980, p. 4). In CLT, as the view of language and the goal of teaching reveal, communication is the basic feature of the instructional process. Hence, the effectiveness of the classroom activities, the teacher's and learners' roles, and student-teacher interaction are assessed according to how much they promote communication. Larsen-Freeman and Anderson (2011), for example, described the roles of the teacher in the CLT classroom as facilitating communication while the learners are communicators.
There are underlying principles of Communicative Language Teaching that the teachers should be aware of to support their students' language development. Richards and Rodgers (2014, p. 105), for example, listed five principles of CLT: Learners learn a language through using it to communicate.
"Ethiopian students need to attain a reasonable degree of mastery of the English language to successfully respond to their academic demands." Although English plays a crucial role in the Ethiopian educational context, it is found to be the weakest part of students and even teachers at all levels of schooling. Alamiraw (2005) for example, reported that learners' knowledge of English was poor, and teachers could not help them since they themselves were not good at it. This situation was not changed even after ten years as Mebratu (2015) rated the status of English in the Ethiopian schools as very poor.
Cognizant of this fact, attempts have been made to improve the status of English in Ethiopia in the last few years. The Ministry of Education, for example, has been organizing the English Language Improvement Programme (ELIP; Girma & Sarangi, 2019). This program mainly focuses on improving the English language skills and language teaching methods of the teachers by providing in-service training and preparing workshops and seminars. Besides the ELIP, revising the English syllabuses has been part of the improvement process.

The English syllabuses for grades 9 to 12
In 2008/09, the English Syllabus replaced the 1996 English for Ethiopia syllabus. The evaluation by the General Education Curriculum Framework Development Department (GECFDD) between 2003 and 2005 was the reason behind the changes. The analysis identified five drawbacks of the then curriculum: a lack of relevance of some of the contents, problems in the assumed methodology of teaching, difficulties in the implementation of continuous assessment, overloaded and often conceptually advanced contents of the textbook, and the teaching-learning materials that do not promote a student-centered approach (Ministry of Education (MoE), 2009c). Thus, the Ministry of Education prepares the English syllabus for grades 9 and 10, and 11 and 12 based on the new curriculum framework for Ethiopian schools, needs assessment, international content standards for similar ages, and learners' grade levels (Ministry of Education (MoE), 2009a). Accordingly, the current English syllabus for grades nine to ten presents six main changes from its predecessor (Ministry of Education (MoE), 2009c). First, the contents of the new syllabuses have been reduced to allow teachers to cover the books in the given time. Second, the new syllabus is designed based on the "Minimum Learning Competencies" (MLCs) which are technically defined as "the skill levels we expect all the students to have reached" (Ministry of Education (MoE), 2009a, p. 7). Thus, learning goals and expected outcomes of each topic, unit or grade are stated, and contents and activities needed to develop them are provided. The students are assessed in accordance with the stated competencies. Third, the complexities of the contents and tasks increase gradually. Hence, the language lessons move from the cognitively simple to the more demanding ones. Fourth, it is noted that the contents in the new English syllabus for grades 9 and 10 are relevant for the three groups of learners after grade 10: those who leave school, continue to grades 11 and 12 or move into the vocational stream. Hence, it is believed the contents at secondary schools are pertinent to learners and enable them to use English in their preparatory classes, at colleges, or in their workplaces. Fifth, the practicality of active learning and learnercenteredness is emphasized in the new syllabus. It is stated that the implementation is determined by three inextricably related language teaching components: having quality textbooks, developing the confidence and competence of teachers, and employing different methodologies. Sixth, the language teaching approach that underlies the new syllabus is communicative and skills-based. According to the syllabus, the language items the learners learn and practice are meaningful and have real purpose and context. The classroom activities are also designed to enable learners to practice the language in pairs, groups, and plenary as language is about communicating with others. The four language skills (listening, speaking, reading, and writing), therefore, are the primary focuses of the syllabus while other language items or components (grammar and vocabulary) are integrated into the practices of these skills. Of all the changes in the new syllabus from its predecessor, the change in the approach that underlies the syllabus is central to this study as it sets what the classroom practices should look like.

The challenges of teaching speaking in the EFL context
The teaching and learning of speaking is an essential part of English language instruction as the mastery of speaking skills is a priority for many language learners through which they evaluate their success in language learning and the effectiveness of English courses (Richards, 2006). One of the reasons can be the distinctive advantages that the ability to speak English provides, for example, speaking ability is one of the criteria that many companies and organizations adopt when looking for employees. Though the ability to speak skillfully offers numerous advantages, speaking is the most demanding of all the four language skills in a foreign language context (Jung, 1995;Rahimi & Quraishi, 2019). It is a difficult skill; even those who have been studying English for many years find it difficult to speak appropriately and understandably (Bueno et al., 2006). Speaking is challenging because it involves a set of physical and mental activities more than just knowing the grammar and semantics of the language (Jung, 1995;Shumin, 1997). According to Levelt (1995), for example, the cognitive process of speaking involves three interacting and overlapping stages (conceptualization, formulation, and articulation) and one ever-present process (self-monitoring) (as cited in Goh & Burns, 2012).
The challenges EFL learners encounter in their mastery of speaking are twofold: the difficult nature of the skill itself and contextual factors. Obviously, the lack of speaking community exerts a major problem as it makes the classroom the only place where the learners learn and practice the language. Taddese (2019) and Asfaw et al. (2021) reported that the following were some of the factors that influence the teaching of speaking in Ethiopia: giving less attention to speaking lessons, excessive use of mother tongue, learners' poor background, shortage of time, frequent use of few classroom activities, poor classroom management, fear of making mistakes and shyness, and uneven participation.

Approaches to teaching speaking
In an EFL context, mastery of speaking is greatly dependent on the approaches to teaching speaking as the classroom is the decisive place where students learn and practice the language. Thus, different approaches to teaching speaking have been adopted and reported. Broadly, Talley (2014) classified the approaches to teaching speaking into two: explicit and implicit. In the explicit approach, the teachers mainly focus on teaching the common expressions and require their students to study and memorize them. The teachers also teach the learners what should and should not be said. The implicit approach, on the other hand, focuses on facilitating situations for the learners to practice the language. Predominantly, the approaches to teaching speaking are categorized into two main types: direct approach and indirect approach.
The direct approach, also known as a controlled approach, to teaching speaking, focuses on the practice and development of isolated language forms (Goh & Burns, 2012). The practice involves a reproduction of predetermined language through drills, pattern practice, and structure manipulation. Novakova (2016) claimed that structural accuracy and the use of language forms are the concern of the direct approach. An indirect approach to teaching speaking, on the other hand, focuses on the creation of conditions for oral interaction to enable learners to engage in communicative activities (Goh & Burns, 2012;Richards, 2008). In the indirect approach, fluency is the primary concern, and learners are exposed to authentic and functional language use. Thus, the practice involves real-life communicative activities such as discussion, information gaps, role-plays, simulation, and so on (Burns, 1998). This approach is also known as the transfer approach to teaching speaking due to the assumption that learners develop speaking skills by communicating with each other which can be later transferred to a real-life situation. As can be understood from the above discussion, the indirect approach is the representation of CLT in teaching speaking skills (Celce-murcia et al., 1997).
Exclusive reliance on one of the approaches to teaching speaking is disadvantageous as both approaches have their limitations (Goh & Burns, 2012). Thornbury and Slade (2006) for example, stated that the direct approach focuses on the component of the language while the indirect approach focuses on the context of use. The focus on one aspect of teaching the skill and neglecting the other forces the learners to be good at the area of focus and weaker on the neglected one. Face-to-face communication is neglected in the direct approach, and the development of accuracy is neglected in the indirect approach. Hence, the direct approach's focus on the discrete element of language and structural accuracy hinders the development of oral skills necessary for everyday communication. On the other hand, the indirect approach's focus on fluency hinders the development of accuracy.
Consequently, an approach to teaching speaking that combines the features of the two approaches is considered to be the solution. Accordingly, Thornbury (2006) introduced a general approach to teaching speaking that consists of three cyclic stages: awareness-raising, appropriation, and autonomy. Awareness-raising is the first stage during which the learners identify their gaps and familiarize themselves with the knowledge designed to address the gap. Appropriation is the second stage at which the learners integrate the new knowledge into the existing one. At this stage, learners progress from other-regulated practices to self-regulated ones. Autonomy is the final stage during which learners use the newly constructed knowledge in a real-life situation. Richards (2008) on the other hand suggested that the functions of speaking determine the teaching approach. Accordingly, each of the speech activities requires different approaches to teaching speaking. He enumerated the skills required for each speech activity. Talk as interaction, for example, involves the following skills: opening and closing conversation, choosing topics, making small-talk, joking, recounting personal incidents and experiences, turn-taking, using adjacency-pairs, interrupting, reacting to others, and using an appropriate style of speaking. Hence, the approach to teaching speaking should aim at helping students master these micro-skills if developing students' interactional skills is a focus. Similarly, the two speech activities involve different micro-skills that determine the approach to teaching.

Context
In Ethiopia, English is a foreign language. Despite the lack of speaking communities, English plays a key role in the commerce and educational system of the country. In education, in particular, English serves as the medium of instruction from grade nine. Although there is much to be done regarding the teaching of English in Ethiopia, this specific study focuses on the comparison of the teachers' classroom practices when teaching speaking and the principles of CLT.
This research was conducted in three government-led secondary and preparatory schools in Gurage Zone, Southern Nations Nationalities Regional State, Ethiopia. These were: Aberus Secondary and Preparatory School, Emdibr Secondary and Preparatory School, and Japan-Ethiopia Friendship Secondary School. Those three schools were chosen purposefully for two reasons. First, the schools were well-established and believed to comprise teachers with varied experiences and insights. Second, they were similar to other secondary and preparatory schools in the country in aspects like administration systems, nature of the classroom, types of instructional materials, qualification of teachers, and the number of students in a class.

Research design
In this study, a case study design was employed. The case study's flexible and context-sensitive nature and the researchers' aspiration to understand the phenomenon in-depth were the reasons why the design was chosen. Stake (1995) classified the case study into three types: intrinsic, instrumental, and collective. An intrinsic case study is employed when researchers aim to understand the intrinsic aspect of a particular case. In the instrumental case study, researchers examine a case in detail aiming to understand the external problem. On the other hand, a collective case study is an extended instrumental case study that involves multiple cases. In this study, a collective case study was employed.

Participants
For the study, four English teachers from the three schools were chosen. The teachers with the highest number of years of teaching services were purposefully selected. This purposive selection is justified because the selection of samples in the qualitative case study does not focus on the representativeness of the population but on the richness of their understanding of the issue (Dörnyei, 2007;Yin, 2018). Table 1, the pseudonyms for the four cases are presented in the first column. Of the four cases, three were males. The teaching experiences also varied from 18 to 6 years. Three of the cases were teaching grade eleven and twelve students at the time this study was conducted, while Daniel was teaching grade ten students. Regarding their qualification, three were first degree holders, but one was a graduate of MA in TEFL.

Informed consent
In scientific research, having the participants' consent is mandatory. In this study, the support letter from the Department of Foreign Languages and Literature, Addis Ababa University, was first delivered to the director, and the four purposively selected English teachers were approached. The teachers were told about the objectives and processes of the study and the expected duration to complete the study. They were also assured that the data from the classroom observations and their responses to the stimulated recall interviews remain confidential and the data analysis would also be done anonymously. Moreover, they were guaranteed that their involvement was voluntary. Once they gave their consent, the teachers signed a letter of agreement to show their participation was voluntary. Once the teachers agreed, the researchers also attempted to obtain the consent of the learners. In obtaining the learners' consent, the classroom teachers played a greater role. They, along with the researchers, told the objectives of the classroom observations to their students. Like their teachers, the learners also gave their consent for the classroom observations.

Data collection tools
In collecting the data, classroom observations and stimulated-recall interviews were used. Nonparticipant and less-structured observations of each teacher's speaking class were conducted to investigate how the teachers taught speaking lessons. The researchers took notes of the classroom practices and also recorded them with an audio recording device. The field notes were preferred as the primary medium as there were procedures and vital signs the audio-recording was unable to detect. The number of the observations was determined by the researchers' judgment concerning whether teachers were repeating their prior practices (i.e., not providing additional data), and also whether they had the required data to answer the research questions, the concept Dörnyei (2007)   referred to as "saturation". Audio-stimulated recall interviews were conducted to elicit the teachers' rationalizations of their classroom practices.

Data collection procedures
Eight weeks were allocated to observe the classroom and do the stimulated-recall interviews. To identify when the teachers taught speaking lessons, they were asked for their weekly lesson plan which they usually prepare and submit on Friday. As the first observation aimed to minimize the influence of the presence of an observer on teachers' and students' behaviors and actions, the researchers did not focus on the speaking lessons in the first round of observations. Thus, the actual observation of speaking lessons started from the second observation. During the classroom observation, the researchers took notes of the focus of the speaking skills lessons, the classroom activities, teachers' and students' roles, the teaching cycles, error correction strategies, and assessment techniques. The classroom observation was also assisted with audio-recording material with the permission of the teachers.
Parallel to the classroom observation, the audio-stimulated recall interview was conducted with the teachers. Immediately after the classroom, the teachers were asked for any convenient time for the stimulated recall. The researchers listened to the recordings of the lessons carefully and selected sections to which they wanted the teacher to react. In their meeting, the researchers made the teachers read the descriptions or listen to the segments and asked for their reactions and justifications. In the stimulated recall interviews, the suggestion Gass and Mackey (2005) gave regarding the timing was duly considered to improve the quality of the data. They suggested a short time interval between the action and the retrospective interview, ideally a time interval that should not exceed two days after the classroom observation. Similarly, Ryan (2012) stated: "The closer the recall is to the event, the more likely it is that the recall itself will not be influenced by memory decay" (p. 155). Hence, the stimulated interview was conducted two days from the classroom observation.

Data analysis
To analyze the data from classroom observations and stimulated recall interviews, qualitative content analysis was used. Dörnyei (2007) reported four phases of the qualitative content analysis process: "Transcribing the data, pre-coding and coding, growing ideas and interpreting the data, and drawing conclusions" (p. 246-257). These four phases of the analytical process were employed in this study. Having the side-by-side collection and interpretation of the qualitative data in mind, the data analysis commenced with the verbatim transcription of data. The transcriptions into word processing documents were done manually by the researchers. As the qualitative component was the primary focus, all the recordings were transformed into texts. Then, the texts were read and reread in their entirety aiming to be familiar with the data. The reduction and interpretation of the data then started with the initial coding of the text. In the initial coding, the texts were broken into segments, and labeling was given to the segments often by using the participants' own words. For the initial coding, hardcopy printouts of the texts were used, and labeling was given on the margin. Then, the initial codes with common concepts were grouped, and higher-order themes were developed. As the first coding may not always work well, the codes were reexamined and relabeled. To assist the researchers in managing the large amounts of qualitative data manually, the data analysis process was assisted with a computerized qualitative data analysis program, OpenCode 4.02.

Trustworthiness of the inquiry
Qualitative researchers take into consideration credibility, transferability, dependability, and confirmability to ensure trustworthiness (Anney, 2014). Credibility refers to the truthfulness of the findings which are believed to be the results of correctly interpreted original data of the participant (Guba & Lincoln, 1982). Of the different strategies used to establish credibility, peer debriefing and member checks were used in this study (Anney, 2014;Guba & Lincoln, 1982). Thus, discussions were made on the data collection and analysis procedures with the three experienced individuals in qualitative studies on language teaching. They also read the final report of the final study and gave comments that helped the researchers to improve the quality of the findings. As far as member checking is concerned, a strategy Lincoln and Guba (1985) considered as the heart of credibility, the interpreted data were given back to the participants to enable them to evaluate if the interpretation was in harmony with what they did and said.
Transferability can be facilitated by providing a thick description of the study and selecting participants purposively (Bitsch, 2005). Hence, information was given about the context, data collection techniques and procedures, sampling strategies, and data analysis methods of the study. Further, the teachers with relatively many years of teaching experience from a relatively well-established secondary and preparatory school were purposively selected as they were believed to have rich and varied insights. To establish dependability and confirmability, the descriptions and/or transcripts of the classroom observations and stimulated recall interviews are kept. In a case study, keeping the records of the descriptions, transcriptions and/or documents to help the readers to evaluate how the research was conducted and its findings is called an audit trail (Bitsch, 2005;Lincoln & Guba, 1985).
In addition to the aforementioned attempts to ensure the trustworthiness of the inquiry, investigators' triangulation was employed to establish validity. Investigator triangulation is one of the four triangulation techniques Denzin and Lincoln (1998) discussed: Data triangulation, investigator triangulation, theory triangulation, and method triangulation. Thus, two of the researchers in this study analyzed the data independently and compared the findings. Validity was ensured as both of the researchers found similar findings and arrived at similar conclusions.

Teachers' observed classroom practices of teaching speaking lessons
In this study, five speaking sessions of each case were observed. Hence, a total of twenty speaking lessons were observed in addition to the four preliminary observations which were done to minimize the influence of the presence of an observer on teachers' and students' behaviors and actions. Of the twenty speaking sessions, eight were taught to students in grade twelve, seven to those in grade eleven, and five for grade ten students. The average number of students in the speaking sessions was 37 while the lessons were taught for an average of 33 minutes. The lowest average number of students (29) was recorded in Daniel's class while it was higher in Abera's class (43). The teacher whose sessions were shortest (averaging 30 minutes) was Awol, while the longest sessions (averaging 38 minutes) were taught by Abera. Table 2 provides information about the speaking topics, the number of students and the minutes used.
Regarding the nature of the topics, fifteen of the twenty topics were originally speaking lessons in the textbook. The remaining five speaking lessons were neither originally speaking lessons nor found in the textbook at all. The teachers' selection of the speaking topics showed that the teachers did not strictly abide by the speaking contents of the textbook. For insistence, Abera in his first session changed the brainstorming questions and reading passage into a speaking lesson. His decision to change the reading lesson into speaking was unexpected due to his description of the speaking parts of the textbook as clear and appropriate. The decision to switch from reading to speaking was not pedagogical but accidental due to most students' failure to bring the required coursebooks to practice reading. On the other hand, Almaz's second and third speaking lessons were not found in the textbook. Similarly, in two of Daniel's classes, the students were made to present grammar lessons that were not initially speaking lessons. The use of lessons out of the textbook and changing lessons of other subjects into speaking were common in Almaz and Daniel's classes.
In addition, even the speaking lessons from the textbooks were modified by the teachers. For example, in Abera's first speaking session, the lesson in the textbook required the learners to engage in a conversation that was characterized by miscommunication to show how it affected communication. The activity in the textbook required the pupils to work in pairs. The classroom teacher however modified the lesson and made the students discuss what communication and miscommunication meant and the causes of miscommunication in groups. Similarly, Almaz's fifth lesson was a modified version of the speaking lessons in the textbook. In the textbook, the issues tourists are concerned about in deciding to visit a certain country were listed, and it required the learners to discuss which of the issues Ethiopia has fulfilled and how to improve the insufficiently attained requirements of the tourists. It also required the students to act as travel agents and persuade tourists to come to Ethiopia. However, the learners in the actual classroom were asked to discuss what tourists could do in Ethiopia, the places they could visit, and the tourist destinations in their region.
The teachers also taught the speaking lessons discreetly though most of the speaking lessons in the textbook were integrated into the preceding lessons, especially the listening and language focus lessons. Daniel's second session, for example, was integrated into the listening lesson in the preceding section. The listening lesson required the learners to listen to a story of Ben and the Devil and answer comprehension questions based on what they listened to in the speaking session, which was the continuation of the listening lesson; the learners were required to re-tell the story they listened to. Re-telling the story was not the only activity as learners were also expected to discuss what they think of the story and tell a traditional story to the class. Though the aforementioned three activities were stated in the speaking lesson of the textbook, the teacher made the students sit in their groups and tell the traditional stories they already knew.
In teaching the speaking lessons, the major procedures the teachers employed were similar despite minor differences. The procedures the teachers went through can be described from two perspectives: teachers and students. The following were the major procedures the teachers went through: providing brainstorming questions, introducing the speaking topic, presenting the discussion questions, explaining the discussion questions, and clarifying technical terms if any, communicating what was expected of the learners during and after the task, encouraging the students to do the activity and facilitating the process, and making the students present what they have done to the whole class. However, it did not mean all these procedures happened in all the classes and all teachers followed the same procedures. For example, the brainstorming questions were not provided in some of the classes, and they did not appear at the beginning of the speaking lessons. The teachers, for insistence, Almaz and Awol, asked brainstorming questions once they introduced the speaking lesson. Almaz, for example, asked the following question at the beginning of her second speaking class: "What does shopping mean?" In her third class, she also questioned: "What do you know about HIV/AIDS?" From the students' perspectives, the procedures went as: listening to the teacher's explanations, answering some of the brainstorming questions, copying the discussion questions, participating in group discussions, taking notes for the presentation, and presenting what they have done in groups to the whole class.
Briefly, the classroom procedures and what the teachers and students did in the process can be categorized into three major phases: pre-task, task, and post-task. The pre-task refers to the first and teacher-dominated phase when the learners were introduced to the speaking lessons and provided with the discussion questions. In the task phase, the students did individual, pair, or group work activities while the teacher was moving around the class and offering help. In the post-task, the students, especially the representatives, presented what they did in groups or pairs to the whole class.
In the classroom, three modes of interaction and different speaking activities were employed. The three modes of interaction the teachers employed were group, pair, and individual work. Group work activities were used in fifteen, individual work in five, and pair work in three of the speaking classes of the twenty sessions. This showed that group work activities were the most used mode of interaction in speaking classes. The modes of interaction however were employed either in combination or alone. Concerning the classroom activities, six speaking activities were employed: discussion, presentation, role-play, picture description, storytelling, and debate.
Regardless of the number of the classroom activities, the discussion was the most used as it was employed in fourteen of the sessions. It was followed by a presentation while the remaining four activities were only used in seven of the sessions. All the group work activities, except the debate, were discussions. That made group discussion the most used classroom speaking activity. Despite the frequent use of group discussions, the classroom activities were sometimes jointly used. Awol for example, started the speaking task in one of his classes with an individual task, proceeded to group discussion, and finally invited the learners for a presentation. The common combination however was between group discussion and presentation as it was observed that almost all the group discussions were followed by a presentation.
The teachers, except Daniel, believed that an English-only policy should be applied in the classroom. The need to use only English in the classroom was the issue the teachers warned the learners to take into consideration. Of the twenty speaking sessions, the learners were warned to use only English in ten of them. The warnings were given either before and while doing the task. Despite that, Amharic was used by most of the learners during the group discussion.
The learners' overt use of Amharic was the reason why the teachers repetitively warned the students to avoid using their first language. The teachers also used Amharic in the classroom but were limited to particular situations. Abera for example, used Amharic in inviting students for presentations and encouraging them to start their presentations when they hesitated. Almaz used Amharic in activating the existing groups for discussions or forming pairs. Contrarily, Daniel and his students freely used Amharic.
Most of the speaking sessions of this study were characterized by a lack of participation. No or few hands were seen whenever the teacher asked questions. For the pair and group work activities, the learners required the teachers' repetitive reinforcement to form the pair or sit in their groups and also to start the activity though it seemed to be difficult to stop once they got going. The learners especially were uninterested in a classroom activity that demanded them to stand in front of their classmates and make a speech. The group discussions were also dominated by one or two of the members while others were passively listening to them and writing down notes for the presentation. Thus, it was challenging for the teachers to find willing presenters. What the teachers commonly did was that they called the names of certain students though they did not raise their hands. Although the students the teachers named were group leaders whom the teachers regarded as outstanding, as reported by Abera, they even hesitated to come out and wanted the teachers' repeated encouragement. The learners' replying to the teachers' questions or presenting used few words and finished it shortly or switched to Amharic. For insistence, the students who were allowed to do fifteen minutes of pair work finished their presentation in less than half a minute. To the teachers, the learners' inability to express themselves in English was one of the fundamental reasons for the dearth of participation.

The relationship between the teachers' actual practices and the principles of CLT and their rationalization
As stated in the English syllabuses for grades 9 and 10, and 11 and 12, the language teaching approach that underlies the syllabuses is communicative (Ministry of Education (MoE), 2009b). As the syllabuses are based on the concept of Communicative Language Teaching (CLT), the governing notions mentioned in the syllabuses are: a) language is a tool of communication, b) learners learn and practice meaningful language that has real purpose and context, and c) the four basic skills are the focus while grammar, vocabulary and social expressions are integrated into the practice of the skills. Richards (2006) comprehensively defined CLT as: "A set of principles about the goals of language teaching, how learners learn a language, the kinds of classroom activities that best facilitate learning, and the roles of teachers and learners in the classroom" (p. 2). Larsen-Freeman and Anderson (2011) listed the important principles of CLT for the teachers to support their students' language development. Richards and Rodgers (2014) precisely enumerated the five basic principles of CLT. The five characteristics of CLT Nunan (1991) mentioned also shared a lot in common with the principles Larsen-Freeman and Anderson and Richards and Rodgers reported. (see Literature Review for further discussion) Accordingly, the matching of teachers' teaching practices with the syllabus is done with reference to principles of CLT. The goal of language teaching in CLT is developing communicative competence that comprises aspects like: the capability to use a language for a variety of purposes, the ability to adapt a language to the situation, the skill to make and comprehend a variety of texts, and the know-how to preserve conversations (Larsen-Freeman & Anderson, 2011;Nunan, 1991;Richards & Rodgers, 2014). As revealed from the stimulated recall interview, the teachers' thoughts about the goal of teaching English in general and speaking, in particular, agreed with the principles of CLT. The teachers believed that the goal of learning English was to develop the skill of communication through the language. Likewise, they reported that the goal of teaching speaking was to enable learners to use the language for communication at any time and in context.
In CLT, learners learn a language through the process of communicating in it (Richards, 2006;Richards & Rodgers, 1999). This is to mean it is the target language that should be practiced and used in the language classroom. Though the teachers described it differently, they believed in practicing the language. Similarly, Larsen-Freeman (2000) stated that the target language should serve as the means of communication in the classroom. Notionally, the teachers' thoughts about how language learning takes place were in line with the principles of CLT as their reaction to the effective way of teaching speaking showed their belief in exposing learners to authentic and functional language use through real-life communicative activities. However, the classroom practices failed to match with teachers' thoughts and principles of CLT in this regard. As witnessed by the researchers and approved by the teachers, the students used Amharic more than English in doing classroom activities. As shown in the following extract from Awol's interview, the teachers assumed that the students' poor background in English is the reason for the use of Amharic in speaking classes.
Really, really, it is adversely-they have poor command of English. Now I am teaching grade twelve and eleventh; my students are somewhat-they have poor background. I do not know why. Really, it needs intervention-immediate intervention. Even students who have learned for consecutive twelve years, they are observed being unable to express themselves, even easy self-explanations, they are unable to do that. So they use Amharic.
In CLT, learning activities that engage learners in meaningful and authentic language use (i.e. pair work activities, role plays, group work activities, project work, etc.) are used (Richards, 2006). Though Richards and Rodgers (1999) stated that the activities compatible with CLT are unlimited, the findings from the classroom observations revealed that group discussion followed by presentation was the most employed. Further, teachers' selection of the classroom activities seemed to be influenced by the contextual factors more than their engagement of the learners in meaningful and authentic language use. Some of these contextual factors were: the learners' experience of practicing the activity, the number of students in the class and the given time, and the adequacy of time to cover the coursebook.
The consideration of teaching as facilitating and teacher as a facilitator has become a dominant notion in the field of education in general and language teaching in particular. Likewise, the teachers in communicative classrooms are assumed to be facilitators of the students' learning (Desai, 2015). According to Richards and Rodgers (1999) and Breen and Candlin (1980), the teachers in CLT are not only facilitators but also organizers of the resources and a resource themselves, researchers, learners, need analysts, counselors, and group process managers. Fosch (2017) summarized the roles of a teacher in CLT into three: facilitator, co-communicator, and advisor. The participants' views about the roles of teachers were consistent with the aforementioned roles of CLT teachers. The teachers revealed that the role of the classroom teacher in speaking class is to facilitate communication. Awol, as shown in the excerpt below, also expected the teachers to motivate the learners by bringing authentic activity to the classroom and scaffolding their practices.
Students need to be motivated and interested in the activities. Actually if it is related to their daily life, they will be motivated. So students should interact actively and productively at the same time. And the teacher should facilitate the classroom activities, help and support, scaffolding. The teacher should play the facilitators role, supportive role, not always intervening the activities.
However, the teachers reported that their attempts to actualize their thoughts were not as fruitful due to the influence of the contextual factors, especially learners' lack of autonomy. The role of learners was one of the areas where the classroom realities were inconsistent with the principles of CLT. According to Fosch (2017), CLT advocates that the teachers and the learners are accountable for the instructional process. Similarly, Breen and Candlin (1980) stated that learners should be responsible for their learning and share responsibilities with other students as well as with the teacher. They, therefore, should be committed to undertaking communicative activities with other participants within the group. These showed how learners' roles are essential in communicative classrooms. As the findings revealed, the students' classroom practices were not in line with the principles of CLT as far as the teaching of speaking was concerned. They for example, often waited for the teachers' repetitive encouragement to be in groups and to start the discussion. In their discussion, though CLT supports that language is learned through the process of communicating in it (Richards, 2006), they most of the time used Amharic. The teachers also reported that the students switched to English when they approached them and switched back to Amharic when they moved to another group. Besides, the group discussions were dominated by one or two students. Often, the teachers' invitations for presenters after the group discussions were not fruitful. These findings of the students' classroom behaviors, although they could be related to the contextual factors, could also be the manifestation of students' lack of responsibility which is the heart of CLT. In other words, it was not only the factors like learners' lack of background skills or the class size that were influencing the teaching of speaking but also lack of commitment. The learners, as revealed in the following extract from Abera, were not ready to use the chances they were given to improve their speaking ability.
Our students do not like to attend the speaking lesson. If they have chances to leave the class, they may even leave the class. They are discouraged to attend the speaking lesson, and they are not actively engaged in the activities in the speaking section.
The teachers' classroom practices of error corrections were also compared to the communicative views of error and error correction. In CLT, as Richards (2006) stated, learners' errors indicate that they are building up their communicative competence. Desai (2015) also reported that the instructional process should not be interrupted by the classroom teacher for the sake of corrections. Thus, it is possible to say that corrections of errors in communicative classrooms may be absent or infrequent. Consistent with the communicative views of error, teachers considered errors as an inseparable language learning process and indicators of students' progress and did not believe in frequent and immediate corrections. The teachers' classroom practices were also in agreement with their cognitions. In Daniel's class, for example, those learners who have been presenting were making mistakes. He neither interrupted them nor gave corrections at the end. Asked about the reason, he said: Even I myself make error in speaking skill, it should be manageable. Even when those students are speaking, I never interrupt them. After they finish their speaking, I may sometimes give them feedback what to be corrected by taking notes while they are just communicating or just speaking.Errors are normal.

The contextual factors that influence the teaching of speaking and/or their enactment of the principles of CLT
As Abera reported that the influence of the contextual factors started at the planning stage as he attempted to match his selection of classroom speaking activities with the problems in the classroom regardless of his cognitions.
In preparing the lesson, I think about the students-the activities the students can do. Also, the time-the minutes given. Also, I try to relate the lesson to the previous years entrance exams. That is good for grade twelve learners. So, sometimes I change the textbook speaking content.
As he reported and was also observed, he regularly employed group discussion though he believed that one of the fundamentals of effective speaking instruction was the use of a variety of classroom activities. As he reported, he did not plan to employ classroom activities like simulation and role play as the students were not interested in them and the classroom facilities were inadequate. Almaz also reported that the contextual factors did create an inconsistency not only between the anticipated speaking teaching techniques and the actual teaching practices but also between the objectives of teaching speaking, which she described as enabling learners to orally express themselves in English, and the classroom speaking teaching practices.
Accordingly, teachers reported diverse contextual factors that influenced the teaching of speaking. Some of the factors were: lack of basic background skills, lack of speaking learning experiences, using Amharic in the classroom, fear of making mistakes, stage fright, lack of interest in speaking lessons, misconceptions about speaking instruction, washback effects of the university entrance examination, low and uneven participation, previous teachers' inadequate focus on speaking, the vastness of the textbook, large class size, lack of exposure and discouraging culture.
The teachers' report of the contextual factors revealed various points. First, the students' lack of basic background English language skills in general and speaking in particular was the predominant factor. Second, the factors the teachers mentioned seemed to be interrelated. The teachers, for example, believed that the teaching of speaking was influenced by the learners' lack of basic oral skills, lack of experience of practicing classroom speaking activities, lack of interest in speaking lessons, previous teachers' inadequate focus on speaking lessons, and teachers' misconceptions. It is obvious that the learners' are not expected to be good in the group they do not practice well. Similarly, it is difficult to expect the learners to be good at the skill their teachers give little attention to. Besides, the learners' lack of interest in speaking lessons was associated with the washback effects of the university entrance examination and fear of making mistakes. Third, some of the factors the teachers mentioned were allied with other reasons. The washback effect of the university entrance examination, for example, was grade specific as it was only mentioned by teachers teaching grade 12 students. In this regard, Abera said the following: Most of the time, I have been teaching grade twelve students and as grade twelve students, they are exam focused, so the worry they have also matters. Their interest I think is on the exam. They want me to practice them the previous sheets. They have no interest to speaking.
In addition, the teachers' evaluation of the influence of the coursebook was influenced by their evaluation of its effectiveness. Abera for example, considered the coursebook as effective in terms of its content and presentation. To Almaz, the textbook was one of the influencing factors because she regarded the coursebook as vast, and some of the speaking contents as ineffective.

Discussion
In this section, the English language teachers' practices of teaching speaking lessons, the comparisons of the practices with the principles of CLT, and the contextual factors that affect the teachers' enactment of the underlying principles are discussed in relation to the existing literature and previous research findings.
Theoretically, the classroom activities the teachers tried to employ were communicative, promoting interaction among the students. This matched with what Alonso (2014) reported that most of the classroom activities the teachers attempted to employ were communicative and interactive, focusing more on fluency and making the teacher a facilitator. Although the activities they attempted to employ were communicative, discussion exercises were overly used. Most classroom studies done on oral skills also proved that discussion exercises were the most preferred and used ones by English language teachers (Gudu, 2015;Kaski-Akhawan, 2013;Taddese, 2019). Though different factors could influence the teachers' overuse of a certain classroom activity, the existing mixed ability groups, known as One-to-Five, were found to be one of the reasons for repetitive use of group discussion by the teachers in this study. They were ready-made groups formed at the beginning of the academic year by the homeroom teachers based on students' ranking of the previous year. Although group discussion, like other fluency-controlled activities, provides enough time to practice the language in context, all classroom activities need to be used equally to promote learner participation and thus enhance the acquisition of oral and communicative competence (Gudu, 2015).
The discussion exercises were followed by presentations. Hence, the teachers' demand of one of the students to come front and present what they have discussed in the group was a common feature of the classroom activity. This originated from the teachers' belief that having a presentation next to the discussion boosts students' devotion and participation in the group discussion. To make the participation in the group discussion even, the teachers sometimes warned the students that they would call the name of the presenters randomly. However, finding willing students for the presentation and making the students start speaking even once they went out was a difficult task for the teachers. In this regard, Tesfaye and GTsadik (2015) reported that activities that demanded the learners to speak in front of the learners were the most anxiety-provoking. More importantly, they reported that a sudden and random calling of names is one of the sources of anxiety in the classroom. The discuss-and-present system also seemed to be influencing students' participation in the group discussion as they focused more on listing points for the presentation than on being involved in the discussion.
The classroom activities were compared to the four characteristics of a successful speaking activity Ur (1991) recommended. The characteristics are: learners talk a lot, participation is even, motivation is high, and language is of an acceptable level. Briefly, the practice of the classroom speaking activities was not satisfactory when compared to the aforementioned characteristics. Though the regularly used classroom activity, group discussion, is known for its provision of a lot of talk time to the learners and facilitation of even participation (Karam & Abdulla, 2021), the actual classroom activities were influenced by the use of Amharic and the dominance of some students. This finding agrees with what Tuan and Mai (2015) who reported that the students tended to use their mother tongue when they discussed in groups. In education, learners' participation plays a key role in the progress of instructional practices. Particularly, learners' participation is pivotal in the learning of a foreign/second language as students are expected to engage in classroom activities. Though the necessity of learners' participation has even been advocated by theories such as learning by doing (Dewey, 1997) and contemporary language teaching approaches like Communicative Language Teaching (Larsen-Freeman, 2000), speaking classes are known for the dearth of learners' participation (Kedir, 2017;Gudu, 2015).
In this study, the sequence of the classroom activities has been considered. The findings showed that the teachers' sequence of classroom activities resembled the first two phases proposed by Willis (1996) to the implementation of Task-Based Instruction, which is described as the extension of CLT (Richards & Rodgers, 2014). The language focus analysis proposed by Willis has got three major phases: pre-task activities, task cycle, and language focus. In the pre-task stage, the learners are introduced to the issue and activities of the lesson. In the task phase, the students do the task in pairs or groups; the teacher walks around and monitors and encourages the students, and the students report their work to the whole class. At last, language focus analyses are done and activities like a choral repetition of phrases, sentence completion, dictionary reference, etc. are practiced based on the language analysis. As the findings from the classroom observation revealed, the sequence of activities was similar to the first and second phases as classes began with an introduction of the topic and task and ended with students' reports of their works to the whole class. In other words, there was no language focus analysis in the observed classrooms.
The comparison between the teachers' classroom practices of teaching speaking with the principles of CLT revealed that the teachers understood the notions of CLT, but there was incongruence between what was articulated and what happened in the classroom. The teachers' attitudes towards CLT were also positive. In other words, the teachers verbalized the principles of CLT and also believed that speaking should be taught communicatively. This was in line with what Pitikornpuangpetcha and Suwanarakb (2021) and Nam (2005) reported that the EFL teachers were optimistic about CLT. However, it was found that most of the teachers' actual classroom practices were not congruent with the principles of CLT. Similarly, Rahman et al. (2018) found that teachers articulated the importance of CLT but did not implement its principles in the actual classroom. Fattah and Saidalvi (2019) also found that teachers did not fully implement CLT. In this regard, Yusof and Halim (2014) stated that the implementation of the communicative methodology in the EFL context is not an easy task due to its demand to use the language productively and receptively.
The teachers' employment of the principles of CLT has been influenced by different contextual factors. According to Borg (2003), contextual factors are the social, psychological, and environmental realities of the school and classroom instigated by parents, school leaders' requests, community, curriculum directives, plans and policies of the school, colleagues, standardized examinations, and resources. The contextual factors the previous researchers (Nevisi & Farhani, 2022;Taddese, 2019;Tesfaye & GTsadik, 2015;Al Hosni, 2014) reported as ones that were influencing the teaching of English in general and speaking in particular matched with what the teachers in this study mentioned. More importantly, consistent with the findings by Farrell and Lim (2005), the factors that influenced teachers' enactment of their assumptions were found to be interconnected. Despite their interconnectedness, they were also diverse, enabling them to be put into different categories. Categorizing the problems that influence the teaching of English in general and speaking in particular into different grouping is also done by other researchers. Musliadi (2016), for example, classified the problems of teaching speaking into two major categories: internal and external. According to the researcher, the internal problems are those that come from the students (i.e. the first language, age, innate phonetic ability, identity and language ego, motivation, etc.) while external problems come from outside students (i.e. lack of time to practice speaking, some irrelevant materials, ineffective language teaching approach used in teaching speaking, etc.). Rababa'h (2005) reported four major factors that cause difficulties in speaking instruction: learner-related factors, teaching-related factors, curricular factors, and environmental factors. Accordingly, most of the contextual factors the teachers reported were internal according to Musliadi's categorization and learner-related based on Rababa'h's classifications. Table 3 presents the contextual factors the teachers in this study reported within the framework of Rababa'h's (2005) categorizations.
In the Ethiopian English syllabuses for the secondary and preparatory schools, it is stated that the syllabuses are designed based on communicative views and communicative language teaching notions (Ministry of Education (MoE), 2009b). Accordingly, the syllabuses indicated that the teaching of English should be guided by three principles: a language is a tool for communication, learners should learn and practice meaningful language that has purpose and context, and the four basic language skills should be the primary focus of the instruction. Though the teachers were aware that they were expected to employ CLT and even understood its principles, the classroom practices were not congruent with the expectations. Different contextual factors are responsible for the incongruence between what was expected to be employed and what actually was employed.

Conclusion and implications
In this study, the EFL teachers' practices of teaching speaking were explored and then matched with the principles of CLT. Besides the matching, teachers were asked to rationalize the practices and mention the factors that influence the teaching of speaking. Of the four major language skills, speaking was chosen to be the focus of this study as it was one of the language skills foreign language learners encounter difficulty with. Four English language teachers took part in the study, and data were collected through classroom observations and stimulated recall interviews.
Considering the findings presented above, the following conclusions can be drawn. First, it was understood that the teachers did not stick to the speaking contents of the textbook; they sometimes brought speaking lessons that were not part of the textbook. Second, the teaching cycles the teachers went through in teaching speaking lessons were similar, allowing the identification of three phases: pre-task, task, and post-task. Third, group discussion was the most used classroom activity. Fourth, the actual classroom practices were found to be incongruent with the principles of CLT. Though different contextual factors influence the teachers' enactment of the underlying principles of the syllabuses, the most referred to by the teachers were learner-related factors.
The findings presented above suggest the following implications: A. The teacher training curriculums should balance the teaching of the theories with the rehearsal of the theories in the actual classroom setting. The findings revealed that the teachers understood that they were expected to implement CLT and even verbalize its basic principles, but the classroom practices were not congruent with the principles of CLT. The incongruence, as the teachers reported, might happen due to the influence of the contextual factors they mentioned.