Can code-switching in preschoolers be determined by context/activity | Education

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.

In our multi-cultural society, bilingualism is becoming increasingly important. When speakers of two or even more languages come into contact with others from similar as well as many different linguistic communities, their language can change in surprising ways. Within the literature on bilingualism an interesting facet of bilingual speech has often been noted: people will switch between their languages during conversations and even during utterances.

This switching between languages has been identified in theresearch in many different communities around the world: in EastAfrica, North America and the India to name a few. What is it, then,that causes or is involved in this switching process and what can ittell us about the speakers and the way that they are interacting withthose around them? This review will identify the meaning of the term'code-switching' before examining, firstly, the theoretical models thathave been used to analyse code-switching, and, secondly, howcode-switching has been researched in bilingual children.

The first problem for linguists is creating a way of talking aboutthese shifts. To help avoid the problem created by the differencesbetween languages, dialects, creoles and so on, the terminology of'codes' is used within this literature, so that changing from one wayof speaking to another is known as code-switching. This needs to bedistinguished from code-mixing. Code-mixing tends to involve the use oftwo languages within one utterance such that there is no clear breakbetween the use of one language and the use of the other. Incode-switching, by contrast, the break between one language and anotheris normally much clearer. There are two general ways in whichcode-switching has been analysed, at the micro-level of conversation,and at the macro-level of social meanings.

Types of code-switching have sometimes been placed into differentcategories in order to highlight the importance of each (Macswan,2004). Poplack (1980), within the variation theory, identifies threedifferent types of code-switching. The first, tag-switching, refers towhere a short phrase or utterance from the second language is placedwithin the first - this is normally used for short, well-known phrasessuch as 'you know'. The second is inter-sentential code-switching, thisis changing codes at end of sentences or phrases. The most complex typeidentified by Poplack (1980) is intra-sentential code-switching. Thisinvolves the insertion of phrases from the other language within thesentence or phrase structure of the first language, but, to contrastwith tag-switching, these are not well-known or standard phrases. Thislast type of code-switching, then, represents the most complex andcreative type.

Another important type of code-switching is that identified byGarcia (1983) as interactional code-switching. This is the situationwhere one party in the conversation speaks in one language while theother party uses a different language and each sticks to their own.This may happen where families have moved across linguistic borders andthe parents are trying to maintain the traditional language of thefamily, while children are instead using the language of the hostculture.

So, within this Tower of Babel how do people choose when tocode-switch? What are the factors that affect their choice of languageand in which situations? Early influential approaches to answeringthese questions were rooted in sociolinguistics, and one early and oftcited theory was that created by Blom and Gumperz (1972). They identifythree main areas that affect the choices a bilingual person makes aboutwhich language to use. These are: the setting in which the interactiontakes place, the participants involved in the interaction and the topicwhich is being spoken about. A further distinction made by Blom andGumperz (1972) was between metaphorical switching and situationalswitching. The latter refers to where a person changes language becauseof the factors involved in the situation. The former refers to factorsthat affected changes in language while the situation remained thesame. Metaphorical shift are also seen in a social light, such that theinterpretation of the listener is just as important as the changes madeby the speaker. It is the social factors that are, in this analysis,often the most important for the choice of code. A number oftheoretical models which each place varying emphasis on the socialfactors will be examined in turn.

There are three main theoretical models identified by both Wei(1998) and Myers-Scotton (2005) that have been influential in analysingthe process of code-switching in bilingual speakers. Many writersconsider the first theory discussed here, the Markedness Model, the onethat has had the greatest influence. The most fundamental aspect of theMarkedness Model is that it sees the way that bilingual people choosethe words they use as fundamentally a process of negotiating theiridentity (Myers-Scotton, 2005). At its heart this model is interestedin the social motivations of the speaker. The model analyses thesesocial motivations in terms of marked and unmarked choices. The modelassumes that any bilingual speaker, when encountering a particularsituation in which they have a number of linguistic choices to make,will tend to make the choice that seems least 'marked'.

An important concept to understand within the Markedness Model is theidea of rights-and-obligations sets (RO sets) (Myers-Scotton, 1993).The RO set refers to an abstract scheme that is set up between membersof a linguistic community. The RO sets control how members of thatlinguistic community use language with each other. There are,therefore, a number of particular RO sets that have become conventionalor socially accepted within a linguistic community. Each type ofinteraction has, what is called within this model, an unmarked RO set,or a normative way of talking. Another way of looking at this is, asMyers-Scotton (2005) explains, that if a person is using an unmarked ROset then their utterances are not considered controversial or unusual -they are fitting in with social rules. In addition to the idea of ROsets Myers-Scotton, (1993) refers to the idea of indexicality - this iswhat links the RO sets to the particular social contexts in which theyare used. Talking about indexicality refers to the idea that making alinguistic choice points to some other social fact. It is not theactual changes themselves that have a one to one correspondence withmeanings, it is that the switches can indicate a number of possiblesocial meanings. It is how other people choose to interpret thesechanges that conveys the social messages between people.

Having a number of 'unmarked' interactions begs the question of howthese types arise. They come about through social and cultural normsbut it will be dependent on the situation. The influence of power iscertainly important, a more high-status individual, for example, mayoften dictate to those of lower status what the 'unmarked' linguisticspeech patterns are.

In contrast to the 'unmarked' choices, then, it is the 'marked' choicesthat stand out in bilingual speech. Myers-Scotton (2005) argues thatwhen a speaker uses an RO set that is unusual in a particularsituation, they are requesting, or negotiating, a change in theinteraction. The way in which this negotiation is carried out isintrinsic to the identity of the speaker, or, in other words, itprovides a social clue about what they are trying to achieve. Inparticular, the importance of choosing a 'marked' RO set is that itaffects the relationships between that person and the others with whomthey are interacting. There are two dimensions on which Myers-Scotton(2005) argues that these changes can primarily be seen to be takingplace: power and solidarity. Solidarity refers to an attempt to allyoneself with those who are from the same social group, the most obviousexample being the family. While solidarity emphasises similarities,power tends to emphasise the differences between people. Certain socialsituations have, built into them for various reasons, inequalities inpower balances. This might operate, for example, between employer andemployee, or, more subtly, inside the family between father and son.

A fundamental assumption of the Markedness Model is that people makerational choices about the types of words that they choose. Theassumption, then, is that a person weighs up the pros and cons of usinga particular language and makes their decision based on what is bestfor their own goals. It is a cognitive calculation in which a personaims at obtaining the best possible final result for their own aims.

The rights and obligations framework provided by Markedness Model hasbeen used by Myers-Scotton (2005) to analyse the way in which peoplecode-switch while they are talking. From this perspective, thecode-switching that people carry out online can be extremely revealingof their aims. Myers-Scotton (2005) identifies a number of differentexamples of types of code-switching which are revealing in this way. Agroup may stake out their claim for a particular identity by theparticular type of code-switching they employ. Rampton (1995) providesthe example of young people in England from Punjabi and Creole speakingcommunities mixing three different languages in one utterance. Withintheir social network this would generally be considered by theMarkedness Model as an 'unmarked' RO set because it represents thesocial norm within that community of speakers. On the other hand, whatthis form of code-switching does is separate this group of speakersfrom, for example, those people who just speak English. In this way,then, the code-switching speakers are negotiating a different identityfor themselves and gaining a sense of solidarity in the process.

A related function of code-switching is described by Myers-Scotton(2005) as 'claiming an identity'. For example, in some societiescode-switching between a local language and English is associated withhigher prestige, because any knowledge of English means that theireducation has been 'better'. In order, then, to claim a higher statusfor themselves a person may code-switch to attempt to gain entry to alinguistic community with a higher prestige. Here the speaker'scode-switching can be seen to be operating on the power axis.Code-switching can often, though, instead of being a method of wieldingpower or manipulating solidarity, simply be the 'unmarked' linguisticchoice. Within many communities, the norm of switching betweenlanguages is well established. Myers-Scotton (2005) provides theexample of teenagers from Nairobi who use combinations of Swahili andtheir own native dialects. This mixing of languages helps to create aparticular identity for this sub-group.

This last example of code-switching might be considered a move onthe part of the speaker to attempt to raise themselves socially. Theopposite manoeuvre can also be achieved, still in an attempt to benefitthe speaker. Myers-Scotton (2005) provides the example of two Kenyanmen who are both trilingual talking to each other in English, theunmarked choice for what is, in this case, a business interaction.When, however, one man wishes to appeal to the other man for a loan, hedoes so on a personal level and moves to the other language which theyshare, Swahili, in order to move the encounter onto an informal level.This is a 'marked' shift in which, by code-switching within the flow ofthe conversation, one man can attempt to influence the other by using acommon language that has lower social status but makes him seem more'ordinary'. In this way the man shows greater solidarity.

Where there are more than two speakers and, while two share alanguage which the third does not, code-switching can be seen as a wayof excluding the third person (Myers-Scotton, 2005). Paradoxically itmay be by adopting a language that is known to only two people, thesocial distance is decreased between those two. Effectively what is an'unmarked' choice for two people becomes a particularly 'marked' choicefor another person with a different RO set. Whether or not a particularcode-switch is an attempt to exclude or talk about another person willdepend on the situation, but it may often be less of an attempt toexclude and more an attempt to bring a particular person closer.

The Markedness Model of code-switching has been criticised on anumber of grounds. Wei (1998) points to the fact that in realconversations the situation is not as clear as the Markedness Modelmight like to make out. The idea that participants in a conversationhave negotiated and agreed what accepted norms of code-switching aremight be considered far-fetched. Similarly, is it really possible forpeople to construct these types of meanings cognitively as theirproduce language? Wei (1998) is essentially asking whether theMarkedness Model is simply reading too much into code-switching andinstead calls on more pragmatic principles, arguing that whatparticipants in a conversation are actually concentrating on is theproduction of meaning. Wei (1998) argues that it is not in the externalmeanings of changing language that code-switching can be understood,but in the particular context in which it happens. This kind ofdistinction is characteristic of the clash between the macro and microapproach to code-switching.

Conversation Analysis provides a much more 'micro' approach tocode-switching. This approach looks closely at the context of theconversation for the reasons for which bilinguals code-switch (Auer,1998). Instead of using categories that might be considered external tothe conversation, within a Conversation Analysis (CA) approach ananalyst is looking more at the local production of meaning. The CAanalyst believes that meanings are created online in the conversationrather than being imposed from outside. Taking a micro approach tocode-switching allows the CA analyst to pay greater attention to theway in which speakers directly affect each other's speech. It should bemade clear that CA analysts still largely agree that the production oflanguage through code-switching is socially significant, but they aremore interested in finding concrete evidence. Wei (1998) explains thatCA analysts try not to make any assumptions about what code-switchingmeans for identity, power relations or formality. On the contrary it ismore important to pick out the individual, local, examples of the wayin which language is produced.

What, then, does code-switching tend to mean for the CA analyst? Wei(1998) argues that it can have many different types of strategic usewithin a conversation. It can, for example, be used to signal the endof a turn, that the subject of the conversation is changing as well asother higher level meanings related to social identity or the contentof the conversation. Wei (1998) provides an example of how importantcode-switching is in contextualisation. Examining the case ofCantonese-English speakers in Tyneside, there are clear examples of theway in which these speakers use their code-switching to identify theelements or topics of the conversation that they are referring to.Code-switching is also found to be extremely important in the structureof the language. Overall the CA analyst argues that it is only througha close examination of the intricacies of the language that anunderstanding of code-switching can be gained. Meaning is seen toemerge in this framework from the use of language, rather than beingimposed from the outside.

Myers-Scotton (2005) explains the difference between CA and theMarkedness Model in terms of levels of analysis. CA analysts argue,therefore, that at the conversational level, the production of languageis sufficiently separated from the social level that the socialanalysis becomes redundant. Auer (1995) argues that in one situation aparticular code-switch can have one meaning, while in another it mightmean something different - the only way to be able to distinguish thetwo is to carry out an in-depth analysis of the local context of thespeaker's production of language. The Markedness Model, then, willimplicate meanings into code-switching, giving social interpretationsto the way in which speakers change code. CA, meanwhile, does notaccept this idea that there are social norms which are widelydistributed relating to the way in which code-switching is used, andthat bilingual speakers of a particular community understand thesemeanings. In other words CA analysts do not believe in RO sets.Myers-Scotton (2005) emphasises that the two approaches are notmutually exclusive. CA analysts, like those who accept the MarkednessModel, see the social factors as being important, it is the level ofanalysis that is different. This is most clearly seen in the importancethat CA analysts place on where a code-switch occurs in a conversation,whereas the Markedness Model is more interested in the particularchoice the speaker makes. Having said that, though, the MarkednessModel has proved extremely influential and is probably more widelyaccepted than CA.

Critics of CA have pointed to its sidelining of socio-cultural factorsas its main weakness. Stroud (1998) argues that the social and culturalfactors are so heavily involved with code-switching choices that toignore them leaves an incomplete analysis. Similarly, code-switchingoccurs within a socio-cultural context specific to a society and so itis difficult to use CA outside of a Western setting.

The third important model that Wei (1998) and Myers-Scotton (2005)describe is that based on Accommodation Theory. Sachdev & Giles(2004) identify three major factors that are important in howAccommodation Theory sees bilingual speakers. Like the MarkednessModel, Accommodation Theory assumes that bilingual speakers haveinterests and concerns beyond the immediate needs of the conversationin which they are taking part. Communication between bilingual speakersoften has a complex socio-historical background that is weavedthroughout the conversation which has an important effect on how itdevelops. Again, like the Markedness Model, the Accommodation Modelsees bilingual speech as being littered with markers relevant forsocial categorisation. Thirdly, bilingual speakers have preconceivedideas about how different types of communication should be carried out.In other words there are often certain norms for intergroupcommunication and other norms for intragroup communication.

All of these ideas revolve around the social psychological theory ofTajfel and Turner, (1986) in which the social category to which onebelongs has great importance for how they are treated, and, in turn,how they treat other people. Classically, 'ingroups' are favoured over'outgroups'. These different groups are not, however, fixed in placeand can, and often do, change as we move through differentenvironments. A person who was once a member of an outgroup can,through a change of participants, become a member of an ingroup.

Accommodation Theory places a great emphasis on the idea of socialdistance and how it is affected by the combination of code-switches.There are two directions in which speakers can travel: they canconverge if they are attempting to close the social distance anddiverge if they are attempting to increase the social distance. Byrne(1969) points out that similarity breeds attraction and it is thissimilarity in language that is deemed to be important in the linguisticconvergence of speakers. Myers-Scotton (2005), meanwhile, likens thesekinds of shifts to those that are analysed in the Markedness Model onthe power and solidarity axis. A convergence can, for example, signalacquiescence or a handing of the power advantage to another party byswitching into their preferred tongue. A divergence might, in thiscase, signal an attempt to increase the speaker's power in thissituation. Myers-Scotton (2005) sees convergence as the moreunconscious of the two processes as this is often what people will dowithout thinking; divergence, on the other hand, often requires aconscious effort to be different.

The three models discussed so far provide powerful ways of analysingthe context and situational effects of code-switching. But, it is alsoclear that an important contextual variable in the type ofcode-switching employed by bilinguals is in whether they are speakingwith other bilinguals or not. Weinreich (1966) points out that when aperson who is bilingual speaks to a monolingual they are subject to anumber of constraints which are different to those operating inconversation with another bilingual. Grosjean (1999) reports the threecategories or 'modes' that have been identified in this regard byHasselmo (1970). Amongst Swedish-English speakers in the US the threecodes used are: speaking English to English monolinguals,Swedish-American with some bilinguals and American-Swedish with otherbilinguals. The last two cases involve a switching of the base languageused but both will involve code-switching in the conversation. These,then, represent the different language modes and have been identifiedby Grosjean (1985) as occurring along a continuum.

Grosjean (1999) identifies how language modes are adopted as beingbased on levels of activation. These levels of activation will varyacross situations as the bilingual speaker comes into contact withthose with varying levels of confidence in their two (or more)languages. Grosjean (1999) points to a wide variety of factors thatmight affect the way in which different combinations of activations aremade. These include: the topic, type of language used, the socialrelationships between speakers and the habits built up betweenspeakers. Within this analysis, code-switching is most likely to occurwhere the activation of both languages is relatively high. Grosjean(1999) argues that, at this level of equal activation of bothlanguages, it is also likely that shifts in the base language mayoccur. These might be occasioned by changes in the situation, topic, orthe people with whom the speaker is interacting.

Research has begun on the idea of language mode and some results havebeen encouraging. Treffers-Daller (1998), for example, examined aTurkish-German bilingual informant. When speaking to the Germanspeaking family that spoke some Turkish the bilingual spoke primarilyin German with only some borrowings and little code-switching. Speakingto Turkish-German bilinguals the informant did not know well, there wassome increased code-switching, but not a great deal. When the informantspoke to a close friend who was bilingual, however, there was a muchgreater degree of code-switching. These code-switches were integratedto a much higher degree with the conversation and occurred with greaterfluidity. Grosjean (1997) found similar results in French-Englishbilinguals where code-switching was clearly affected by language mode.

Looking at the wider implications of this analysis, the importanceof the linguistic context of the speakers clearly has a pronouncedeffect. Monolingual situations tend to reduce the activation of thebilingual's 'other' language whereas multilingual situations do thereverse. In particular, some studies in children have found evidence tosupport this claim. Genesee, Boivin & Nicoladis (1996) examinedbilingual children whose average age was just over two-years old andanalysed how they interacted with their mother, their father and astranger who did not speak their strong language. They found that thechild tended to concentrate on the language that the mother spoke whenspeaking to the mother, and the same for the father and for thestranger. The amount of code-switching the child carried out was foundto be highly correlated to the amount of code-switching that the parentused. The disadvantages of this study were that it only involved asmall sample size of four infants. In addition only two of the infantswere able to code-switch with any efficiency, while it seemed to bebeyond the capabilities of the other two.

Early research into bilingualism tended to suggest that code-switchingin children was the result of a difficulty differentiating one of theirtwo or more languages from the others. This could be seen as a problemfor bilingual children because, as Meisel (2004) points out, theiroverall language acquisition would be held back. The worry was thatchildren would develop a kind of fusion of two different languagesystems such that it would be difficult for them to disentangle twogrammars. By the 1970s, though, according to Meisel (2004), researchersagreed that bilingualism did not seem to have any observable adverseeffects for the growing bilingual child. The question, therefore,turned to the stage at which languages began to be differentiated bythe child. As children certainly display a mixing of their two (ormore) languages in the first few years of speech production, does thisrepresent a fusion of grammars and the consequent problems that mightfollow, or does it simply represent a particular way of talking thatbilinguals use? This question is similar to asking whether children aretruly code-switching or just code-mixing.

The more recent evidence tends to suggest that children are, infact, able to keep their languages separate, but mix them naturally asa new way of talking, and they can do this from a very young age. Ithas been suggested that children demonstrate code-switching from asearly as 2 years of age. Köppe & Meisel (1995), reviewing theliterature, find that children have sufficient competence in theirlanguages to be able to make a number of important code-switches. Theycan, for example, choose the code on the basis of the person who istalking to them, and do not generally make grammatical mistakes thatare the result of mixing their codes. Köppe & Meisel (1995) explainthat it can be difficult to tell with younger children who have justbegun to talk whether they are code-switching or code-mixing but theevidence available doesn't support one hypothesis more than that other.Lanza (1992), however, examined code-switching in two-year-olds inrecordings of spontaneous speech in the family environment. This studyfound that the children did differentiate their language specificallyin line with the context. These results tend to support the idea thateven two-year-olds can engage in code-switching.

An important way in which the use of code-switching has beenanalysed in children within the Markedness Model is on the power axis.Jørgensen (1998) draws on data gathered from a study of Turkishspeaking children attending a Danish school (Jørgensen, Holmen, Gimbel& Nørgaard, 1991) in an analysis of how power is wielded bychildren through code-switching. The participants involved in thisstudy were mostly of primary school age.

One strong finding was that Danish, the 'official' language wasprimarily used in the public discourse. When the children wanted toshow that they were playing by the rules and following the instructionsthat had been given to them, they primarily used Danish - the'official' language of the environment. When the children wanted tomove over to private conversation between themselves, they tend to moveinto Turkish. Turkish can also be seen to be used as an emotionallanguage - while the Danish is for 'playing nice'; the Turkish is usedin sudden outbursts or rowdy behaviour. In this way the language usedis attached to the attitude adopted by the child. Jørgensen (1998)analyses these shifts in terms of the social environment surroundingthe two languages. Danish is used as the official language of theschool while Turkish is seen as the language of home and privatelyexpressed emotion. This is why it is used in this way by the children.These types of shifts are an example of what Jørgensen (1998) refers toas global factors, in other words all the children, on average, withthe same languages demonstrate these sorts of shifts in language.

More individual-based code-shifts are seen in children in Jørgensenet al.'s (1991) study. The issue of power-wielding can be seen to actat an individual level as, for example, the children interact with eachother to try and gain the upper hand in various conversations. Aconversation is presented between two bilingual girls by Jørgensen(1998) which shows these interactions clearly. While arguing, two girlsare shown to code-switch continually throughout the conversation, butrather than signifying shifts in subject-matter, attitude or theme,they are used as an attempt to redirect the conversation. Thecode-switch is used here to emphasise a point that has been won in theargument which is then sometimes acknowledged by the other party whofollows into the same code. This kind of interaction shows theimportance of code-switching among bilinguals within the same communityand how power can be wielded within these codes in order to achievespecific aims within the conversation.

Jørgensen (1998) presents a second analysis of four Turkish boys whospeak only a limited amount of Danish and who are involved in aschool-type task. They primarily carry out their conversation inTurkish, perhaps mostly for the practical reason that they don't speakthat much Danish. However Danish is certainly used for power-wieldinghere as the boys are aware that Danish has the greater power attachedto it and so they code-switch when they want to assert their power overthe other boys and show their own dominance.

Amongst children who are still engaged in learning languages, the skillwhich they possess in each will be an important factor in theircode-switching. It may not be possible for them to express themselvesin one language and so they are forced to move to a different language.An alternative situation is where the child, perhaps because of levelsof activation described by Grosjean (1999), has greater access at aparticular time to one language rather than the other. In thissituation socio-linguistic factors can be seen to take a back seat tomake way for simple practicality. This fact highlights one of thedifficulties with this type of research, especially with children,where the socio-cultural analysis relies on assuming that code-shiftshave particular meanings. Clearly sometimes they don't, they are simplythe result of practicality or more individual concerns.

There has been some limited research carried out into thecode-switching of young children. Some of this research provides someevidence of, and can illustrate the difficulty of, searching forsocio-cultural meanings. Köppe & Meisel (1995) provide an examplethat is more practical in nature. They emphasise the importance ofcode-switching in role-playing with different codes being used toidentify with different roles. This appears to be one of the earliestforms of code-switching to emerge. Another early function ofcode-switching to emerge is using it for emphasis. Redlinger & Park(1980), for example, find that one word can be repeated in the otherlanguage to emphasise that word. Other reasons for repeating a word inthe other language might be practical in that they want to ensure thatthey have been understood (Lanvers, 2001).

Some code-switches are thought to be the result of children temporarilyforgetting, or making a mistake about which code or mix of codes isappropriate within a particular context. Lanza (1997) finds evidencethat children sometimes carry out 'initiated repair' of theirutterances. This means that they are occasionally reminded by otherpeople that they are using the wrong language. This is not alwaysinitiated from outside although Redlinger & Park (1980) point toexamples of this being carried out by a child of his own accord.

Some support is found in the literature for aspects of the theoreticalmodels reviewed earlier. Two particularly interesting examples involveyoung children using code-switching on both of the axes on theMarkedness Model: solidarity and power. Bauer, Hall & Kruth (2002)highlight how one child uses code-switching to emphasise the powerrelationship. When leading an activity the child makes her own codechoice, whereas when following, acquiesced to other's code choices.Oksaar (1970), meanwhile, interpreted a child's deliberatecode-switching out of the 'accepted' norms as an attempt to increasesolidarity and decrease social distance. Similarly, Köppe and Meisel(1995) find evidence that supports aspects of the Markedness Model thatmake reference to inclusion and exclusion. They find that childrencode-switch in order to either include or exclude other children.Cromdal & Aronsson (2000) also found these features when examininghow children code-switched during play. In this situation children werefound to code-switch as an aid to contextualisation.

In conclusion, while social psychological factors are important, asGrosjean (1982) points out, code-switches do not always have a meaning;especially in children, there are certain pragmatic considerations tobe taken into account. Preschool children are still learning theirlanguages and so have a number of practical constraints around whatthey can achieve. Despite this, the theoretical models examined pointto some of the meanings that can be attributed to code-switching andthe context within which it occurs. These included the socio-culturalfactors made manifest by shifts in solidarity and power. Researchcarried out by Jørgensen (1998) has made the power axis particularlyclear. Both the Markedness Model and Accommodation Theory can be seento share a similar viewpoint about the meanings of code-switches withinconversations. In addition, Grosjean's (1999) analysis provides aparallel emphasis on the importance of the linguistic environment oncode-switching. The research on code-switching in young children isstill somewhat sparse, but the case studies have shown important linksbetween empirical findings and theoretical models.

Auer, P. (1995) The pragmatics of code-switching: a sequentialapproach. In L. Milroy and P. Muysken (eds.), One Speaker- TwoLanguages: Cross-Disciplinary Perspectives on Code-Switching.Cambridge: Cambridge University Press.

Auer, P. (1998) Introduction. In Peter Auer (ed.), Code-switching in Conversation. London: Routledge.

Bauer, E. B., Hall, J. K., Kruth, K. (2002) The pragmatic role ofcodeswitching in play context. The International Journal ofBilingualism, 6(1), 53-74.

Blom, J. P., Gumperz, J. (1972) Social meaning in linguisticstructure: Code-switching in Norway. In J. Gumperz and D. Hymes (eds.)Directions in Sociolinguistics, New York: Holt, Rinehart and Winston.

Byrne, D. (1969). Attitudes and attraction. Advances in Experimental Social Psychology, 4, 35-89.

Cromdal, J., Aronsson, K. (2000) Footing in bilingual play. Journal of Sociolinguistics, 4(3), 435-457.

García, E.E. (1983) Early Childhood Bilingualism. Albuquerque: University of New Mexico Press.

Genesee, F., Boivin, I., Nicoladis, E. (1996) Talking withstrangers: A study of bilingual children's communicative competence.Applied Psycholinguistics, 17, 427-442.

Grosjean, F. (1985) The bilingual as a competent but specificspeaker-hearer. Journal of Multilingual and Multicultural Development,6, 467-477.

Grosjean, F. (1997) Processing mixed language: Issues, findings, andmodels. In A. M. B. De Groot and J. F. Kroll (eds.), Tutorials inbilingualis: Psycholinguistic perspectives. Mahwah, NJ: LawrenceErlbaum Associates.

Grosjean, F. (1999) The bilingual's language modes. In J. L. Nicol(ed.). One Mind, Two Languages: Bilingual Language Processing. Oxford:Blackwell.

Hasselmo, N. (1970) Code-switching and modes of speaking. In G.Gilbert (ed.), Texas studies in bilingualism. Berlin: Walter deGruyter.

Jørgensen, J. (1998) Children's acquisition of code-switching forpower wielding. In Peter Auer (Ed.), Code-switching in conversation:Language, interaction and identity. London: Routledge.

Jørgensen, J.N., Holmen, A., Gimbel, J., Nørgaard, I. (1991) FromKöy to Køge: a longitudinal study of the bilingual development ofTurkish immigrant children in Danish schools, Language and Education,4, no. 3: 215-217.

Köppe, R. and Meisel, J. M. (1995). Code-switching in bilingualfirst language acquisition. In L. Milroy and P. Muysken (eds.), OneSpeaker- Two Languages: Cross-Disciplinary Perspectives onCode-Switching. Cambridge, UK: Cambridge University Press.

Lanvers, U. (2001) Language alternation in infant bilinguals: Adevelopmental approach to codeswitching. The International Journal ofBilingualism, 5(4), 437-464.

Lanza, E. (1992) Can bilingual two-year-olds code-switch? Journal of Child Language, 19(3), 633-58.

Lanza, E. (1997) Language Mixing in Infant Bilingualism: aSociolinguistic Perspective. Oxford: Clarendon Press; New York: OxfordUP.

Macswan, J. (2004) Code Switching and Grammatical Theory In T. K.Bhatia, W. C. Ritchie, (eds.) The Handbook of Bilingualism (BlackwellHandbooks in Linguistics). Oxford: Blackwell Publishers.

Meisel, J. M. (2004) The Bilingual Child. In T. K. Bhatia, W. C.Ritchie, (eds.) The Handbook of Bilingualism (Blackwell Handbooks inLinguistics). Oxford: Blackwell Publishers.

Myers-Scotton, C. (1993) Social Motivations for Codeswitching, Oxford: Clarendon Press.

Myers-Scotton, C. (2005) Multiple Voices: An Introduction to Bilingualism. Oxford: Blackwell Publishers.

Oksaar, E. (1970) Zum Spracherwerb des Kindes in zweisprachiger Umgebung, Folia Linguistica, 4, 330-358.

Poplack, S. (1980) Sometimes I'll start a sentence in Spanish yterminó en español: toward a typology of code-switching. Linguistics,18, 581-616.

Rampton, B. (1995) Crossing, Language and Ethnicity among Adolescents. London: Longman.

Redlinger, W. E., Park. T. Z. (1980) Language mixing in young bilinguals. Journal of Child Language, 7, 337-352.

Sachdev, I., Giles, H. (2004) Bilingual Accommodation In T. K.Bhatia, W. C. Ritchie, (eds.) The Handbook of Bilingualism (BlackwellHandbooks in Linguistics). Oxford: Blackwell Publishers.

Stroud, C. (1998) Perspectives on Cultural Variability of Discourseand some Implications for Code-switching. In P. Auer (ed.),Code-Switching in Conversation: Language, Interaction and Identity.London: Routledge.

Tajfel, H., Turner, J. C. (1986) An integrative theory of intergroupconflict. In W. G. Austin and S. Worchel (eds.), Psychology ofIntergroup Relations. Chicago: Nelson Hall.

Wei, L. (1998) The 'Why And 'How' Questions In The Analysis OfConversational Code-Switching. In P. Auer (ed.) Code-Switching inConversation: Language, Interaction and Identity. London: Routledge.

Weinreich, U. (1966) Language in contact: Findings and problems. The Hague: Mouton.