Soo Kheng Soon
Universiti Kebangsaan Malaysia

Code-switching (or C-S) has been defined in various ways. As quoted in Soh (1984), Haugen (1956) said it is 'The alternative use of two languages'. As quoted in Soh (1984), Ervin-Tripp (1964) limits it - perhaps overly so - to just 'Stylistic alternation when a speaker changes his speech stylistically to suit different social roles'. This implies the use of just one language. Gumperz (1982) classifies code-switching into three varieties. He concentrates on conversational code-switching. He defines it as 'the juxtaposition within the same speech exchange of passages of speech belonging to two different grammatical systems or subsystems'. As this paper will look into examples from written discourse as well, Gumperz's definition is perhaps too restrictive. Marasigan (1983) 'C-S or code choice... refers to the use of two languages in the same sentence or discourse'. This will be the definition adopted for this paper which will seek to illustrate various functions of code-switching in texts obtained from bilinguals in Malaysia and Singapore.

The Linguistic Composition of a Code-switching Society

Hudson (1980) says '... code-switching is permitted in some societies and not in others Permitted' may be a rather strong word; 'possible' might be a better choice as he explains that code-switching 'is not something which the bilingual… does except when talking to a fellow-member of a community which permits it'. It is logical that C-S can only exist in societies which are at least bilingual. Such societies must be linguistically, if not ethnically, heterogeneous and the members of these societies the 'multilingual in a variety of languages or dialects that are functionally differentiated....' (Kuo, 1985). Due to historical and economic reasons, most South-East Asian countries fit this description. Malaysia, for example, has a sole national language, Bahasa Malaysia, with an unofficial second language, English, and a multitude of ethnic languages, chief among which are the Malay dialects, Hokkien, Cantonese and Tamil. Singapore too, has Malay as the national language but it also has three other official languages -Mandarin, English and Tamil. Together with Hokkien, they are the major languages spoken in Singapore today (see Fig. 1). Thus, linguistically these two countries represent what Rustow (1968) describes as a language pattern involving 'a variety of unrelated languages each with its own literary tradition'; and what Fishman (1972) designates as a 'multimodal nation'.

Figure 1: Percentage of Population Aged 15 and Over Who Can Understand Various Languages in Singapore, by Ethnicity, 1972 and 1978.

The Communicative Index, (Index 1) in Fig. 2 gives a clearer picture of the heritage of linguistic diversity in Malaysia and Singapore today. More and more Singaporeans are becoming multilingual. This is probably partly due to the official policy of multilingualism. With rapid economic development, there is increasing ethnic integration with the breakdown of traditional identification of occupations along racial lines as well as increasing ethnic mix in public housing schemes.

In that the majority of the population in both countries is multilingual the possibility of C-S is so much greater. Due to a need for inter-ethnic communication, nearly every member of a multi-ethnic and multi-cultural society learns a second or even third language. It probably starts with what Gumperz describes as diglossia or situational C-S where the speaker 'switches from one language to another as he moves from one sphere of activity to another' (Noss & Llamzon (eds.), 1986). It is an easy step for the same speaker to start C-S in the same discourse. While the linguistic profile of a society can predict the existence of C-S, the occurrence itself can only be explained by examining the psychological and sociological functions of language use of which C-S is a part.

Theoretical Framework of C-S

Corder (1973) advances the concept of the common semantic structure as the only means by which two languages can have identical expression. This means that the same meaning can be coded in two or more languages as long as the speaker has the corresponding 'labels' (words for the same experience; meaning) in the different languages and is aware of the interchangeability. As noted in Soh (1984) although Kohlers (1968) and Taylor (1976) found evidence that this does not occur with equal ease for every category of words, Slobin (1971) suggests that linguistic and cultural universals are more important than linguistic and cultural relativity. As quoted in Soh (1984), Bruner (1975) is in agreement when he says 'Today the emphasis has shifted more towards language (and cultural) universals and how, in general, they impart a universally human character to all information processing, whatever the culture'. In learning an L2, the L2 learner merely has to learn the 'L2 labels' which are 'hooked on' to the first language encodings of the particular experiences or meanings. This Soh refers to as positive transfer (as opposed to negative transfer) which facilitates learning and creates no problems.

Therefore, if the relationship between thought, symbol (language) and referent in Li is this:

In C-S, the speaker merely has to learn the relationship AD as the relationship ABC (L1) has already been internalized. The L2 equivalents (D) are just 'hooked on' to the L1 symbol (A). Once this has been achieved, both A and 0 are available for expressing C. This paper has attempted to answer the questions, what is C-S, where does it occur and now it can occur. The one remaining issue is why does it occur. This can only be answered by looking at the larger issue of language functions.

As quoted in Marasigan (1983), Austin (1962) has identified the illocutionary act as the central component of language use. One can make an utterance - a locutionary act which carries the propositional content but the propositional content may not be the intended message. The illocutionary effect 'determines how the utterance is to be taken' (Marasigan, 1983). The distinction is important as an utterance can carry any of several different illocutionary intentions, although the words and literal meanings are the same. For example, 'Aku sayang pada mu' can be an expression of feelings or a factual statement.

Many philosophers and linguists have attempted to define and group utterances into various illocutionary acts. Searle's (1976) system is considered the most powerful. As quoted in Marasigan (1983) he divides illocutionary acts into six main types. They are representatives, directives, commissives, expressive, declarations, and representative declarations. But explaining how 'Aku sayang pada mu' is to be taken, still could not explain why 'I sayang pada you' for example is used instead, when the former is equally available and indeed more logical. To do so requires taking into account the perlocutionary intention of the speaker. In saying 'I sayang pada you' the speaker (if he is a Malay) has switched to English in making reference to himself and the listener. This may be because open expression of love is not part of Malay culture and in voicing a 'foreign topic', he expresses it in the Westerner's code with which the topic seems more natural. But English, not being the native language, sounds more distant. In this situation, the speaker must convince the listener that he is truly involved and needs a code that would convey greater intimacy. With regard to this 'we' and 'they' dichotomy, Gumperz (1982) says "…participants are likely to interpret 'we' code passages as personalized or reflecting speaker involvement and 'they' code passages as indicating objectification or speaker distance'. Thus the speaker shifts back and forth between two languages trying to convey both intentions. This example shows the possibility of deriving the speaker's perlocutionary intentions by studying the language functions of C-S in spoken and written discourse.

Functions of Code-switching

This study will now attempt to look at examples from each type of illocutionary act as defined by Searle (except that his representative declaration will be subsumed under declaration as Marasigan had done). if this could be done, it would be reasonable to assume that C-S, like language itself, is a universal phenomenon as it occurs in every kind of utterance.

Representative

Example: A: 'Cinderella ye mei you shen me story, (Cinderella also don't have much of a story) Nothing to tell one you see'

A was in a drama class choosing a play to put on. The discussion has gone to and fro for some time in Mandarin and A was apparently getting impatient so he offered his opinion, basically, in Mandarin. The use of 'Cinderella' was understandable as it is an English fairy-tale and probably the students did not know the Chinese version of the title. He was being persuasive and so used the 'we' code. However he also intended to end the discussion on Cinderella by making the others accept his view a form of covert assertion. If his following utterance had been in Mandarin, the tone would have been so weak that it would come across as a statement of fact and nothing more. By switching to the 'they code', the utterance acquired an assertive force. Because this assertiveness was covert, A would not have been embarrassed if his assertion failed. In this way, C-S functions simultaneously as a coersive and a face-saving device that Li could not.

Gumperz (1982) says 'Frequently a message in one code is repeated in the other code, either literally or in somewhat modified form. In some cases such repetitions may serve to clarify… amplify or emphasize a message'. This function surfaces in this example in which B provides an explanation to C:

B:    Fauziah, you must be more flexible, you know.
C:    Apa?
B:    Flexible, lebih lembut sikit.

This type of C-S occurs even in more formal situations. In an example in which a teacher was explaining the finer points of comparatives to her class, the following discourse emerged.

A:    Ar... that's right. Nearer menunjukkan satu perbandingan antara dua, betul?

Culture plays an important role in C-S. The Roman calendar is a Western invention; its way of measuring time is often an inherently 'foreign' concept to Asians. Marasigan (1983) speculates that as this 'time' concept is not inherent in their cultures they express it in the 'they' code or English. This seems to be the case in another example in which a sister wrote to a female student in UKM saying, 'Kami punya semester break will be on the 23rd Januari 1987-30th, sekali dengan raya China'. Another example occurred when a teenager was explaining to her new boyfriend how she spent her time. She wrote, 'saya dibenarkan keluar rumah twice a month'. Here again the time clause is in English. This seems to show that the topic can trigger C-S.

Directives

Example: A: No, ask from Fandi. Takkan tak ada?

In this example A, a Malay, switched to English in an attempt to get another student to do something. This switch could have two functions. The switch to a more formal code could be to signal that it is a serious request and if at all possible, the addressee should act as required. The other function is what Gumperz (1982) calls addressee specification. The addressee is a Chinese and A might have felt his L1, Malay, would probably sound more distant than English to the addressee. However, he himself might have found English distant and so switched again. His next utterance was a plea in Malay. This switch back to informality could be to soften the demanding tone created by his first switch. It may be interesting to note that his directive failed. The implication is that C-S is extremely prominent psychologically and its influence can override that of utterances in L1.

Another two examples from Appendix 6 show the same phenomenon.

A: Tengok. Sekarang tiga-tiga jalan.... Sam in the middle.
A: Fauziah. Jalan biasa, jalan biasa.... duduk sini dan pura-pura go to the middle, aa!

Commissives

Example:    Sekiranya ketiga-tiga perkara yang menakutkan saya itu tidak ada/wujud pada din saudara, pastinya dan tentu sekali 'I'll choose you to be one of my friend'.

In this example, the author constructed a complex conditional. The author committed herself to a course of action provided the addressee could convince her of his sincerity.

It is curious that the author not only switched code for the commissive but she put it in inverted commas as if she was quoting or copying someone's speech. Yet the grammar mistake makes it unlikely. One possible explanation is that she wanted it to be taken orally, that is, as if she was actually there, saying it. Committing oneself to a future action is better done orally if only because it is more convenient should the speaker fail to fulfill his promise (Marasigan 1983). This could account for the written-as-if-spoken style here. The switch itself could be due to the writer's perlocutionary intention of playing down her commissive by making it sound a little like a joke. The switch to a more distant code upsets the expectation of the warmer and more positive emotions that should accompany such commissives. This reduces the gravity of the commissive and thus makes it less binding.

Expressive

Example:    Akhir kata 'Happy New Year'. 'May God Bless you'.

In the above example, a Malay female undergraduate in UKM sent New Year greetings to her Chinese girl friend in another hostel. She could not use her 'native' greeting because the addressee was of a different culture, so she chose her greetings from a language and culture the addressee might understand better. In so doing, she also switched code. The code-switch shows her consideration for her friend and from this the addressee can infer her sincerity. This expressive conveys the writer's feelings for her friend more successfully because of the code-switch. It should be noted that the English greetings are very much Malay in character too, especially the mention of God's blessing. So the speaker has successfully conveyed her intention and at the same time did riot compromise her own identity.

In another expressive, a friend wrote to another UKM undergraduate, Yatie, and began asking questions about Yatie's studies and life on campus:

Tentu sekali Yatie sibuk sebab baru naik semester. So far macam mana sekarang? Hope everything okay!'

In another example, a group of students were discussing their preparations for the final rehearsal.

A:    Three-thirty.
B:    Aa?
A:    Tomorrow.
C:    Eh, three thirty! Kalau barang semua ada O.K. for final rehearsal tapi barang semua tak ada.

Declarations

Utterances of this category are rare. People are not often in the position to hold something as the truth simply because they say so. When declarations occur, it is not surprising that they are usually uttered in the 'formal' code as they are serious and binding (in the sense that both speaker and listeners alike will have to adopt certain attitudes and actions consequent of the utterances). In one example, the writer wrote to her new found boyfriend, accepting his overtures of friendship, saying:

Mulai dari hari ini, we're friend, okay....

On the other hand, C-S can backfire. In another example, almost the same utterance was made.

Well, we're friends. Only time will tell if we can progress into better friends.

But I wish you would not be so arrogant and 'xiao qi'.

The entire letter was almost totally in English because the addressee could not read Mandarin. The writer was educated in Chinese and a careful reading of her letter shows her mastery of English is less then good, although competent. Partly due to cultural constraints (she wrote of trivial matters and came to the point only in the last paragraph) and party due to being forced to operate in her 'they' code, her discourse sounded cold and evasive. This was probably not her intention. For the first time, she opened the letter with 'Dear Ah Kow' instead of the usual 'Dear Mr Tan' or just 'Mr Tan'. This was significant because later in an angry letter she reverted back to 'Mr Tan'. For the first time, she talked of her everyday life - information usually reserved by conservative people for one's inner circle only. The Chinese saying following "xiao qi' is a wish for better things to come in a new year - an indication that she hoped the relationship would flourish. In other words, switching to a distant code 'cramped her style' communicatively and effectively destroyed her attempt to convey her intentions. It dominates the entire discourse in a way the English word 'sensitive' never can. It turns an acceptance into a rejection so that even the subsequent reaffirmation of friendship and invitation (in another letter) were nullified.

Consequently, her agreement to be 'friends' did not sound convincing. This is an example of how the addressor's perlocutionary intention (to establish the relationship on a formal footing) to produce an intended perlocutionary effect (a stronger hold on the loyalty, friendship, etc.) by producing the communicative effect of the addressee recognizing her illocutionary act (expressive: complaining) fails because the actual perlocutionary effect was unexpected (the addressee thought he was being rejected and the relationship was really inadvertently killed before it could even bud).

Interjection

Sometimes people switch codes when they utter interjections. In one example, the writer wrote:

'Wow!!! It is nice to have a new friend.'
'Oops! Sorry ye, kerana bercelotih panjang sangat.'

Loan words

The last function we will look into may be more a phenomenon than a function. 'Foreign' words are incorporated into one's discourse and used as exact substitutes for Li words. In one example, the word 'change' was used by itself in three phrases in a direct replacement of /zhuan/.

C:    Bu'shi. Na'ge Greek Greek to shuo wo men qu change. Na'ge lecturer understand, zhe ge wen ti Ni'de ke'yi change, zen me wo men nan de bu ke yi change.

Concluding Remarks on the Functions of Code-switching

After looking at these few examples of C-S, the one thing we can say of C-S is that it is not idiosyncratic. All the functions discussed previously can be subsumed under the umbrella function of facility of expression. Nababan (1978) is of the opinion that the main determinants of C-S seem to be ease of expression. To this might be added efficiency of expression. It has been shown the C-S is more efficient in expressing certain meanings. Some functions are even unique to the C-S phenomenon. Soh (1984) says 'Choice of language or code has important and often subtle social significance... makes such communication more effective, affectively speaking, between speakers sharing the same languages'. Basically, C-S works by increasing or decreasing the social distance between speaker and listener. These adjustments in themselves are not meaningful until interpreted in the context of the discourse. Therefore, we can postulate that speakers have a certain perlocutionary intention in mind, take into account the topic, the listener and what has been said so far (among other things) and attempt to create the perlocutionary effects through a code-switch.

Implications for L2 Learning

C-S's contribution to our knowledge of the L2 learning process has been unsung but immense. By studying C-S, Kohlers (1966) discovered that no extra memory capacity was taken up in using two languages - the bilingual stored linguistic information from two languages together in only one storage. Similarly, according to Soh (1984), Dulay and Burt (1974) conclude that universal cognitive mechanisms are the basis for the child's organization of a target language.

In the light of these findings, the vigorous exclusion of C-S from education is theoretically baseless. If the bilingual stores linguistic units 'as semantic representations which could be activated through the use of either language irrespective of the original input language' then C-S may be the natural way multilinguals use languages. As Soh (1984) puts it, 'True bilingualism may not only require facility in two languages separately considered but also interchangeability between the languages'. According to his definition, the ability to C-S may become a measure of bilingualism. If the ultimate objective of ESL programmes can be taken to be to produce bilinguals (passive ones minimally: and co-ordinate ones ideally), then the status of C-S must be drastically elevated.

On the other hand, the L2 learner's L1 may be useful in the acquisition of an L2. Wode (1978) in his Kiel Project, found that 'in naturalistic L2 acquisition and in FLT, children/students do rely on their prior L1.... they do so in a highly systematic way'. No learner comes to a learning situation as a tabula rasa. L2 learners rely upon prior learning of the L1 to facilitate their new learning and the semantic system of one language acting like a filter helps, indeed forces interpretation of the other language. The conclusion to be drawn from this is that the common semantic structures shared by two languages 'should be capitalized on or made the fullest use of when teaching a second or foreign language' (Soh, 1984).

Soh's investigation into bilingual testing shows that pupils in Chinese schools 'performed at a definitely higher level when the information (question stems) was made available to them in Chinese than when it was presented in English'.

This raises the question as to whether L2 learners have been made to under-perform consistently because of the monolingual approach in ESL education and evaluation where the L2 learner is expected to 'switch off' his L1 while learning and using an L2. In view of Soh's findings, it is certainly plausible that a bilingual approach to testing may better reflect L2 learners' mastery of the L2.

If these findings are confirmed, then ESL education will be revolutionized. C-S will become an important pedagogical tool in ESL teaching. As a foretaste of what might be coming, Lado (1978) said 'the common practice of using monolingual materials in both languages without due regard to the special learning capabilities, cognitive mapping and linguistic and cultural needs of bilinguals will create problems...'

1. Bilingual curriculum materials

Soh (1978) says that in a revised view on the role of contrastive analysis, Lado (1978) suggests applying the method to develop bilingual materials. These bilingual materials should ideally have sizeable common elements through the use of similar if not exactly the same topics presented through two languages. Examples are bilingual texts, bilingual word lists and dictionaries and even exercises involving C-S. The rationale is that pupils need not learn two sets of non-linguistic content in two languages, instead they will learn through one set of non-linguistic content how the 'message' can be coded by using two linguistic systems.

2. Bilingual instructional methodology

Lim (1970) says 'One of the basic issues in L2 teaching is whether one should use the first or native (L1) language in the process'. As it is now assumed that no learner comes to a learning situation as a tabula rasa, the implication is that the C-S can be employed so that the pupil's non-linguistic knowledge as well as his experience in learning the L1 could be more directly and fully made use of in learning an L2. 3. Bilingual teachers

To handle bilingual instructional materials and methodology, teachers will ideally have to be 'balanced bilinguals who can switch with ease between languages' (Soh, 1984). In practice, they only have to be proficient enough to be able to make reference to the other language in the course of instruction when necessary. 4. Bilingual testing

Soh (1984) has found a strong enough correlation between results in bilingual tests and results in monolingual tests utilising the two languages involved. Bilingual testing may offer a more economical way of testing ability in two languages simultaneously.

Conclusion

C-S is still a largely unknown area. What has been discussed in this study represents only a fraction of what more there is to know. And yet from this fragmentory knowledge, the implications to education are already quite obvious. As more research is done to determine the conscious and unconscious causes of C-S, surely language education will have to change to take into account this neglected language tool.

References

Corder, S. Pit (1973), Introducing Applied Linguistics, Hormondsworth: Penguin.

Fishman, Joshua A. (1972), 'National Languages and Languages of Wider Communication in the Developing Nations', J. A. Fishman, Language in Sociocultural Change, Stanford, California: Stanford University Press.

Gumperz, John (1982), Discourse Strategies, New York: Cambridge University Press.

Hudson, R. A. (1980), Sociolinguistics, Oxford: Cambridge University Press.

Kuo, Eddie C. V. (1985), A Quantitative Approach to the Study of Sociolinguistic Situations in Multilingual Societies, Singapore: SEAMEO Regional Language Centre.

Lim Kiat Boey (1970), 'The Use of the First Language in Second Language Teaching and Learning', RELC Journal, No. 1.

Marasigan, Elizabeth (1983), Code-Switching and Code-Mixing In Multilingual Societies, Singapore: Singapore University Press.

Richard Noss and Teodoro Liamzon (eds.) (1986), Sociolinguistic Aspects of Language Learning and Teaching, Singapore: SEAMEO Regional Language Centre.

Rustow, Dankwart A: (1986), 'Language Modernisation and Nationhood -An Attempt of Typology', in J. A. Fishman, C. A. Ferguson and J. Dass Gupta, (eds.), Language Problems of Developing Nations, New York: Wiley.

Slobin, Dan I. (1971),

Psycholinguistics, London: Scott, Foreman.

Soh Kay Cheng (1984), 'Code-Switching of Primary School Children', unpublished Ph. D. thesis, National University of Singapore.

Steinmann Jr., Martin (1982), 'Speech-Act Theory and Writing', in Martin Nystrand (ed.), What Writers Know, New York: Academic Press.

Wode, Henning (1979), Studies in Second Language Acquisiton, Singapore: Regional Language Centre.

Wong, Irene (1979), Code-Switching in Bilingual Contexts: A Malaysian Perspective, Singapore: Regional Language Centre.