Education Department Proposed Streaming Students Education Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

During the period from 1994 to 1996, the Education Department proposed streaming students of different abilities into three types of secondary schools: English medium schools, two-medium schools and Chinese-medium schools. The policy later was revised to separate schools into English medium ones and Chinese medium ones. Although the streaming policy has met with some resistance from the whole society, a report called the ��Review of Medium of Instruction for Secondary Schools and Secondary School Places Allocation- Consultation Document��, issued in 2005, supported the continuing use of the streaming policy (Poon, 2004). The enforcement of the dual stream policy is regarded as the major issue in the education field, as it affects all the parties concerned: namely, students, teachers, parents and schools. In EMI schools, many teachers, who used mixed code in the past, have to get used to using English as the sole medium of instruction, while in CMI schools, teachers have to prepare all teaching materials in Chinese, using Chinese textbooks and to speak in Cantonese during the whole class session. Students, on the other hand, will be under pressure to get into an English medium Secondary School, as English medium education is still regarded as superior and beneficial for their future study and career.

Though code-switching is based on a historical and linguistic background, it is still used by teachers in secondary schools. It is mainly against this background the next chapter is set. In this chapter, the author, first, introduces the basic definition of code-switching, and then outlines its brief literature and the function of code-switching play in the classrooms. Then, it analyses the code-switching situation from the political, affective and pedagogical side so as to demonstrate its relationship to bilingual education and the MOI policy. The aim is to point out both its possible advantages and any limitations, and to conclude whether that the proper use of code-switching would better facilitate language teaching, as well as other subjects learning, in Hong Kong Secondary classrooms.

4.2 Current Language Situation in Hong Kong

Indeed, numbers of findings related to teachers' classroom language use reveal a clear difference between EMI and CMI streams (Evans, 2009), and the subjects from each stream has a clear orientation toward each type of MOI. However, neither EMI nor CMI subjects appear to have total pure linguistic environment in classroom as required in the streaming policy and fine tuning policy. Teachers in both CMI and EMI schools have found it difficult to fully implement the policy in their classrooms.

According to a study of identifying teachers' perceptions using language patterns (Tsui, 2004), on the one hand, in CMI schools, when teaching different subjects such as history or mathematics, teachers would not only choose to use Cantonese during the whole lesson, rather, they will introduce many English key words or terms after explanation in Cantonese. On the other hand, interestingly, in EMI schools, although teachers claim to be confined in certain MOI policy, not only do they all feel pressure to strictly obey the rules but also admit that they would use Cantonese for illustration during the lesson. For instance, as one English teacher said "when it is something as complicated as concepts like logic, if I use English they can not understand, so then I have to use Cantonese". For another, one math teacher states that usually when students' English standard is not good enough or they can not hear clearly or they may not understand what he says, he will switch to Cantonese. This clearly show that though teachers may be divided into different categories and some of them may claim to be the strongest advocates of exclusive use of L1, most of them tend to inevitably use code-switching more or less during teaching.

From the time the streaming policy has been implemented, teachers in EMI schools are required to make changes to their classroom language practices. Teachers once from EMI schools are compelled to shift to CMI that the teaching materials should be switched from English to Chinese for written communication whereas CMI teachers should get rid of mixture teaching with Cantonese and English, a pure Cantonese teaching is firmly adopted for spoken communication. Accordingly, the students in CMI schools should continue the practice of reading and writing in Chinese, while speaking and listening in Cantonese. The teachers claim that after strict streaming, CMI schools use Cantonese as their sole medium of instruction and interactions in Form 1-3. However, a study shows that in numbers of classes, academic-related functions are performed either equally in English and Cantonese or mainly in English (Evans, 2009). What is more, findings (Evans, 2009) concerning CMI schools have indicated that although Chinese is the required language of teaching materials and written work, a third of the subjects reports that the medium of reading and writing is either English (19%) or a mixture of Chinese and English (11%). Therefore, the results have shown that English continues to be used as a written medium in some CMI schools regardless of the streaming policy. More than that, an interviewee from CMI school say that some teachers, particularly those of science subjects, would like to use English textbooks at secondary level in order to prepare their students for later examination-oriented teaching in English in higher grades (Evans, 2009). It is the adverse effect of streaming policy that maybe some CMI teachers have already been accustomed to using such materials before their school were obliged to switch to Chinese and they have not fully adapted to the streaming policy. More importantly, this further indicates that English is still used in CMI classrooms despite the clear-cut division.

In addition, the streaming policy only limits in Form 1-3, after that, CMI schools are permitted to switch to EMI in Forms 4-7 if their students elect to take the English version of the HKCEE and the HKALE (Hong Kong Advanced Level Examination) (Evans, 2009). Many CMI students would all take this option since if they want to enter Hong Kong's English-medium universities or professional employment in business area, the proficiency in English is an indispensable and essential factor. However, this could cause a problem is that when students from both CMI and EMI students come into one class in higher grades, teachers are facing difficulties to cater both needs as some CMI students�� ability to learn English is limited, all their choice is made by their parents rather than themselves. Under such circumstances, teachers begin to use Cantonese or a mixture of Cantonese and English to explain English materials.

In terms of reading and writing in the reformed EMI schools, the majority of the subjects report that English is the sole medium of teaching materials and written work in Forms 1-3 (88%). Forms 4-5 (92%) and Forms 6-7 (87%) (Evans, 2009). Unlike their counterparts in the CMI stream, EMI teachers could keep using English textbooks and handouts. However, concerning the language use in EMI schools, some surprising observations have occurred that Chinese or mixed-coding teaching practices are still being used widespread.

Indeed English is the usual medium of whole-class instruction in EMI schools. However, EMI teachers still use more or less Cantonese rather than English when they talk to students individually in non-academic matters in the class. This is considered important because in Hong Kong, the transmissional model of teaching is usually adopted, and the teacher is essentially controlling the patterns and language of classroom interactions as well as "determining the nature of classroom discourse" (Sridhar and Sridhar, 1980, p.412). Therefore, using Cantonese in EMI classrooms decide the extent to which the MOI policy is being implemented. By quoting the Secretary for Education, Mrs Law's comment (2005, p.5), ��teachers in EMI schools use Chinese to teach English textbooks. EMI should mean full immersion into English, but students are immersed in Chinese".

As we have seen from the above discussion, no matter the streaming policy or the new fine-tuning policy, they are to eliminate the use of Cantonese in EMI classrooms and English in CMI classrooms. However, both languages are still used in each context and the actual classroom practice in both streams has not changed dramatically when compared with the time before 1998.

Therefore, since the mixed-coding teaching is still popular in two types of schools, in the following parts, it will be specifically explained, analysed and discussed to testify whether it is a new but practical way for Hong Kong future language education.

4.3 The Definition of code-switching

When two languages exist in a community, the bilinguals naturally tend to switch from one language from another. In Hong Kong classrooms, it is common occurrence for teachers and students to switch from English to Cantonese. In general, the switches can be subcategorized into inter-sentential and intra-sentential ones. To be more specific, inter-sentential is called code-switching, which refers to language switch across sentences, or the alternation of two languages within a single discourse, sentence, or constituent. Intra-sentential is termed as code-mixing, which means switch within sentences (Luke, 1984; Kamwangamala, 1992; Sridhar, 1995). The distinction between these two terms is problematic, since some scholars (e.g. Clyne, 1991) prefer to use ��code switching�� as a cover term for ��the alternate use of two languages, either within a sentence or between sentences�� (Clyne, 1991, p.161) while others (e.g. Bhatia, 1992) tend to use �� mixing�� to designate both inter-sentential and intra-sentential switches. To reconcile the terminological problems, in this chapter, the code-switching phenomena will be regarded as a "continuum" of the above concepts. In addition, since code-switching and code-mixing practices refer to the complementary switch of L1 and target language, as the instructional language in the English classroom, in this chapter, �� mixed-code teaching�� is readjusted to mean the alternation of Cantonese and English in CMI and EMI classrooms of Hong Kong Secondary Schools, and includes both code switching and code mixing practices.

4.4 Controversial Issues Concerning the Definition

According to Chan's research (2007), in terms of individual lessons in different schools, the word-level and phrase-level switch are much more used by teachers than the sentence-level switch. That is to say, the mixing within sentences is more frequently used than between sentences. Although the inter-sentential and intra-sentential mixing all belong to code-switching, from the perspective of pedagogy, intra-sentential exchange is more beneficial for students' language as well as subject learning. Firstly, the students from EMI schools are more in favour of teachers talking in complete English or Cantonese chunks. Just as one student said, "chunks of English render full understanding of grammar, and vocabulary. ��If you get a bit of English and a bit of Cantonese, you cannot learn much" (Chan, 2007, p.193). What the students said, also, was in line with the teachers' thought, which is that though mixed sentences may provide easy comprehension for students, and convenient practice for teachers, it cannot provide a better answer for English language learning. Jacobson (1990) also demonstrates in his ��New Concurrent Approach�� that there should be less intra-sentential code-switching in the classroom, since it is not a good model for bilingual education. Therefore, teachers cannot choose the most convenient way of code-switching for their own sake, nor sacrifice English language acquisition to promote subject learning, instead, they should consider code-switching from an all-round perspective, and really cater for students' needs, that is by trying to use much more sentence-level switching rather than the other one. By keeping the two languages apart at the inter-sentential level could be of greater help, as it would expose students to total English sentence structure, rather than mixing them in a single sentence. This kind of mode, in turn, would enhance students' listening and speaking skills, which could be used more when the students reach a relatively proficient level.

4.5 A Brief Consideration of General Issues about Code Switching

There are four Grammatical Constraints and Theories concerning code-switching��, but before focusing on the issue of code-switching in Hong Kong classrooms, this section will analyse the general linguistic aspect of grammatical constraints in code-switching. First, it considers ��The Equivalence Constraint��. This is a term defined by Poplack (1980, p.586) to mean that when code-switches take place in a discourse of which the L1 and L2 elements conform to both languages' syntactic rules. However, Poplack's theory has raised serious doubts, since in different language pairs, it is unrealistic to satisfy two languages' syntactic rules simultaneously. Secondly, the Free Morpheme Constraint, which means that a switch may not occur between a bound morpheme and a lexical item unless the latter has been phonologically integrated into the language of the bound morpheme.' (Sankoff and Poplack, 1981, p.5). However, the constraint is also severely challenged by many counter-examples in Swahili English (Scotton, 1983) and Cantonese-English (Leung, 1987, p.91). Thirdly, the Dual Structure Principle, which according to Sridhar and Sridhar (1980, p.412), is where the 'guest constituent' is different from the 'host constituent'. The constraint is reflected in the internal structure which is where the guest constituent need not comply with the rules of the host language, as long as the host language obeys the rule of the host language. Nevertheless, this principle also has counter examples in the data of Lingala-French code-switch (Lingala'French; Kamwangamalu, 1989, quoted in Myers-Scotton 1993, p.154). Lastly, the Matrix Code Principle suggested by Kamwangamalu (1989, p.157) which means in code-switching in which L1 is identified as the Matrix Code (i.e. host code) and L2 as the Embedded Code (i.e. guest code), the grammar of L2 must obey the morpho-syntactic structure rules of L1. Nevertheless, it also receives criticism in that more often it is hard to make out differences between the roles played by the two languages (Jacobson, 2001). In addition, the principle may be betrayed when sometimes the embedded language such as English play the dominant role instead of following the rule of the host language such as Cantonese.

Beside all the main four grammatical constraints above, there are also theories like Muskysken and Sigh's Government Constraints (1986), Satorini and Mahootian's Null Theory (1995), Macswan's Minimalist Approach (1999). Nevertheless, all these are more or less criticized by its counterparts. In addition, they only consider describing how code-switching occurs within, but are rarely concerned with how code-switching happens in the classroom and what would it be its possible functions.

Therefore, in the next part, I will adopt a more realistic view by introducing the main scholars' arguments about the useful effects and functions that code-switching can bring to the classroom.

Johnson & Lee (1987, p.107) found that in bilingual classrooms, the general pattern of presentation is often sequenced in topic statements in English followed by explanation and illustration in Cantonese and the conclusion in English again. Merritt et al. (1992) also mentioned that the aim of changing codes is often to clarify or reinforce lesson materials. Switches from English to first language across sentences are often used in summaries or to emphasize the vital points in the teaching materials. More than this, code-switching is also supported by a number of researchers. For instance, Ho and Van (1986) identify a range of functions which include helping students to understand, motivating students to learn, reprimanding students and saving time. Such findings regarding the function of Cantonese in English classes were later echoed by Lin (1990).

Lin (1990) states that mixed code has the effect of ��acculturation, efficient classroom management, establishing a closer relationship with students, establishing bilingual academic knowledge and pragmatic response (of teachers and students) to the problems created by the imposition of a foreign language as the medium of instruction, despite their having a common native language �� (Lin, 1996, p.74). Code-mixing is therefore beneficial to language learning and acts as a communicative resource for the teachers in the classroom.

The uses of mixed code are further identified in other research.

In the classroom, the change of codes is used by teachers and students in their negotiation of different role-relationships. Lin (1988) proposed a Model of Pedagogical and Para-pedagogical Levels of Interaction in the classroom and explained that code-mixing behaviour is used with regards to the relationships between different levels of interaction negotiated by the teacher. That is to say, if the teacher wants to maintain his authority in teaching, he would use English. If he would like to build up a friendly rapport, Cantonese is a better choice. Later, he also found, in a mini-study (Lin, 1998), that the lower achievers benefit more from mixed-code teaching in English classes, since code-switching would greatly enhance their motivation to learn. His further studies on the use of mixed code identify that ��mixed-code teaching has its merits, if it is done in a systematic way�� (Cheng, 1998, p.7).

This is also believed by Lin (2000). He illustrates that L1 can be used strategically to arouse students�� interest and build an harmonious relationship between them and the teacher. Besides that, it will help students catch the essence of academic knowledge and pedagogically increase class participation.

In 1996, Lin had done further research and recorded 8 excerpts using code-mixing, in one of the Hong Kong secondary schools, to explore the benefits of using code-mixing. To be more specific, please not the following parts. (Please see appendix 1)

Excerpt 1

Because the students have omitted to finish the assignment, the teacher switches from the English teaching to Chinese questioning. Actually he is emphasizing his concerns and intends to arouse students' attention to the fact that not to complete their assignments is an unacceptable behaviour. Under this circumstance in Chinese culture and norms, it is appropriate for the teacher to admonish the students who are lazy, and that students cannot use excuses such as "I forgot the homework". This inside Chinese logic is shared by all Chinese teachers and students.

Excerpt 2

The teacher is angry at students' lateness for class, thus, he uses Cantonese when asking them to settle down as soon as possible. In this situation, when the teacher tries to pass on some urgent messages, he will switch to Cantonese.

Excerpt 3

In this situation, the teacher says something which is not in the textbook and intends to establish a warm and friendly atmosphere in Cantonese. Then, she switches back to English to facilitate a pedagogic task in textbook again.

Excerpt 4&5

In the chemistry class, the teacher introduces the chemical terms in Cantonese, because the origins of the vocabularies are in modern Chinese scientific literature, whereas those in English are the translations. By this means, Gagni (1993) demonstrates that by knowing both English and Chinese chemical terms, students could not just understand them, but also form more comprehensible scientific conceptual connections.

Excerpt 6

When giving geography lessons, the teacher uses Cantonese to explain the term "climate" and make students notice the differences between climate and weather. In this situation, the teacher uses annotations to explain the English academic knowledge being taught, with the students' native language being used as well to facilitate their understanding.

Excerpt 7

When introducing a new topic in L2, the teacher uses the mother tongue to extend the topic so as to arouse students' existing knowledge and life experiences. This L2-L1 discourse format is more familiar and less alienating for students when they are attempting to grasp knowledge and it serves two functions: firstly, the linguistic brokerage effect to ensure students' comprehension of L2, and secondly to bridge the gap between the unfamiliar L2 and more relevant L1 life experiences.

Excerpt 8

This excerpt could be concluded into this E(L2-L1)-R(L1)-F(L1-L2) discourse format which show the teachers' attempt to elicit and judge the contributions, in their mother tongue, from students in the process of knowledge attainment. As a result, the incorporated L1 knowledge has been transformed into L2 corpus. The effect here is the teacher's pragmatic response to the paradox presented in valuing their first language as well as fulfilling the requirement of teaching second language.

All these excerpts have gone some way to prove that bilingual classroom practices can reveal teachers' highly creative use of communicative resources (Lin, 1996, p.74). The strategies involving in the classroom teaching, through effective discourse formats with teachers as academic brokers, would render English as the suitable medium of instruction and education, and is more accessible and interesting through it using students' familiar schemata.

Except for all the advantages that teachers bring about in mixed-code teaching, two issues arise: one is that in excerpt 4, after the teacher explains the issue in Cantonese, he switches back to English immediately and says the Chinese knowledge is beneficial but not required in the exam. This phenomenon could be treated as" English-dominant academic bilingualism" (Lin, 1996. p.69). Besides that, another investigation revealed that when teacher introduces English and Chinese terms side by side, they will do it in English-Chinese sequence and pay more attention to English since the examination will be English-based. This sequence also shows that Chinese suffers, if not inferior but subsidiary status in China.

Due to the two questions above, a long discussed issue has been raised again, which is the debate on whether the exclusive use of the target language or added with first language in second language classroom teaching is more beneficial for students.

4.6 In Favour of Target Language Learning

It appears to be the situation that in second language learning, the less L1 is used, the more beneficial it is for the students. According to Franklin (1990, p.20), it is generally agreed by theoreticians that it is crucial that the instructional language in classroom should be the target language, regardless of the teaching methodology. The theoretical framework which in favour of the first language, is rooted in Krashen's Input Hypothesis and the later Natural Approach, advocated by him and his colleagues, Burt and Dulay (1975). They argue that language is subconsciously acquired rather than consciously learnt, and this kind of acquisition only occurs when the students' input is comprehensible. As well as this, some scholars believe that input to learners could be made more comprehensible through the use of a simplified target language, contextualized cues, and abundant visuals (MacDonald, 1993).

Since the 1960s and 1970s period, there have been a number of studies that discussed the correlations between teacher use of target language and the learners�� proficiency and achievement. In the 1980s, with the widespread use of the Natural Approach (Krashen & Terrell, 1983) and the Immersion Approach, exclusive target language use in second language classrooms was widely favoured. To be more exact, supporters of the Immersion Approach advocate their assertions with many empirical findings. These findings suggest that learners in these programmes are far more proficient in the target language than their counterparts in the traditional language programmes (for example, Genesee, Holobow, Lambert & Chartrand, 1989; Heitzman, 1994).

In the 1990s, more works appeared that supported the exclusive use of target language in foreign language classrooms. These included the books of F. Chambers (1991), Duff and Polio (1990) and MacDonald (1993). For instance, Chambers (1991) argues that the teacher is the only source of spoken target language that acts as the paralinguistic support to students, therefore ��teacher talk is crucial in the development of students�� listening skills�� (ibit, 1991, p.28) and ��the natural use of the target language for virtually all communication, is a sure sign of a good language course�� (ibit, 1991, p.27). In addition, Turnbull (1999, 2001) advocates a near exclusivist position. To conclude, all of those authors quoted above believe that teachers should maximize the use of the TL as much as possible, especially in foreign language contexts where students have little contact with the TL outside the classroom (Atkinson, 1987; F. Chambers, 1991; Duff & Polio, 1990; MacDonald, 1993; Tang, 2000; Turnbull, 1999, 2001; Wang, 2003 in Chinese; Wang & Wen, 2002).

4.7 In Favour of First Language Learning

However, some researchers, including those in Hong Kong, Mainland China, and non-English-speaking countries (for example, Ho, 1985; Lin, 2000; Ustunel & Seedhouse, 2005) have emphasized the practical use of L1 in L2 classrooms. Atkinson (1993, p.3) argue that ��for most teachers in most circumstances, teaching entirely in the target language is simply too tall an order�� and 100% use of the target language in an English classroom is not necessarily desirable or feasible in most classrooms. A similar point is identified by Phillipson (1992) who illustrates that if the mother tongue is forbidden in the classroom, students may be deprived of their cultural identity and, therefore, this action could be treated as cultural imperialism. In addition, Anton and Dicamilla (1999, p.237) further confirm that the use of L1 is beneficial to the learners' needs to build up the L2. Without the use of L1, for example, teachers who adopt the task-based approach are often frustrated when they have to implement in-class activities. It can prove difficult for the students to learn and understand technical terms, such as chemistry terminology, without the help of L1 explanations. Cook (2001) has expressed reservations about the exclusive use of L2 in second and foreign language classrooms. He questions the limitation of Krashen��s Input Hypothesis with first language acquisition. He also suggests that L1 may play ��an integral role in L2 learning as well as in L2 use�� (Cook, 2001, p.408).

Indeed, the argument about whether to teach L2 language by total immersion or with proper L1 illustration has long being discussed by theoreticians as well as practitioners. From the above arguments, it is still hard to judge or prove which side is more advantageous or beneficial for language learning. But what is certain is that to deal with different language situations, we need a distinct way to interpret each specific one. Therefore, in the following parts of this study, the author will discuss the code-switching in Hong Kong from the political angle, affective side and pedagogical side, so as to explore the issue deeper therefore to prove that the benefits and feasibility in Hong Kong education, especially in secondary schools.

4.8 The Political View of Code-switching

The case for not using L1 has for a long time had support from both the government and children��s parents, because they think the use of L1 might "jeopardize their children's future" (Tsui, 2004, p.99). The Hong Kong Government (1982), Education Commission Reports No. 4 (1990) and No. 6 (1996) all opposed the code-switching phenomenon, claiming that mixed-code teaching in classrooms will not only block second language learning but also retard students' development of the mother tongue. In the Education Commission Report No.4 (1990), it recognized that teaching and learning are generally more effective if the medium of instruction is either the mother tongue or English. It followed that they believe that the use of the mixed-code should be reduced as far as possible. The government regards code-mixing as the major, if not the sole cause of learning difficulties by students in schools.

In addition to this, the Education Department of Hong Kong (1997) issued the "Medium of Instruction Guidance for Secondary Schools", which blamed mixed-coding as a leading factor in the decline of students' language proficiency and prohibition of the use of English in schools. They stated that any teacher in schools would be severely punished if found using this method. To avoid this situation, the Education Department even sends inspectors to schools from time to time. With regards to this situation, the Education Journal (Swain, 1986, p.6) and Poon (1998, p.96) argued that the problems of mixed-code teaching are serious, and they doubted its effectiveness. They maintained that, by continuing to teach, it could result in the students not being able to speak either good English or Chinese. Thus, the conclusion is drawn that the mixed-code is harmful for students in both languages and the problems should be rectified.

However, this judgement is drawn on an often taken-for-granted, language ideology of linguistic purism, and is used to condemn the bilingual classroom practices. The famous Llewelyn Report in 1982 suggested that the government could acknowledge the widespread use of bilingual oral practices in the classrooms of English medium schools, and that it should try to improve the effectiveness of these oral practices by developing genuinely bilingual curriculum strategies. In responding to the Llewelyn Report��s recommendation, Johnson referred to the practices of using both L1 and L2 in the Anglo-Chinese secondary school classrooms as ��various language modes of presentation�� (Johnson, Chan, Lee, & Ho, 1985).

Furthermore, according to Flowerdew (1998), the mixed-code teaching is a discursive and prevailing practice, which can be found in bilingual classrooms in many domains throughout the world. Therefore, the mixed-code does not exclusively belong to Hong Kong. Concerning the historical and current social situation in Hong Kong, it is hard to totally eliminate mixed-coding in the foreseeable future. Besides that, there is little data or empirical evidence to prove that code-switching might give rise to the decline in proficiency in L2 learning. In contrast, research has suggested that code-mixing is a rule-governed and systematic method which can demonstrate the grammatical rules and reflect the operation of syntactic regularities (Toribio, 2001, p.404).

Therefore, what should be done is for those involved with language teaching is to come up with strategies on how the L1 can be used constructively in classrooms, rather than using L2, as a substitute for it. A strict policy change would only restrict teachers' abilities, rather than offer them more space to develop their own methodology. Many experienced teachers in both CMI and EMI schools hold neutral or even negative attitudes toward the changing policy, suggesting that what they would like is to adopt various teaching strategies according to different situations. According to Lam (1999, p.43), a study in both EMI and CMI schools reveal that the majority of teachers' use of instructional language remains unchanged even after the implementation of the streaming policy. This indicates that from the perspective of the policy makers, the streaming policy is there to encourage language purity and eliminate the mixed-code use in the classrooms. However, these views have had little effect on teaching methods, as teachers tend to use their own way of teaching, which they consider it could best benefit their students.

4.9 Attitudes towards Mixed-code Teaching

4.9.1 Arguments on Students' Confidence Building and Habitual Effect

Some scholars claim that in order to build up students' confidence, mixed-code teaching is used for lower achievers, as the students with higher language capabilities do not necessarily need this type of help, as in most cases they can handle learning difficulties. The students insist that too much reliance is placed on teachers' explanations in L1, and that it will lead to less time for listening to English and hinder better acquisition both in L2 and the related subjects. Just as earlier studies pointed out, teacher talk has a great influence on learners�� acquisition of a second language, since not only it is vital for the development of students�� listening skills, but also "a prime model of true communicative use of the language" (Cook, 2001). More than this, too much Cantonese in the classroom would build a habitual effect that every time students are facing language barriers, they intuitively hope for the mother tongue explanation, even if the teacher has already paraphrased the terms in English, the students may still feel insecure and suspicious about the knowledge. This conclusion was also drawn, earlier, by Atkinson in1987, who argued that one of the disadvantages of using the mother tongue in second language teaching is that students may feel they have not really understand an L2 word unless it is translated into L1.

However, for teachers, their teaching should not only be focused on high level students. Satisfactory learning progress in class would not be made if only the students with higher scores understood the new lesson, then the teacher could just step into a new part, ignoring those students who have been unable to understand, fully, the new work. A responsible teacher, according to Edstrom and Leung (2004, 2010) does not only have to adhere to the pace of more competent students but also needs to make adjustments for less competent students to catch up. What is more, based on teachers' perception, by using code-switching, they could boost students' self-confidence, especially among relatively low proficient students. As secondary language students still belong to the elementary level or are beginners, therefore they may become more easily frustrated or anxious during the learning process. More importantly, tasked-based teaching is a common approach used in Hong Kong secondary language classrooms which require learners to complete the tasks in order to achieve certain goals set for the use of the target language (Willis and Willis, 2001). During this process, the students may be under great pressure if the language barrier is too hard for them to deal with. However, with the help of code-switching, the situation could be eased, which as one teacher demonstrates that in her class, if the students could cognitively manage the tasks aided by the teacher��s explanation in Cantonese, without being discouraged by the language obstacles, then their learning efficiency could be greatly enhanced and they would have more confidence in performing harder tasks (Leung, 2008, p.42). Moreover, the use of code-switching only occurs at certain times and situations such as when the students cannot understand an English term. Instead, it should be used when the teacher finds that an explanation in English would make students more confused or frustrated. However, the danger exists that students could come to rely too much on having explanations in the mother tongue, but explanations in the mother tongue is favoured by teachers as it tends to speed up the lesson. Therefore, we could conclude that, from the effective side, using L1 properly could have positive advantages for both students and teachers.

4.9.2 Arguments on Students' Using "Survival Skills"

Some scholars doubt whether the relaxed atmosphere that L1 creates is really powerful and efficient enough to facilitate the language learning. The English teaching mode may not be as stressful as some people claim to be; in contrast, proper use of target language could also boost the confidence of participants in learning English. According to Shum's study (2008), more than 60% secondary students' response, in focused interviews, is that they did not find that there is too much pressure in L2 teaching, and that with the proper teaching techniques, such as a "riddle-solving" game, where the challenges could be transformed into motivation, can give rise to "a sense of competence" among students. Even if the students feel a certain degree of anxiety, the language barrier in the L2 classroom could be alleviated through the help of "survival skills��, these are the skills of negotiating, interpreting and expressing, which Chinese are needed in the process of learning (Crawford, 2004). In another focused interview by Shum (2008), some of the participants used different survival skills to overcome learning problems. For instance, they could ask help from their peers or teachers in after lessons. By using these means, the anxiety or low motivation problem among students could be, to a great extent, successfully resolved. Therefore, the goal of learning another language is achieved since the use of the English-only mode is not just for input but also for developing other complementary skills (e.g. survival skills) needed for acquiring a new language.

Nevertheless, not all students could still use their "survival skills" when facing challenges or learning difficulties. Some of the less able students have already experienced numerous failures in their study life, what they need, rather than more stress, is encouragement from others, recognition and confirmation of any successful achievement. This will help them to improve and continue to do their best. According to Berry (2008), the teachers' constructive feedback on students' performance is indispensable in quality learning. What is more, Leung's study (2010) has pointed out that using L1 can achieve an efficient and powerful effect, so as to make teachers' feedback explicit and comprehensible. Therefore, using L1 with constructive and comprehensible feedback would help students to build up confidence and motivation in learning.

5.0 Pedagogical Effect of Mixed-code Teaching

5.0.1 Arguments about the Exposure Time on L2

Concerning code-switching in the classroom, various teachers have different opinions about its use. In general, except for the optimal L1/L2 booster in the use of code-switching, another category is the maximal L2, as advocates believe L2 learners should be exposed to English as much as possible, since the more they use L2, the more reliant the students become on it. This has been identified by a group of both CMI and EMI teachers in a study asking for their attitudes toward L1 and L2 use in classrooms (Leung, 2010). In addition, within the limited class time, the use of L1 will reduce the time available for L2 teaching, from what it is at present.

However, the above mentioned exposure time could be used for the enhancement of English proficiency, on condition that students have already gained some basic English language, otherwise, the exposure time will only cause more problems. For instance, from the teachers' perception as shown in Leung's study (2000), one female teacher, Carolyn, states that for weaker students who are not so motivated, wide exposure is, not as important, relatively as knowing the essentials of the language. The reason lies in that if L1 has not been used, little learning of L2 would occur, and some students may even develop feelings of resentment towards their study and therefore hinder their further L2 cognitive development. This finding is also in line with Krashen's affective filter theory that teachers should attempt to reduce students�� possible negative feelings toward L2 study, so as to reach maximum efficiency in language acquisition. Moreover, due to the real teaching situation in the Hong Kong classroom, they have rather tight teaching schedule in limited class time. In order to finish the regulated amount of teaching, teachers have to use code-switching to speed up their teaching and make their explanations more brief and clearer. Last but not least, using L1 could mediate the power relationship between the teacher and the students. This is also identified in one English class (Leung, 2010, p.20) where students tended to react more readily with teachers who can speak their L1 in a positive and friendly manner later.

5.0.2 Arguments for Alternations in L2 Learning

Pedagogically speaking, there exists many alternative ways for students to grasp abstract ideas without using the L1 to facilitate understanding. For instance, "abundant visuals" and "contextualized cues" in L2 could be offered in powerpoint presentations when teaching grammar (Shum, 2007). For another, complicated vocabulary teaching could be aided by synonyms or antonyms (McCarthy, 1990).

However, the large number of grammatical rules or vocabularies cannot simply be clearly illustrated by the above means. To be more specific, some technical words cannot be explained in a superficial way, by using L2, since students need to understand, more often than not, the background knowledge of a word, but due to class time limits and students' L2 proficiency, the L2 explanation is not only inefficient but time-wasting, so, using L1 could be a much better option. For instance, according to Tsui's research (2004, p.45), in a history class, when the teacher mentions a place, somewhere in Middle East ,that no longer appears on the present day map, for students with a weak concept of this place, even if they know the English name for, it can only be a "noun" for them. Therefore, in order to dig out the underlying meaning of the word, the teacher needs to selectively and briefly introduce the background knowledge with the help of L1 about the geographic locations of the place therefore to help their literacy as well as to activate their existing schemata in L1 (Krashen, 1992, p.41). By introducing the background knowledge in L1, the teacher could then switch to L2 to further explain the word and with the information told by mother tongue, the students are better prepared to catch deeper meaning about the word and the following lesson. Therefore, based on the above discussion, we could see that L2 teaching does not often work effectively even with certain help of techniques. In contrast, using L1 for explanation is time saving and could not only fill the gap between L1 and the L2 but also better facilitate students' deeper understanding about the knowledge in the subjects.

5.0.3 Arguments about L1 Facilitating Four Skills' Learning

In CMI and EMI schools, teachers use less English to teach reading comprehension, grammar and composition but adopting code-mixing practices because it is more effective. However, people against this idea believe that less English is used in the above aspects mainly because most Hong Kong secondary schools adopt the thematic approach which means learning the four skills is linked to the same theme in the same teaching unit. As vocabulary and reading comprehension always proceeds listening and speaking, students, naturally, pay attention to the code-switching in the first two parts and support its usefulness in language teaching, but neglect the other two important aspects which are mainly taught in English (Chan, 2007). In fact, statistics reveal that 98% of EMI school teachers and more than 50% teachers in CMI schools use only English when teaching listening and speaking (Lam, 1999, p.36). Thus, it can difficult to simply conclude that code-switching facilitates all language learning in the classroom because in listening and speaking part, English is exclusively used.

In addition to this, although evidence shows that mother tongue translation may have the advantages of brevity in explanation, there are no statements to support that mother tongue teaching could facilitate long-term memory of knowledge. On the contrary, the beneficial effect of short-term memory in vocabulary lessons only lasts for a matter of hours and this short-term memory, by the using mother tongue, helps only immediately (So, 1988, p.35). Thus, it seems that supporters argue that the advantages of code-mixing by teachers in classroom may tend to be more of an exaggeration rather than a fact.

However, the so-called short-term memory of using L1 is not comprehensive at all for the study only consists of several classes with relatively lower English level students in one single Chinese-medium of school. Therefore, the short memory judgement is lack of reliability to represent a common phenomenon of the classrooms using code-switching. Admittedly, too much use of code-switching is not beneficial for students' learning. What teachers should consider is how much and in what subject should they use code-switching, rather than hold a totally negative attitude. There is much empirical evidence in the following could prove that the proper use of code-switching facilitates language learning in vocabulary, reading, writing and grammar.

As for vocabulary teaching recorded in Leung' study (2008), both teachers suggest that students understood and memorized vocabulary more easily if some Cantonese was used. In fact, L1 translation could be regarded as a useful students�� vocabulary learning strategy, since it could help learners acquire information about a new word by building a bridge between the second language and the L1 life experiences. This was also identified by Schmitt (2000), who believes that an L1 translation by the teacher is a ��discovery technique��, that connects the word or its synonyms in L1 with the target language, and thus forms a kind of mnemonic which strengthens the consolidation of a word after it is introduced by the teacher.

In terms of reading, both CMI and EMI teachers, in Leung' research (2008), agree that code-mixing could help students to learn better, as one teacher comments that since non-fictional texts can be unfamiliar to students as compared with other types of reading, such as novel or stories. Therefore, she tends to use more Cantonese when doing explanation to convey more accurate information about the non-fictional passage.

As to grammatical rules, they are often more abstract and many of them do not exist in the Chinese language, therefore despite the fact that teacher may spend more time in explaining a single grammar rule, the use of pure English explanation may still confuse and is incomprehensible to students. According to Leung (2010), the feedbacks from teachers point out that their students would not understand if the explanations are in pure English due to their low motivation and L2 proficiency. One of the teachers in Leung's interview (2008) cites an example that according to her experience, "gerund" teaching is much more efficient and time-saving with the help of Cantonese explanation.

Since the western writing style and pattern are different from their mother tongue��s writing, the use of writing skills such as paragraphing and organization could be more easily to be mastered through certain Cantonese explanations. This opinion has also been supported by the teachers in Leung's study (2008). Furthermore, during the stage of essay planning in the writing class, using English may not be the best way to give students the opportunity to express their creative ideas. Instead, the use of Cantonese may be allowed for stimulating students�� thinking and for brainstorming ideas for the writing topic. Therefore, it could be concluded that code-switching does have pedagogical advantages to facilitate students' understanding in vocabulary, reading, writing and grammar.

5.1 Conclusion

��The linguistic purist��s argument, at best, misrepresents and, at worst, totally ignores the sociolinguistic reality of Hong Kong society�� (Lin, 1996, p.51). In line with Li and Tse��s (2002) research, it is unrealistic and impractical for Hong Kong citizens to stay in a ��pure�� code of Cantonese or English, given the wide- spread use of the mixed code in everyday communication. Indeed, the mixed code has become an important linguistic marker of Hong Kong identity. By discussing the literature concerning code-switching development, its definition, functions, and debates about pro-L1 camp and pro-target language camp, the advantages of code-switching politically, affectively and pedagogically have been considered. This is not suggesting that L1 use in L2 learning is a panacea. The use of Cantonese should be used properly as it depends on the English proficiency level of the students and concerned with the lesson objectives and content. More Cantonese could be used if the lesson objectives are about Chinese literature, history, complicated terminological words or a collaborative task to encourage students' participation. What is more, the proper, effective use of code-switching to a large extent is decided by teachers' proficiency. Therefore, better teacher training is needed in future so as to enhance the teachers' qualifications with the aim of raising the quality of teaching.

There is no single way to generalize or predict how individual teachers or students view the code-switching, as their choices are related to multiple factors. Indeed, L2 acquisition has long emphasized the value of L2 use in the classroom. However, it is not true that the government attributes the decreasing standards of students' English to code-switching. Appropriate use of L1 can play a supportive role in L2 learning in such aspects as facilitating students' understanding, managing discipline, attracting attention, catering for students' diversity, building a harmonious rapport and saving time for achieving learning goals. These advantages and functions have been confirmed by the empirical and theoretical evidence shown in the above discussion. All these could lead to greater participation in L2 learning, and hence result in great advantages for students�� learning and potential achievements.

Chapter 5 Conclusion

In terms of Hong Kong��s MOI, it has been more than a pedagogical issue, rather, economy and politics have also played an important role. It could be concluded that CMI is mainly considered for educational value while EMI is for the high market value. Therefore, as long as these three main factors continue to exist, the debate on whether to adopt CMI or EMI will not cease. From the author's perspective, the future MOI in Hong Kong does not have to be clear cut. In the current situation, code-switching, if effectively used, could lead to a unique bilingual education in Hong Kong, which seems to be the best option for future Hong Kong language policy.

As for Putonghua teaching, by taking note of the political reasons and pedagogical considerations, a variety of perceptions of teachers and students have been examined including both reservations and support. Admittedly, nowadays, Putonghua has already become one of the core subjects in most of the secondary schools. While the students have shown increasing interest and motivation to learn Putonghua, and the teachers, though under pressure, endeavour to achieve better teaching results. While from the schools�� perspective, the Chinese curriculum is still in need of improvement in order to make it more effective. In spite of these efforts and the potential improvement of the Chinese curriculum, it is unlikely that in the near future, Putonghua will become the main MOI in Hong Kong and as Putonghua continues to be taught as the third language, it will be hard to replace Cantonese or English in Hong Kong people's eyes.

As the above mentioned, three MOIs, all have obvious disadvantages and contradictions, code-switching practice has come to be the main subject for discussion. In fact, it is a part of Hong Kong culture to code-mix, given the psycho-sociolinguistic and pragmatic environment. In daily interactions, people regard using it as a habitual behaviour and, more importantly, in classrooms, both students and teachers have continued this kind of practice. This indicates that mixed-code teaching is perceived as desirable and inevitable in English language classrooms.

Indeed, when code-switching is used randomly or excessively, it can reduce the students' exposure time for listening to English, or to correct English in usage and pronunciation which would hinder the natural language acquisition process.

However, to embrace only the negative side of code-switching in an English-only or Chinese-only policy is unrealistic and impractical. Though it is claimed that the best way to acquire a second language is to be completely immersed in it, teachers should consider using the students' abilities to code-switch to their advantage, especially for students at the beginner level in Hong Kong��s secondary schools. Rather than carrying on with the myth of teaching purely in L2, teachers should use L1 in a way that teachers and students both feel comfortable and confident.

Therefore, what I am suggesting is that instead of setting strict regulations, there is an urgent need for the government to reconsider and re-evaluate the single MOI policy. In order to raise the students' English proficiency level, there is the need for them to understand the lesson��s objectives and content, this is where the teachers play a crucial role in deciding the amount and patterns of code-switching to be used. A more suitable MOI could be a school-based one which allows not only school administrators and parents, but, also, more importantly teachers and students to decide the most helpful medium of instruction. After all, every school faces different situations, therefore, it would be wise to provide opportunities for students and teachers to express their ideas about whether the mother tongue should be used in classrooms and, if so, how should it be used effectively both in English lessons and for other subjects, especially in secondary classrooms.

In addition, I consider that code-switching should be used in a controlled manner. Since, firstly, it would be extremely useful for teachers to plan their lessons and decide the language mode which best suits the needs of the students, and secondly, teachers should have the capability to identify their students' level of English proficiency, and then decide the way code-switching should be used. All teachers in Hong Kong must be trained in language usefulness and awareness, so that they will know how to use code-switching positively while minimizing its negative points, so as to benefit students linguistically and academically.

Code-switching in second language learning is a field in which much more examination and exploration needs to be done, but it is the hope of this author that teachers could be soon to be able to use code-switching as a learning and teaching tool with more strategic, effective and proper use. No matter what kind of MOI is used, in the final conclusion, it is an unchangeable principle that to benefit the largest interests of every individual student should be the paramount concern.

As for future study, more research needs to be done on finding ways to achieve effective relationships between the English language and the use of code-switching. The outcome of this research could be of maximum benefit to students and as it could go some way towards establishing real bilingual education in schools.