Retention of Mother Tongue as Instructional Language at the Junior High School Level in Ghana: Teachers’ and Students’ Perspectives

: The early-exit transitional education policy in Ghana demands teachers at the junior high school level to exclusively use English as the language of instruction. However, there are enormous reports that some teachers constantly violate the implementation of this policy by retaining the Mother Tongue (L1) of students as the instructional language. The present explanatory case study, therefore, aimed to investigate the rationale behind the retention of students’ L1 (Twi) as the language of instruction as well as how such practice affects students’ academic progress. The study included five public schools in the Sunyani-West District of the Bono Region, Ghana. Data from 100 participants (including 25 teachers and 75 students) were gathered through classroom observations, focus group discussions, and interviews. It was found that teachers retain students’ L1 at the junior high school level because they believe it is the most effective means of ensuring a better understanding of lessons, which would, otherwise, be difficult given the low English proficiency of the students as well as some teachers. The findings of the study also indicated that students had much interest in L1-based instruction because they believed that the use of their L1 created an enabling environment that promoted active participation and learner-centred pedagogy. In this regard, it was recommended that, given the difficulties teachers and students encounter in implementing the early-exit policy, the late-exit transitional education policy should be adopted by the Ghana Education Service.


Introduction
Language impacts classroom communication, comprehension, and academic achievement. It is considered as an inevitable element in education as it is the channel through which knowledge is shared between instructors and students (Lupogo, 2014). It contributes immensely to the effectiveness of students' learning, regardless of the subject under consideration. Thus, progress in any form of education largely depends on the effective use of language. Learners integrate new concepts when they listen, speak, read, and write. This implies that if the learners have low proficiency in the instructional language, their learning is likely to be affected adversely (Anani, 2019). Relevant research on language use and students' academic achievement has shown that students whose language of instruction is a foreign language produce poor academic performances compared to those who receive instruction in their L1 (Bull, 1955;Cummins, 1984;Nicholas & Lightbown, 2008;Pinnock & Vijayakumar, 2009). Unfortunately, over 221 million children

Literature review
The literature review section begins with a historical account of the state of language in education policies and continues with a review of theoretical issues related to foreign and mother tongue education.

Historical overview of Ghana's language in education policy implementation
As with most African countries' education systems, the implementation of the language the education policy of Ghana appears to be influenced by socioeconomic, linguistic, political, and historical conditions (Ssebbunga-Masembe et al., 2015). Scholarly debates on issues pertaining to the topic under discussion divulge sundry issues causing inconsistencies with regard to the implementation of the language in education policy of Ghana. This subsection of the paper seeks to offer a brief insight into how the language in education policy of Ghana has experienced inconsistencies.
The earliest form of bilingual education in Ghana can be traced back to the days of the Christian missionaries and colonial rule (Owu-Ewie, 2006;2013;Owu-Ewie & Eshun, 2019). The debate about the use of foreign languages in the classroom started with the establishment of the castle schools (Owu-Ewie & Eshun, 2019). Between the years 1529 and 1874, no Ghanaian language was employed as the language of instruction. The Portuguese, the Dutch, the Danes, and the English, utilized their languages as a medium of instruction during their respective regimes (Owu-Ewie, 2013). After this era, however, the missionaries, especially those from Basel and Bremen, contributed greatly to the development and use of the local languages of Ghana in education. For religious and educational reasons, the Basel missionaries developed the Akuapim Twi, whereas the Bremens and Wesleyans developed the Ewe and Fantse, respectively (Graham, 1971). The uncertainties concerning which language to legitimise as the language of instruction were particularly prominent in the post-colonial period because a change in government meant a change in the language policy. The British proclaimed the 1882 Ordinance, which was the first education ordinance to encourage a better education. Under this ordinance, the English language was the only language accepted to be used as a medium of instruction. The Ordinance for the Promotion and Assistance of Education in the Gold Coast Colony was the name given to ordinance (Owu-Ewie, 2006). In 1925, however, the Guggisberg's Law amended the previous ordinance by placing emphasis on the use of local languages as the language of instruction for the first three years of education (primary 1-3), after which the use of local languages would be phased out in favour of English-only policy. Interestingly, many indigenous elites saw this initiative as a deliberate colonial attempt to provide Africans with substandard education. Despite the resistance, the plan was implemented. The policy was adopted notwithstanding the objections that had been leveled against it. It also stipulated that the native languages must be taught as a subject from elementary school to higher education (Andoh-Kumi, 2002).
Subsequently, the Accelerated Development Plan (ADP), introduced during the 1951 educational reform, also emphasised the use of the local languages as instructional language at the early stages (P1-P3) of education, and as a subject to be studied in upper grades (Owu-Ewie, 2006). However, after the revision of ADP in the year 1957, the use of local languages was prohibited, thereby paving the way for English-only instruction from the first year of schooling to the tertiary level of education in Ghana (Owu-Ewie, 2013). This policy could not persist because, upon the recommendations of the Kwapong Committee, led by Professor Alex Kwapong, the National Liberation Council (NLC) reverted to the use of L1 as an instructional language as was previously practised under the Guggisberg administration. This reversion was informed by the Committees' realisation that most teachers were deviating significantly from the implementation plan for the English-only policy since they were rather using the local languages for instruction (Owu-Ewie, 2013). Subsequent administrations (the Progress Party in 1970, the National Redemption Council in 1974, and the Provisional National Defence Council in 1987 reaffirmed the use of L1 for the first three years of education. The historical account given above shows that there have been a series of alterations regarding the language in the education policy of Ghana. Nonetheless, one of the most prominent and recent changes which raised a great debate among educationists in Ghana was the change initiated by the New Patriotic Party (NPP) government in the year 2002. This policy abolished the use of the local languages as the language of instruction at all levels of education. The rationale for the abolishment of the use of L1 was that teachers, especially those in rural areas, were abusing the use of L1 to the detriment of the English language. Ansah (2014) asserts that this decision by the government to abolish the use of L1 was backed by parents. This policy was eventually reversed in 2007 when it was faced with lots of criticisms by different educationists and interest groups. Table 1 summarizes the historical overview of Ghanaian languages in the education policy implementation. As presented in Table 1, the current policy requires that L1 of learners can be used only from Kindergarten to Primary 4. Yet, at the JHS level, reports indicate that teachers still use L1 as a language of instruction; hence the need for this study, which, particularly, seeks to ascertain the rationale behind the retention of L1 as instructional language at the JHS level.

Theoretical basis of the use of L1 as language of instruction
The contention over which language is deemed appropriate for promoting quality education appears to be a perennial issue among scholars and educationists (Ball, 2010). Most of these debates are premised on the linguistic situation of African countries, because the majority of children are, to date, receiving instruction in foreign languages (Mackenzie & Walker, 2013). Obsession with using English as the language of instruction, for instance, is prominent among most teachers in Ghana. This is because, per the language in education policy of Ghana, teachers teaching at the lower primary (Primary 1 to Primary 3) have been mandated to stick strictly to the use of the children's L1 as the main language of instruction. In line with the tenets of the early-exit model of bilingual education, this framework provides children with the opportunity to gain content knowledge and also gain linguistic proficiency in the English language before they transit into English-only instruction. The foregoing notion reflects the assertion that "... the level of L2 competence which a bilingual child attains is partially a function of the type of competence the child has developed in L1 at the time when intensive exposure to L2 begins'' (Cummins, 1979). Cummins' assertion lends weight to the fact that if Ghana's educational model is effectively implemented, most students are likely to attain English language proficiency at the JHS level, which would, in turn, make the implementation of the English-only instruction policy successful.
Several theoretical assumptions and empirical studies support the idea that effective instruction in one's L1 is a prerequisite for the development of a foreign language and cognitive development. Cummins submits that there are two thresholds for the attainment of a well-grounded bilingual status. As Daller and Ongun (2018) suggest, there is a likelihood that bilingual students will face academic difficulties if they fall below a certain threshold of proficiency in both languages. In examining striking issues relating to the transition of Grade 4 students to English-only remote communities in South Africa, Steyn (2017) observes that the difficulty teachers encounter transiting into English-only instruction at the upper level of primary education is the inadequate native language foundation and the standard of teaching offered. It was, thus, recommended that the age of entry in Grade 1 should be seven years and that the real transition time should be extended. This provides credence to Cummins' theoretical assumptions.
Another theoretical framework that provides credence to L1 use in education is the Common Underlying Proficiency (CUP) (Cummins 1976;1979;1980;1991). This framework supports the view that there is a close link between one's L1 proficiency and foreign language proficiency (Cummins & Swain, 1986). Several studies support this hypothesis, as it has been shown that English language learners who are effectively engaged in a bilingual program perform relatively better than their monolingual peers (Greene, 1997;Francis et al., 2006;Guglielmi, 2008). With the acquisition of content knowledge, studies in the Ghanaian context show that the local languages offer many academic benefits to students (Kwapong, 2006;Opoku Amankwa, 2009;Mwinsheikhe, 2009). Despite these research findings, it has been shown that in situations where students have limited exposure to the English language, as in the case of Ghana, their literacy development could be impeded . For instance, Amua-Sekyi and Nti (2015) express the view that the limited proficiency of college students in Ghana is a result of teachers' obsession with native language usage in and out of the classroom.
The general impression, based on the literature review, is that native language use is undoubtedly a catalyst for cognitive and foreign language development among minority language students. However, for students to acquire a well-balanced bilingual status, they must be effectively engaged in both L1 and L2 instruction. However, reports suggest that there has always been an inappropriate implementation of the early-exit policy at the primary level due to the issues of linguistic diversity and insufficiency of instructional materials (Owu-Ewie, 2006;Fredua-Kwarteng & Ahia, 2015), and thus the early-exit model is, arguably, having a minimal impact on students' English proficiency since the situation, as observed in some JHS, seems to suggest that teachers are struggling to use English only as the instructional language, thereby violating the policy. Since the policy is not being implemented as expected at the JHS level, it was considered imperative to further investigate issues related to factors that accounted for that and to find out the perspectives of both students and teachers on how LI retention has impacted learning at the JHS level. This was important because limited or no research on language policy has specifically looked into that.

Research methods
This explanatory case study employed interviews and focus group discussions with teachers and students, respectively in five purposively sampled JHSs in the Sunyani West District of Bono Region, Ghana. The rationale for focusing on these schools is that though students are of different ethnic backgrounds, they are all competent in Bono, which is the predominant language in this district. In total, 100 participants engaged in this explanatory case study. In all, 5 teachers from each school were interviewed. Also, 15 students from each school were engaged in a focus group discussion. Thus, the total number of participants engaged in this study was 100 (25, teachers and 75 students). These participants were selected for a series of classroom observations to ascertain teachers who frequently use the local language as the language of instruction. A semi-structured interview guide was used as the data collection instrument because, with this type of research instrument, the interviewer can explore unanticipated issues as and when they emerge in the interview (Taylor et al., 2016). The interviews with teachers were conducted in English, while the focus group discussions with students were done in Twi, which was later translated into English with the help of competent translators from the Department of Ghanaian Languages and Linguistics, University of Cape Coast. The decision to interview the students in Twi was informed by the realisation that most of the students experienced difficulties communicating effectively in the English language. Hence, a Twi version of the interview guide was developed to aid the discussion with students. In particular, the interviews were done during break hours in order not to interfere with students' study hours. The Zoom H5 Handy recorder was used to record the interviews.
The gathered data were analysed inductively by following the Qualitative Research Data Analysis technique of Creswell (2014). Following this approach, the data were first transcribed and translated. After reflecting on the data, the author coded them and organised the identified themes into categories. The categorisation was based on the research questions. These themes were then used as the basis for building arguments in the analysis.

Results
This section is devoted to discuss the results of the study. The discussion focuses on factors that accounted for the continuous use of L1 as the language of instruction at the JHS level and how that impacts classroom interaction. The final discussion explores how students at the JHS level respond to the retention of L1 as the language of instruction. The major themes that emerged from the data show that teachers retain the L1 of students because instruction in English affects classroom communication. It also emerged that instruction in the L1 of students facilitated effective instruction which leads to a better understanding and thus, students exhibited much appreciation for instruction in their native language.

Factors that account for the retention of L1 as instructional language at the JHS level
Teachers engaged in the interviews expressed the view that the core mandate of every teacher is to ensure that knowledge has been acquired by the students. It is, therefore, believed that the worth of the teacher is measured by his ability to achieve successful student learning. Thus, to ensure that students understand what is being taught, the language that they are highly proficient in must be used to teach. When they were asked to give a reason for using Twi as the language of instruction despite the spelt-out plan for implementing the English-only policy at the JHS level, a science teacher stated the following: The most important aspect of teaching is the students' ability to understand whatever is being taught at the end of the lesson. So, I think violating the policy is not a good idea. In my opinion, I wouldn't support that, but we do so to help the students comprehend lessons.
Another teacher from the same JHS confirmed the preceding statement, stating that; It is better for students to understand than to use English throughout, and at the end of the day, they are not getting what you are saying.
The teachers' response thus contradicts the widely held belief that most schools violate the English-only instruction policy due to a lack of rigorous monitoring (Owu-Ewie & Eshun, 2019). Teachers indicated that students' ability to demonstrate an understanding of what they have been taught is the primary essence of schooling. For this reason, they would rather violate the policy to facilitate learning. As a justification for the breach of the policy's implementation, it was made known that violating the policy is necessary because if teachers fail to use students' L1 frequently, comprehension of lessons may be quite challenging.
Teachers were also of the view that exclusive use of English is quite challenging because most students experience difficulties regarding the use of English as a communicative medium. They agreed that even teachers sometimes struggle to articulate proficiently when explaining quite difficult concepts to the students. Hence, to avoid any difficulty with the use of the English language, they use Twi as the medium of instruction. Concerning issues of low English proficiency, a male teacher expressed the view below.
If you use English only as the instructional language, you may find it very difficult to express yourself on certain topics. However, when you use Twi, you would be able to explain concepts clearly to the understanding of the students.
In line with the opinion expressed above, a science teacher said that there could be instances where teachers find it difficult to expound on technical concepts using English only. This is how he puts it: I teach science, and there are some complex concepts in science that are not necessarily English terms. Students may not be familiar with some of these terms. For instance, I once made mention of 'Photosynthesis'. Students may not understand certain terms like these if they adhere so much to English-only instruction. This teacher believes that his proficiency in the English language is not high enough to allow him to express himself when explaining complex concepts. One issue of concern is that even if the teachers are proficient enough to use the English language only, the language competence of the students remains questionable. This is a situation the teachers ascribe to limited exposure and use of the English language outside the classroom. Commenting on this, a teacher reveals that: Unlike students in this area, students in cities such as Kumasi, Sunyani or Accra are always exposed to the English language wherever they go. Yet, I don't think teachers in such places would use English only for instruction. This suggests that the cause of students' low proficiency in the English language is their limited exposure to the language outside school. It could, therefore, be contended that using English as the sole language of instruction in such a rural setting where exposure to English is relatively limited would be a setback to the academic progress of the students. As a result, the local language, and in this case, Bono, becomes the perfect means of facilitating comprehension among students.

How the use of L1 as instructional language impacts classroom interaction and students' learning
Findings from the interviews and a review of students' formative assessment results revealed that the use of Bono in rural and urban schools fosters the academic progress of students. Data gathered from the interviews revealed that the use of L1 in the classroom helps students to transfer knowledge, to emancipate them from affective factors such as anxiety and low self-esteem, and also to help the teacher to deliver their lessons easily.
When asked whether the students found any difficulty in transferring the knowledge they acquire in Twi instruction to English, a teacher responded:

I have never seen a child providing answers in Twi instead of English when you give them a test. You would rather realise that he can transfer whatever knowledge he has acquired in Twi-instruction into the English language perfectly.
A social studies teacher in the same school also corroborated this claim as follows: When students acquire the knowledge in L1, it becomes easier for them to demonstrate it on paper because they can write the English language. Students supported this claim as they opine that despite being taught in their L1, they do not experience much difficulty in transferring what they have learnt into the English language. A student responded to this as follows: My problem is not about writing English. I can read and write everything in English. The only problem I face is speaking in class. When the teacher uses English to ask me a question in the classroom, I find it very difficult to express myself well. However, when writing exercises and assignments, I don't face so many difficulties.
Teachers' and students' opinions on the issue under discussion seem to suggest that students at the JHS level may be suffering from a low Cognitive Academic Language Proficiency (CALP) in English. As Cummins (1978) warns, the mere fact that students can use English for communicative purposes does not guarantee their readiness for exclusive instruction in a foreign language. Per the opinions given by the teachers, though students can read and write, they find it very difficult to understand lessons when the language of instruction is exclusively English. Thus, they use their L1 to mitigate the language-related difficulty that would otherwise arise in class. This lends weight to Cummins's theoretical assumptions about instruction in a foreign language.
Further issues raised were that using L1 takes away anxiety and feelings of low self-esteem from the students. The interview and classroom observations revealed that strict adherence to the policy rather affected classroom interactions. It was observed that whenever teachers used English in class, students became passive and hardly participate in the lesson. A teacher explained that students would rather stay quiet throughout the lesson than be ridiculed by peers for an awkward expression. He explains: Students exhibit low self-confidence when they are to provide feedback on learning in English. The reason is that they fear being laughed at, so usually, they lower their voices when they are answering questions unless you order them to speak out. Students also supported this by expressing the view that they sometimes feel reluctant to speak or respond to classroom discussions because their fellow students might ridicule them for committing grammatical errors. The extract below is a response of one of the students in the focus group discussion:

I always feel shy about speaking English in the classroom. The reason is that when you make a mistake in your speech, your friends will laugh at you in class. That is why I don't want to answer questions in the classroom. But when I am speaking in Bono [the L1 of the student], I don't fear being laughed at.
The student's comment is a clear indication that using English to engage them in class has a negative impact on their classroom engagement. They feel reserved in class as compared to their L1 which makes them less anxious to speak or express their ideas in class. Teachers posited that students may have tangible ideas to put across, but their inability to put them into a language they are fluent in may lead them to keep mute on whatever thoughts they have. This is how he puts it: Sometimes a student may be finding it very difficult to use the English language to express himself or herself, but when you ask him/her to use English only the student will say that "I won't talk again". But if you allow them to use Twi, they will be able to express themselves better. Describing students' engagement in L1 medium of instruction, another teacher posited that "… the reserved students, the quiet ones, when you go down to their level, they also participate well. Those who are quiet, if you bring it to their mother tongue, you see everybody is willing to participate in". The teacher was further probed to give a possible reason why those he termed as "quiet students" would only get involved when the instruction is the local language. Responding to this, he stated that, I believe that since they all get involved when the conversation is in the local language, it can be that they are afraid to speak English. But some of them too, it is their nature.
Besides helping the students to be active in the teaching and learning process, teachers also reported that the use of Twi makes their job easier as it enables them to easily explain concepts, which would, otherwise, be difficult if they used English. As stated by one of the teachers, the core mandate of a teacher is to help students to understand every lesson. Hence, it would make sense if you choose a language that would enable you to achieve successful teaching and learning instead of adhering strictly to the implementation of the language policy. When asked about what would happen if he does not use Twi as the language of instruction at all, he explained that: It will affect you [teacher] as well as the students. The reason why you would be affected is that it is your objective to get your students to understand whatever you are teaching in order to get the expected results. Hence I believe it would be a waste of time if your students do not understand whatever has been treated by the end of the lesson. It could be inferred from his comments that, in the context of this school, it is the use of Twi as the language of instruction that helps them to reach their set objectives of every lesson. Therefore, from the teachers' standpoint, Twi as the language of instruction is an inevitable factor in the successful delivery of all lessons.
Contrary to the idea that the use of the local language is one of the main factors that cause the underachievement of students (Etsey, 2005a), it was found, from the perspective of the teachers, that the use of Twi rather helps students to understand lessons better, and, therefore, Twi cannot be considered a cause of students' poor academic performance. When asked to give other factors that may be the probable cause of the underachievement of students, Teacher 1 added that: There are various factors, one of them is that a child would come to school without 1 cedi to buy food. He would not even bother to buy books that will facilitate his learning, and even if there is an organised extra class, as we previously practised, he would not get 1 Ghana cedi or 50 Ghana pesewas to take part in. Drawing on this, it is necessary to note that the language factor, from the perspective of teachers, plays a lesser role in students' underachievement. The data from the interviews revealed that factors such as inadequate motivational support and inadequate teaching and learning resources for learning rather affect students in the school.

Students' reactions to the use of L1 as instructional language at the JHS level
Initial interviews with the teachers revealed that students showed a great deal of interest in the Twi medium of instruction to the extent that they could even alert the teacher to translate aspects of what had been said in English into Twi for them to understand. When asked whether the students show some sort of disinterest in the overuse of Twi as the language of instruction, a teacher quickly replied emphatically that: That is not the case at all. When you speak English continuously for about 10 minutes without using Twi, one of the students may tell you that "sir, you have to explain some of the points in Twi". This implies that, from the perspective of students, Twi is the preferred language of instruction. Additionally, it was revealed that lessons that aroused the interest of the students were the Asante Twi lessons. The rationale behind this assertion is that the Twi subject is the only lesson that allowed unrestricted use of L1 as the instructional language in class. Commenting on this, one teacher revealed that: They [the students] have more interest in the subject, Asante Twi. They have more interest because that is where they can express themselves. Even the less academic student [the slow learner], someone who is not performing so well in his academic work, will be able to express himself or herself in that particular lesson. She further explained that it allows a slow learner to participate effectively in class because he can express himself effectively. It could, therefore, be stated that in the case of these schools, students showed great interest in Twi as the medium of instruction.
Apart from showing great interest in Twi as the instructional language, active participation among students in classroom activities was also identified as one consequence of using Twi as the language of instruction. Hence, it was perceived by most of the respondents that the use of Twi strengthen the teacher-learner relationship. As a confirmation of the classroom observations conducted, a teacher emphasized that: I can say with all certainty that students show interest when you use Twi as the instructional language all the time. Every one of them expresses his/her thoughts effectively and asks questions. Also, Teacher 4 added that: Students prefer Twi more than English because if you speak English, somebody will stand there for about 30 minutes trying to analyse everything in their head before they proceed with whatever they want to do. But for the Twi, when you say it at once, they just proceed with it. This shows that comparatively, the use of Twi accelerates the lesson delivery as it is perceived by teachers that the students grasp concepts quicker when the teacher uses Twi as the language of instruction. Conversely, insight from the focus group discussion with students revealed discrepancies in students' responses with regards to how they react to using L1 as the language of instruction in the classroom. While some favour the Twi medium, others subscribe to the English language medium of instruction. The rationale given by students who prefer English instruction was that it is the official language of the country, and hence, it is imperative to gain proficiency in the English language at an early stage of education. Those who were in favour of the Twi medium of instruction also opined that it is essential to receive instruction in Twi because it facilitates learning. Responding to whether or not they favour the Twi medium of instruction, one of the students responded that: Yes, because we may not understand most of the things he says in English, so when he teaches in English, he later explains it in Twi to let us know the meaning of what he said in English.
Also, another reason students gave when commenting on why they favour the Twi instruction was that it is their language, therefore, it becomes interesting when they receive instruction in it. Affirming this, one of them said that "I like it because it is my language". She further added that:

I like the instruction in both languages. I like Twi and English. Because the English language is the language I write. I write English in every assignment I do. Also, every learning material I read is written in English.
From the points raised by these students, it could be said that they are in favour of Ghanaian languages as the medium of instruction. Their rationale behind this is its ability to help them comprehend the lessons. On the other hand, they also favour the English medium of instruction since they believe that in any academic work, they would be tasked to express their thoughts in it. The rationale for liking the English instruction appears to be utilitarian since they were in doubt whether or not the use of Twi could help them demonstrate knowledge of English. Additionally, a student added that she expects to be instructed in the English language only. Below is her reason for saying that: I would like it [English-only instruction] because when you grow up and get a job, you are not going to use Twi, but you will be required to speak English only. She was, therefore, asked whether she would be happy with English-only instruction even if understanding the lesson is compromised. She added that after that I would ask that I didn't understand this part. In line with the findings of Burden (2000), it could be said that for the sake of understanding and explanation of complex concepts in every lesson, students expect teachers to use their mother tongue. However, they also need the English medium to prepare them for other academic tasks such as reading and answering questions in examinations.

Discussions and conclusion
The purpose of this study was to investigate the rationale behind the retention of L1 as the language of instruction at the JHS level in Ghana and how such practice impact teaching and learning. The findings indicated that teachers violate the early-exit transition policy to ensure effective teaching and learning. It was also revealed that using the L1 had a positive impact on lesson delivery and as such, students showed much interest in L1 instruction.
Regarding the factors contributing to the retention of students' L1 as the language of instruction at the JHS level, it was noted that the most prominent factor is teachers' willingness to ensure students' understanding of the lesson. Generally, the consensus among scholars in the field of bilingual education is that L1 use in the classroom offers much to students' academic progress. In Ghana, the educational structure is such that using children's L1 at the JHS level is somewhat an illegal practice (Owu-Ewie & Edu-Buandoh, 2014). Nonetheless, it appears that most students are unable to reach the requisite threshold of English language proficiency before the transition period. Hence, per the assertions of the teachers who participated in this study, adhering strictly to the implementation of the English only policy at the JHS level is somewhat detrimental to students' learning. Thus, teachers constantly resort to the use of students' L1 as a means to foster an effective classroom interaction which, in turn, enhances students' understanding.
It could also be argued further that the retention of the L1 at the JHS level could be a direct result of the dilemmas of the implementation of the early-exit bilingual model at the basic education level. Reports on the implementation of the early-exit policy suggest that factors such as linguistic heterogeneity and inadequate teaching and learning materials impede the effective use of the L1 as the instructional language during the early stages of schooling (Ampiah, 2008;Bretuo, 2020). This, therefore, presupposes that students do not acquire requisite linguistic skills in their L1 which would, in turn, expedite their L1 development before transition. This reflects Cummins ' (1979) assertion that the competence students acquire in their L1 is a prerequisite for the development of L2. Therefore, the inability of students to engage in English-only instruction at the JHS level might probably stem from the ineffective implementation of L1 instruction in the early grades.
The findings of the study further contribute to the debate on the role of L1 use among teachers in the Ghanaian context. In most examinations, students' failures have always been ascribed to teachers' obsession with the use of L1 as the language of instruction. While most studies (Opoku-Amankwa, 2009;Mwinsheikhe, 2009;Bretuo, 2020) credit the positive role of children's L1 in fostering academic progress of students, other studies (Etsey, 2005b;Amua-Sekyi & Nti, 2015) rather attribute students' failure in academics to the frequent use of L1 as the language of instruction. The findings of current study, however, seem to suggest that students in this context would be at a great disadvantage if they received instruction in any language other than their L1. As stated by most of the teachers, if they insist on the implementation of the English-only policy at the JHS level, they will be doing so to the detriment of the students' academic performance. The foregoing lends weight to Cummins' threshold theory, which holds that bilingual students would face challenges if they fell below the requisite threshold of proficiency in both languages. As indicated by the teachers, students' proficiency in the English language is not well-grounded, and, as a result, they face difficulty in the language of instruction. Therefore, given that the English proficiency of most students in Ghana appears to be a general problem.

Recommendations
Based on the findings of this study, it is recommended that considerations be made for both languages to be used complementarily at the JHS level. The Language in Education policy should therefore be revisited if teachers are required to stick strictly to its implementation.
It is also recommended that since students advocate for English language instruction, teachers should minimize the use of students' L1 (Twi). The rationale for this recommendation is that the English language is the language of the curriculum, and students need to develop their proficiency in it to progress in all other academic activities that require language competence.
Most importantly, it is also recommended that teachers should not forcibly impose English-only instruction on students, as practised in some schools in Ghana. The reason is that it could make most of the students passive in the teaching and learning environment.
Finally, given the struggles with the implementation of the early-exit model in Ghana, it is suggested that it should be discarded for the adoption of the late-exit model to ensure that students gain mastery of the English language before transitioning into a mainstream English class.