RUSSIAN-ESTONIAN CODE-SWITCHING AMONG YOUNG ESTONIAN RUSSIANS: DEVELOPING A MIXED LINGUISTIC IDENTITY

RUSSIAN-ESTONIAN CODE-SWITCHING AMONG YOUNG ESTONIAN RUSSIANS: DEVELOPING A MIXED LINGUISTIC IDENTITY Anastassia Zabrodskaja Tallinn University Introduction The study is concerned with conversational functions of Russian-Estonian codeswitching in Kohtla-Järve, which is a specific region in terms of population structure: Estonians from 20.9 % and Russians 79.1 %. The demographic situation is not homogenous throughout the whole Estonian territory. There are predominantly Russianspeaking areas in North East (Kohtla-Järve, Narva etc), predominantly Estonian-speaking Western, Central and Southern Estonia, and the capital Tallinn with approximately the same size of the two speech communities (see M. Rannut (2004) , Verschik (2007) for discussion). The internal diversity within the Russian-language community (indigenous group vs. newcomers, self-identification with Estonia or Russia) actuates variations in Estonian language skills, degree of contacts with Estonian, and in linguistic and cultural identity (for more detail see Verschik 2002, 2005). Young Russian-speakers of Kohtla-Järve have a more sovereign command of the Estonian language than their parents and use their Russian as a strategy to communicate with their parents, but are essentially moving towards primary use of Estonian as they become socialized into the larger society. Changes in language attitudes, self-perception, and linguistic repertoire have occurred during the last ten years. Considerably, the


Introduction
The study is concerned with conversational functions of Russian-Estonian codeswitching in Kohtla-Järve, which is a specific region in terms of population structure: Estonians from 20.9 % and Russians 79.1 %.The demographic situation is not homogenous throughout the whole Estonian territory.There are predominantly Russianspeaking areas in North East (Kohtla-Järve, Narva etc), predominantly Estonian-speaking Western, Central and Southern Estonia, and the capital Tallinn with approximately the same size of the two speech communities (see M. Rannut (2004) , Verschik (2007) for discussion).The internal diversity within the Russian-language community (indigenous group vs. newcomers, self-identification with Estonia or Russia) actuates variations in Estonian language skills, degree of contacts with Estonian, and in linguistic and cultural identity (for more detail see Verschik 2002Verschik , 2005)).
Young Russian-speakers of Kohtla-Järve have a more sovereign command of the Estonian language than their parents and use their Russian as a strategy to communicate with their parents, but are essentially moving towards primary use of Estonian as they become socialized into the larger society.Changes in language attitudes, self-perception, and linguistic repertoire have occurred during the last ten years.Considerably, the proficiency in Estonian has significantly increased: in 2000 38.6 % of Russian-speakers of Kohtla-Järve claimed to know Estonian (Estonian Statistics Office).Although the population census does not define proficiency and the data are anonymous and selfreported, this is nevertheless an "act of identity".There are several strategies in communication between Russian-and Estonian-speaking people.Zabrodskaja (2006 b) demonstrates how Estonians use participant's language to show their solidarity or to help during the conversation the Russian-speaking teachers/pupils, whose Estonian is very poor.As Russian-speaking persons consider a monolingual conversation in Russian as inconvenient, sometimes interlocutors use each other's mother tongue.In one context this phenomenon can be named language politeness, because the partner's dominant language is used.On the other hand this phenomenon can be described as language training as both wish to practice and improve their knowledge of another language.Thus, the language choice patterns are to a great extent related to the level of Estonian language proficiency.This study is based on five sets of data collected at different occasions: the Estonian language summer camp of Kohtla-Järve Third Secondary School, the summer camps of Kohtla-Järve Pärna School.The data have been collected between 2000 and 2006.The analysis is based on the framework of conversational approach (Auer 1984(Auer , 1995(Auer , 1998(Auer , 1999)).
Seven conversational functions were found in the data in connection with intrasentential/extrasentential switches.Language play and expressive function are encountered most frequently.The language of the quotations is usually Estonian when describing a situation with an Estonian-speaking person.Code switches as reiterations are PDF created with pdfFactory trial version www.pdffactory.comused for clarifying, emphasizing or amplifying a message.The nature of short interactions restricts the possibility of detecting switchings related to the change in participant constellation or topic, but some examples do appear in the data.Quite infrequently, switching may serve to mark side-comments.
Different factors (reporting what Estonian-speaking person has said, highlighting something, discussing Estonian-related topics, emphasizing a particular social role etc.) push Russian-speaking children to develop code-switching skills.Children use the Estonian language speaking with each other in common environment where they are able to test out new patterns of code-switching without fear of ridicule.
This study is organized as follows: firstly, the aims of study and the theoretical background are outlined.Then the socio-ethnical and historical issues of Estonia's Russians are described.Part 3 presents the data and subjects included in this study.Part 4 summarizes earlier theories on conversational code-switching.And at last, conversational functions of Russian-Estonian code-switching are shown and analyzed.

Aims and scope of the study
In this study, code-switching is a term that covers the alternating use of two or more "codes" within one conversational episode (Auer 1998: 1).In addition to grammatical factors, sociolinguistic and conversational factors may play a role in codeswitching and can override rules of two monolingual grammars (Romaine 2000;Thomason 2001;Zabrodskaja 2005).It seems that Auer's model that combines the conversational and grammatical points of view is the most appropriate for analysis of the given case.Many instances demonstrated below show that grammar of Russian-Estonian PDF created with pdfFactory trial version www.pdffactory.comcode-switching is closely related to its pragmatics.The purpose of this paper is application of Auer's (1984Auer's ( , 1995Auer's ( , 1998) ) conversational approach to Russian-Estonian data, and the analysis of conversational functions of code-switching.
The term 'code-switching' is used in two related yet different fields of linguistics: second language acquisition (hereafter SLA) and studies on bilingualism.In the former, code-switching is analyzed in terms of learning strategies, whereas the in latter codeswitching is believed to be linked to competence.Arnfast and Jørgensen (2003) intended to detect the borderline between the two approaches by investigating the use of codeswitching in first-year learners of Danish.Their study points out that code-switching appears as a skill used in early attempts of playing with the languages involved in the conversation (Danish/English and Danish/Polish/German/English).In the literature on SLA, code-switching is little more than a communication strategy to overcome difficulties with expressing oneself (Faerch, Haastrup, Phillipson 1984;Odlin 2000).For a comprehensive overview of the two concepts see Arnfast and Jørgensen (2003: 24-28).
Code-switching in contact linguistics is discussed as conversational phenomenon (Auer 1998).
Following Arnfast and Jørgensen's (2003) concepts, who believe that codeswitching by learners can be considered a skill in the same way as can code-switching among bilinguals, this study tends to differentiate Russian-Estonian code-switching patterns among Russian-speaking children in everyday spontaneous speech from the conversational point of view.
A fundamental starting point for conversation analysis is that no feature of interaction can be regarded as irrelevant, because every speech event contains a structure PDF created with pdfFactory trial version www.pdffactory.comrather than being accidental.Auer (1984) used the theoretical approach of conversation analysis to further develop Gumperz' (1982) interactional perspectives on code-switching in conversation.His continuum model is based on conversational analysis of codeswitching (Auer 1999).For Auer, code-switching is first and foremost a conversational event and it is mainly guided by discourse factors.The starting point of the model is that even bilingual speakers have a preference for one language interaction until certain circumstances force them to use another language.In earlier versions of the model Auer (1984) distinguishes between two main types of language alternation: code-switching and transfer (which was later called insertion).Transfer refers to alternation of a unit of speech "with a structurally provided point of return into the first language" (Auer 1984: 26), whereas code-switching entails alternation at locations in the unfolding exchange that do not allow for projection of the point of return to the language of previous talk.
Both alternations are further divided into discourse-related and participant-related alternations.The former term focuses on instances of code alternation that "cue" the unfolding interaction, while the latter refers to alternation that pertains to participants (speakers' or recipients') language preferences.In Auer's publications of 1995 and 1998 participant-related alternation is renamed to preference-related.In Auer's terms, insertion means intrasentential switching or a switch which takes place within a sentence or rather within a clause.Another treatment of term insertion is found in Muysken's (2000: 3-10) terminology, where it is a term for using of lexical items or entire constituents from one language in another.In the current paper following type of switching is referred to as insertion: into Russian, a base language structure, lexical items or entire constituents are PDF created with pdfFactory trial version www.pdffactory.cominserted from the Estonian language.In the data Russian-Estonian switches are intrasentential: they take place within a sentence or rather within a clause.
Once the data on Russian-Estonian code-switching had been collected, a conversation analysis similar to the method used in bilingualism studies such as Auer (1998) was applied.The linguistic context and the reaction of interlocutors as well as their linguistic awareness are taken into consideration.

Historical background of Russians and Russian language in Estonia
A brief sociolinguistic history of Estonia explains why and how Soviet-time Russian-speaking population differs from autochthonous minorities and immigrant communities from all over the world.
First Russian-speaking settlers appeared approximately in twelfth century on the northern coast of the lake Peipus.At this time, the area had been populated by Vots, a small Finnic people whose language is closely related to Estonian.In the sixteenth century more Russians (fishermen, handicraftsmen etc.) arrived from Pskov and Novgorod area.Some of them subsequently embraced Lutheranism and underwent a language shift to Estonian (Külmoja 1999: 516-518).Starting from the seventeenth and well into the eighteenth century, Russian Old Believers, seeking to escape persecutions, settled on the western coast of the lake Peipus.In sum, this was a mixed area populated by Estonians, Russian Orthodox and Russian Old Believers.Both groups of Russians had at least some proficiency in Estonian.
In the nineteenth century some urban Russian population added but it did not have any impact on the overall demographic picture (see for more detail Issakov 1999).
World War II overrode Estonia twice.Germany conquered Estonia in 1941, and in 1944 the country was again occupied by the Soviet Union.Resulting from mass deportation and imprisonment of the autochthonous inhabitants, the population of Estonia decreased by 104,000 in the course of the first year of Soviet occupation (1940)(1941) Rannut 2004).
After the Second World War, Soviet Communist ideology began to encourage newcomers or migrants to Estonia, who considered Estonia as a part of Soviet Union (M. Rannut 1995Rannut , 2004)).Immigration into Estonia which began after World War II reached its peak in the last decades of Soviet power.In 1934, 87,000 non-Estonians lived in Estonia but in 1959 this figure was already 304,000 and in 1989 even 602,000, growing from 8.34% to 25.41% and 38.47% of the total population, respectively (Viikberg 2000).M. Rannut (1995) describes the Russian migration to Estonia in detail and claims that immigration was favored as a means of producing loyal personnel with "clean papers".The immigrants settled mainly in: 1) the town of Narva (North-East Estonia), which was bombed by the Soviet air force, and lay in ruins, and where during the postwar years resettlement of Estonians was restricted by the central authorities; 2) Sillamäe, an area closed to Estonians as members of a "suspect" ethnic group, due to the uranium mining, and, later, uranium processing; 3) the Kohtla-Järve oil-shale mines; 4) the PDF created with pdfFactory trial version www.pdffactory.comcountry's capital, Tallinn (large factories and Soviet bases), and 5) the submarine base Paldiski, where for Estonians were prohibited to live, and where all monuments reminiscent of Estonia, including cemetery squares, were demolished (M.Rannut 1995: 198-199).
The language situation was made particularly complicated because the migrants settled in a concentrated manner in the north-eastern Estonian industrial area or in the capital of Estonia.Industrialization was the basic economic doctrine in the Soviet Union starting from the late 1920s onwards.Reconstruction of factories began after World War II, and the country's industry was adapted to the Soviet requirements.After the war a new gas-pipe was built to Leningrad in 1948 andto Tallinn in 1953.In the race to develop the atomic bomb the Sillamäe uranium enrichment plant with a company town sprang up rapidly  in the North-east of Estonia.Vseviov (2002) gives an excellent overview of the formation and structure of urban anomaly in North-East of Estonia after World War II.The newcomers grew up in the spirit of propaganda that constantly promoted Russian at the expense of other languages.Non-Russians were expected to master Russian while Russians enjoyed a right to remain monolingual.That is why very few bothered to learn any Estonian.Verschik (2004 a: 135) describes Russification policy up to 1991, the aims of which was to establish Russian as not only single official but also the most popular and used in daily communication language among non-Russians in the Soviet Union.The As communication between the two groups was limited and "everybody knew Russian anyway", as the popular perception was, there was no real need to practice Estonian.Except for the members of indigenous minorities, so-called third ethnicities (i.e., neither Russians nor Estonians) who had only a small share in the population would typically send their children into Russian-medium schools for the following reason: as education in languages other than Russian and Estonian was not provided, they had to choose between Russian and Estonian future for their children and opted for the former because they had some proficiency in Russian as all-union lingua franca, while being completely unfamiliar with Estonian.More than 140 ethnic groups are represented among the migrants that settled in Estonia (Viikberg 1999;Lagerspetz 2005: 15).Many of them had shifted to the Russian language already before arriving in Estonia.
The outcome of Russification policy was, in fact, a voluntary segregation and polarization of the two language communities that led separate lives.Russian was a compulsory subject in all Estonian-medium schools; however, it would be wrong to claim that Estonians were all proficient in Russian.Functional bilingualism was characteristic of those Estonians who had to interact with clients or to work in the public sphere.Only in north east the proficiency in Russian was crucial.There was no motivation for Russians to integrate into the Estonian society and to study Estonian.Contrary to the expectations of the central authorities, the prestige of Russian among Estonians remained low.As it is known from various language shift studies, a very often cited prerequisite for a language shift is low self esteem and belief in the inferiority of one's own language (see PDF created with pdfFactory trial version www.pdffactory.comdiscussion in Wertheim 2003 on Tatarstan).This is a distinguished feature of the Baltic countries where the titular ethnicities did not become Russified as elsewhere in the Soviet Union: cf.Kazakhstan (Akhatova, Smagulova, Shaimerdenova, Suleimenova 2006), Tajikistan (Nagzibekova 2006), Uzbekistan (Pardaev, Mamasoliev 2006), Kyrgyzstan (Orusbayev 2006).The reasons for that are manifold: first of all, prior to the occupation, Estonian had already functioned as a language of sophisticated literature and academia, public administration, courts, higher education, army, etc.In other words, the society was not in need of any other "high" language because Estonian was adequately equipped for the mentioned functions.Second, Russian was considered as a language of the occupier.
Third, Estonian continuously held a high prestige among its speakers.Thus, this was a situation of two competing language hierarchies.Needless to say, the anti-Soviet feelings could not be voiced and the resistance against Russification, albeit strong, had to be expressed in other ways.The declaration of proficiency (or, rather, non-proficiency) in Russian in the census of 1979 is remarkable example of the silent resistance.Raun (2001: 210) observes that the self-reported decline in the proficiency (from 28.3 % claiming to know Russian in 1970 to 24.1 % in 1979) is not credible by any objective means.
Estonia regained its independence in 1991 after more than 50 years of occupation.
Few Russians left Estonia after 1991, but due to the general decrease of the birth rates the proportion of Russians in the population is 25.6 %.Considerably, the proficiency in Estonian has significantly increased: in 1989 only 15 % of Russians claimed to know Estonian as opposed to 44.5 % in the census of 2000.Although the census does not define proficiency and the data are anonymous and self-reported, this is nevertheless an "act of identity" (recall the previous example where Estonians chose to declare the lack PDF created with pdfFactory trial version www.pdffactory.com of proficiency in Russian).The reaction to the Language Law and to the re-establishment of the independence varied among different sets of Russian-speaking population.
According to M. Rannut (1995: 201-202), three groups can be distinguished on the bases of their attitudes: (1) those who identified themselves with the Soviet Union and perceived Language Law as a violation of human rights; (2) those who remained indifferent if the changes would not affect their economic situation; (3) those who supported the independence and acknowledged the symbolic significance of the Estonian language.It has to be noted that the transition from the previous dominant status to a group on a par with other, numerically smaller, ethnic groups such as Tatars, Armenians, Ukrainians, etc., was a dramatic and traumatic experience for many.
After 1991, a solid body of literature on the language situation in the Baltic countries has emerged (see Kolstoe (1995), Laitin (1998), Smith (1998) to name just some major contributions).However, language policy and minority rights issues continue to dominate the research agenda (Pavlenko 2006 b: 90 and references therein).It must be noted that a discussion on the literature dedicated to language policy and language legislation problems in Estonia and in the Baltic countries in general is outside the scope of the present study.According to Pavlenko (2006 b: 90), the changing status of Russian also raises several theoretical challenges.As stated above, Russians in the post-Soviet countries cannot be treated as minorities.There is an ongoing debate on the terms "diaspora" and "post-colonial" (see discussion and summary in Pavlenko 2006 b).The term "diaspora" appears as too general and vague because it refers to all Russians living outside Russia regardless the reasons and conditions of their migration.The concept of "near abroad" is often used in Russian political discourse but it is politically charged and PDF created with pdfFactory trial version www.pdffactory.com is perceived as offensive at least in the Baltic states: it is implied that the post-Soviet countries are not viewed as "real" abroad but remain in the sphere of Russia's influence.
The notions of majoritized minority (Russians) and endangered minoritized majority (non-Russians), suggested by Skutnabb-Kangas (1992: 178) are helpful.Obvious differences between Soviet Union and colonial empires of the past notwithstanding, the notion of post-colonial setting may also prove to be helpful.Ozolins (2002) uses a somewhat similar term, post-imperial.
Russification policy and its consequences on local language use across the former Soviet Union is discussed by Kolstoe (1995), more detailed by Hogan-Brun (2005)  The relationship between language and identity is complicated.Pavlenko and Blackledge (2006) point out that while national/civic identity was a subject of heated contestation in the 1980s and early 1990s, linguistic identities do not appear to be contested in the same way.Giles and Byrne (1982) considered language to be a salient marker of ethnic identity and group membership.However, the one-to-one correlation between language and identity is criticized for its monolingual and monocultural bias, which conceives of individuals as members of homogeneous, uniform, and bounded ethnolinguistic communities (Pavlenko & Blackledge 2006: 4-7).To the best of my PDF created with pdfFactory trial version www.pdffactory.comknowledge there are few if any studies on the links between civic/ethnic identity and increasingly complex linguistic repertoire of Russian-speakers who are emergent bilingual speakers in Estonia.Now Russians constitute the largest minority group with 26% of the entire population of almost 1.4 million inhabitants (Estonian Statistics Office) in Estonia.
Russian is one of the "migrant" languages spoken in Estonia and it has intensive contacts with Estonian, the majority language.As the group of Soviet-time migrants included also people of non-Russian ethnical background assimilated into (Soviet) Russian culture, it is reasonable to call them Russian-speakers or Russophones (for a detailed analysis of the term see Diachkov 1992).
Russian-speakers of Estonia do not form a homogenous social or ethnic community.One can draw a line between the two groups -notably, indigenous Russians, or Estonian Russians, and non-indigenous Russians, the settlers of the Soviet era.
Therefore, in general terms the former do not identify themselves with the latter: they have formed a distinct group in today's Estonia with its defined local identity.Indigenous Russians are so called Old Believers, whose ancestry settled down on the coast of Lake Peipus in the 18th century (Viikberg 2000: 476-477)  Russian-speaking Tallinn University students shows that some sub-groups of the latter do not identify themselves with the Russia's Russians, but have formed a distinct group in today's Estonia with its defined local civic identity (cf.symbolic appreciation of Estonian statehood, customs, material culture, etiquette).Verschik (2005: 289) stresses that there is no sociolinguistic research on generational differentiation on micro level.More and more Russian-speaking children learn Estonian as a second language.Ehala (1994;2000) studies Russian influence on Estonian and discusses second language learner's impact on its structure.He believes that a widely spread bilingualism among Russian-speakers is achievable and this will certainly brings changes in the grammar of Estonian (Ehala 2000: 24).Ehala's pilot study (2006)(2007)  Based on Franceschini (1998) concept of the portrait of a CS-speaker ("CSspeaker" means a multilingual speaker that uses code-switching), Verschik (2004 a: 436) states that a CS-speaker in Estonia is: (1) a younger person that comes from a Russian monolingual setting (typically his or her parents would barely know any Estonian); (2) has typically acquired Estonian as L2 in school or in a university; (3) uses CS for both inand out-group communication.This portrait precisely applies to the young Russianspeakers of Kohtla-Järve.Criticizing Laitin's (1998) approach, Verschik (2005: 289) argues that this new identity group should not be treated as a merely sociopolitical contrast, since its development is manifested in language behaviour, language use and deliberate changes in speech (on this phenomenon see Verschik 2002)."Russian-speaking population" is not a new category of identity in the post-Soviet Estonia as Laitin says (1998: 263).As far as regional (cf.95 percent Russian-speaking Narva and predominantly Estonian-speaking Tartu) and generational differences are concerned, it could be suggested that types of identities are different and Russian-speaking community could become more fragmented as time goes on.Laitin's (1998: 264) claims that "Russian-speaking" population has clearer boundaries and is a more powerful identification in the Baltic states than it is in Ukraine or Kazakhstan.This approach has monolingual bias which was criticized by Le Page & Tabouret- Keller (1985) and Backus (1999).A speech community is thought of as the collection of people that live together and speak the same language.But people's languages differ in pronunciation, grammatical constructions, use of lexical options etc (Keller 1990).Thus, every speech community has a code matrix (Backus 1999: 14) -the set of codes its members use.In a monolingual community, all codes would be varieties PDF created with pdfFactory trial version www.pdffactory.com(dialects or sociolects) of the same language; in Estonia as a multilingual community, they are varieties of Russian and Estonian.Estonia's Russian language is not Russia's Russian and sometimes its grammatical constructions can be perceived by Russia's monolingual speakers unintelligible or strange (Verschik 2004 b).Some segments of Russian speech community has switched from highly planned standard language to more flexible forms.These sociolinguistic factors have to be taken into account.For a microsociolinguist, patterns of Russian-to-Estonian communication appear more subtle and diverse than just two monolingual varieties.This is especially true of Tallinn where people with a various degree of command in Estonian and in Russian interact on an everyday basis in institutions, over the counter, in universities, etc (Verschik 2002(Verschik , 2004;;Zabrodskaja 2006 a).Another question that a microsociolinguist might ask is exactly what kind of Russian (Estonian) the speakers use: actually produced utterances may not be classifiable as belonging to the respective monolingual varieties or even to learner varieties of Estonian (Verschik 2006).Pavlenko (2006 a: 4-5) also argues against the monolingual bias in the study of language and emotions.The monolingual bias also obscures the fact that, in many cases, languages may be linked to professional, rather than national or ethnic, identities.Individuals are agentive beings who are constantly in search of new social and linguistic resources which allow them to resist identities that position them in undesirable ways, produce new identities, and assign alternative meanings to the links between identities and linguistic varieties (Pavlenko & Blackledge 2006: 5-27).

Some sociohistorical facts about Kohtla-Järve
According to the census of population and residential premises of 2000, Estonia is home to predominantly (91%) urbanized Russians who constitute 25.6% of the total population.There are altogether 15 counties in Estonia, whereas 85% of the Russian population is residing in two counties: Ida-Virumaa and Harjumaa.
PDF created with pdfFactory trial version www.pdffactory.comKohtla-Järve is located in the North-Eastern part of Estonia in the county of Ida-Viru.It has always been an industrial town that has become widely known as the capital of the oil shale basin and the chemical industry.Kohtla-Järve is one of the youngest towns in Estonia.It received its designation as a town on June 15th, 1946.Its origin is due to the discovery of the largest recoverable deposit of oil shale in Estonia and the former Soviet Union (see Mettam & Williams 2001: 144-145).
Table 1 shows the ethnic composition in Kohtla-Järve according to the 2000 census (Estonian Statistics Office).Standard Estonian is now the language used and accepted at all levels of society.
Even children of Russian/Estonian marriages tend to grow up speaking Estonian (Daskalovski 2003).While Russian has still been used very frequently in everyday communication, Estonian is taught and used in the school system, and competency in it is a requirement of Estonian citizenship (Hint 1991;M. Rannut 1995).
PDF created with pdfFactory trial version www.pdffactory.com For Russian-speaking youth in central and southern regions of Estonia the Estonian language is more available, because Estonian-speakers prevail there (Ü.Rannut 2005).In those language environments the Estonian language is needed everywhere, it is highly valued and used in everyday situations.However, in Kohtla-Järve the language environment is almost Russian (Ü.Rannut 2005: 37-130).Estonians constitute a minority here, making 20.9 % of the whole population.The most frequent language of communication used to be Russian, in which both Russians and Estonians are fluent.
After Estonia regained it independence, local Estonians began to use more Estonian, whose prestige has increased since it became an official language.

Data
The research is micro-sociolinguistic, dealing with everyday linguistic behaviour.The The sociolinguistic profile of the informants, 77 Russian-speaking primary school pupils, can be described as follows: The average age of children is 9.They all come from Russian-speaking families and live in Russian-speaking Järve area of Kohtla-Järve.The basic language of communication among Kohtla-Järve residents is Russian, in which Estonians are also fluent.
As it was mentioned above, the children attend two Russian-speaking schools: Kohtla-Järve Third Secondary School and Kohtla-Järve Pärna School.The traditional method of Estonian language teaching is prevailing there.Children attend Russian PDF created with pdfFactory trial version www.pdffactory.comschools, where all the subjects are taught in Russian, Estonian is taught as foreign language five times per week.Estonian is as a school subject taught from the first grade; in the kindergarten children learn the basics of the language.Based on this, it can be suggested that all informants have a more or less equal proficiency in Estonian.
As the analyses presented below focuses on the conversational functions of codeswitching in general and not on the linguistic repertoire of a particular participant, the speakers are coded as follow: a girl -G; a boy -B; a teacher -T.If there are many participants in the conversation, then the following abbreviations are used G1, G2, B1, B2, T1, T2 etc.

Current conversation analytic approaches to code-switching
A number of investigations on bilingual language use have been carried out since the mid 1980s which all apply the model proposed by Auer, but which, in addition, have contributed to the development and/or innovation of this approach.There exists for example, a number of studies on code-switching where the functions and patterns of language alternation have been connected to the ideas of conversational functions of code-switching (Grosjean 1982: 152;Appel & Muysken 1992: 118-120;Baker 1995: 77;Auer 1995).According to François Grosjean (1982: 152), bilingual speakers seem to combine languages in order to: fill in a lexical gap, set phrase, discourse marker, or sentence filler; continue the last language used (triggering); quote someone; specify addressee; qualify message: amplify or emphasize ("topper" in argument); specify speaker involvement (personalize message); mark and emphasize group identity PDF created with pdfFactory trial version www.pdffactory.com(solidarity); convey confidentiality, anger, annoyance; exclude someone from conversation; change role of speaker: raise status, add authority, and show expertise.
Similar results have been demonstrated by René Appel ja Pieter Muysken (1992: 118), who discuss the question why people switch between languages.Using the functional framework of Jakobson (1960) and Halliday et al. (1964), switching can be said to have the following functions: ( 1 According to Colin Baker (1995: 77) code-switching has such functions that would vary with age and change with increasing age and experience.For example, codeswitching may be used to: (1) emphasize a point, (2) because a word is not yet known in both languages, (3) for ease and efficiency of expression, (4) repetition to clarify, (5) to express group identity and status, to be accepted by a group, (6) to quote someone, (7) to interject in a conversation, (8) to exclude someone from an episode of conversation, (9) to cross social or ethnic boundaries, and (10) to ease tension in a conversation.Auer (1995: 119-120) suggests, that code-switching requires a sequential account of language choice, in which the language chosen for one speech activity must be seen against the background of language choice in the preceding utterance.From this perspective, the question is not what verbal activities are associated with one language or the other, rather during which activities do bilinguals tend to switch from one language to the other.In answering this question, Auer (1995: 120) shows such conversational categories: (1) reported speech, (2) change of participant constellation, (3) parentheses or PDF created with pdfFactory trial version www.pdffactory.comside-comments, (4) reiterations, (5) change of activity type, (6) topic shift, (7) puns, language play, shift of "key"(8) topicalization, topic/comment structure.
Some of the research on conversational code-switching among children raised as bilinguals has assumed code-switching to be pragmatically differentiated.Analyzing code-switching patterns of both bilingual adults and children, Bani-Shoraka (2005) has found that code-switching is used for a variety of purposes in argumentative and narrative opposition.The pragmatic role of code-switching among bilingual children was demonstrated by Shin and Milroy (2000), Bauer et al (2002).Ervin-Tripp and Reyes (2005) discuss link between children's code-switching as an aspect of pragmatic developing and the adult division between both languages.
The present study attempts to fill the gap in research on conversational codeswitching among bilingual children, who are certain bilinguals (see Skutnabb-Kangas (2000: 106, 573) on definitions of bilingualism) but they study the second language at school.For the qualitative study of Russian-Estonian code-switching a conversation analytic approach has been chosen to analyze bilingual interaction.This means that context is regarded as something essential.
PDF created with pdfFactory trial version www.pdffactory.com Most code-switching functions from Auer's list occur in Russian-Estonian bilingual data: language play, quotations, reiteration, topic shift, change in participant constellation, parentheses or side-comments, and expressive function.The different functions will be illustrated in what follows.As this study is qualitative, statistical analysis was not carried out.Here their occurrence will only be referred with categories "a lot", "often", "rarely".

Language play
The reasons for language play can be numerous.Through different examples it will be possible to show how puns impact on the process of code-switching.
In this switch type children use a lot of diminutives: they add to Estonian word in nominative Russian diminutive suffix.Excerpt 1 contains an example of such language play.

Excerpt 1
Estonian-language summer camp of Kohtla-Järve Third Secondary School (June 2000), primary school children (the Russian parts are given in italics and the Estonian parts in bold typeface): During an excursion to Toila [a small town near Kohtla-Järve], after a long walk in the park, one of the children says: In Estonian -ke(ne) derives diminutives mostly hypocoristically: pojake 'sonny' from poeg 'son'.In Russian the same function is served by suffix -čik.Children realize that adding of a Russian suffix to an Estonian stem is funny.They wish to mark codeswitched word, to emphasize it using language elements that add to code-switching some kind of pun.
In the second type of language play, trilingual sentences are produced.Excerpt 2 contains an example of such an activity.Here the teacher and children are talking.As we see in line 1, the teacher asks the question in monolingual Russian.In fact, even though the girl repeats it using elements of three languages -Russian, Estonian and English.

Sergej 'Sergej'
This trilingual question is directed to the co-pupils.The girl is sure, that it would be understandable.Code-switching takes place after the ambiguous lexical item on, meaning in Estonian 'is ', and English 'on'. Clyne (2003: 168) views such instances as lexical facilitation.

Quotation
PDF created with pdfFactory trial version www.pdffactory.com Direct quotations or reported speech occur often in the data.The language of the quotations is Estonian as it was in the situation described by the Russian-speaking child.
Excerpt 3 presents the clear example of quotation.Boys want to play football, but they can not find the ball (lines 1, 2, 3).They are looking for it everywhere.In line 4 the teacher helps, offering the potential location of the ball in Estonian.When asking for the result of searching in line 6, B1 answers in Estonian in line 7.

Excerpt 3
Summer camp of Kohtla-Järve Pärna School (June 2003).Pupils of the 3rd form: 'Do not cry!We will find them now!All will be OK, tip-top'

Topic shift
The nature of short interactions constraints to a certain extent the possibilities to detect this conversational function.Nevertheless, some examples do exist in the data.It can be claimed that children switch code when speaking about topics referred to their studies.
In the Excerpt 5 two boys are talking about half term marks (lines 1-4).Then topic shifts and they begin to speak about Christmas (line 5-7).

Excerpt 5
Kohtla-Järve Pärna School (December 2003).Pupils of the 4th form: Well and go then 'Let's go then' In line 5 B2 responds in Estonian.Thus, Estonian language functions as a tool for specifying addressee.But the conversation ends in Russian, language which was mainly used.

Parentheses or side-comment
Code-switching is also used to distinguish general talk and comment.Parentheses or side-comments are very close to this function.In many cases, the switched Estonian comment contains the same semantic information as Russian lexical item and the purpose of switch can be hardly identified.
PDF created with pdfFactory trial version www.pdffactory.com In Excerpt 7 two boys are talking about their mothers' birthdays.In line 3 sidecomment represent a well-known Estonian song "Valgeid roose" 'White roses' which a child could here during Estonian as a second language lesson.

Excerpt 7
Summer camp of Kohtla-Järve Pärna School (June 2004).Pupils of the 4th form: The whole conversation is in Russian.Again, this is one example of an instance of code-switching consisting of one word.

Conclusion
The object of the present research is multilingual communication among Russianspeaking schoolchildren with a special focus on Russian-Estonian code-switching in the town of Kohtla-Järve.
The data confirm Arnfast and Jørgensen's (2003)  Also it is suggested and proved that Russian-Estonian code-switching is used for expressive reasons too.The results of the present study may be summarized as follows: Most conversational functions are found in the data in connection with switches, whereas language play occurs most frequently.Unexpectedly, the expressive function is present in the data in to a great extent.Russian-speaking children of Kohtla-Järve emphasize their emotions through the use of two languages in the same discourse.Direct quotations or reported speech occur often in the data.The language of the quotations is often the same as it was in the situation described by the child.Thus, the quotations are usually in Estonian when describing a situation with an Estonian-speaking person.Code switches as reiterations are also often used for clarifying, emphasizing or amplifying a message.As this study deals with short conversations, switching when changing topic is PDF created with pdfFactory trial version www.pdffactory.comvery rare in the data.The nature of the setting (short interactions) restricts the possibility of detecting switching related to the change in participant constellation.However, some examples do appear in the data.Code-switching may also serve to mark side-comments.
This type of switch is rare but still observable in the data.Conversational functions that occur more often: language play, quotation, reiteration, expressive function.
In the quest of communicating meaning, children use different code-switching strategies.It could be suggested that Kohtla-Järve Russian-speaking children are well endowed with higher levels of creativity and the characteristic property of the environment pushing them to use the Estonian language in creative ways.Young Russian-speakers use Estonian words or phrases with the interlocutors to extend their communicative competence by using the resources of both languages.The main point here is that a significant number of instances of code-switching can be interpreted to reflect the child's developing communicative competence.In other words, it is a sign of resourcefulness.Their Russian-Estonian code-switching, like adult code-switching (see Verschik 2004Verschik a, 2006)), is not random and chaotic but systematic in accordance with the grammatical principles of the first language.Code-switching strategy works because all the children have some knowledge of Estonian.
The emergence of bilingual communication during the past decade was described in the paper.Taking this point to account, the main future research question is investigation of differences and similarities in code-switching among Russian-speaking children in other areas of Estonia, with special emphasis on its conversational functions and type.
PDF created with pdfFactory trial version www.pdffactory.com promotion of Russian language led to the maintenance of Russian monolingualism.The majority of the immigrants of non-Russian ethnicity preferred to communicate in Russian both at work and in social life.Expanding usage of Russian in administration and mass PDF created with pdfFactory trial version www.pdffactory.comcommunication put Estonians to speak Russian with Russian-speaking interlocutors.Only some of the post-WWII migrants could speak Estonian.
and also old-time inhabitants of Eastern Estonia, descendents of post-1917 emigrants from Russia.The group of indigenous Russians in Estonia numbers about 50 thousand, which is a tenth of Estonia's Russians.Thus, indigenous Russians of Estonia form a distinct minority group and it is correct to call this group Estonian Russians.At the same time, they are not able to shape the ethno-cultural character of the whole body of Russians among whom recent soviet-time migrants dominate.For instance, of the non-indigenous PDF created with pdfFactory trial version www.pdffactory.comRussians living in Estonia only up to 40 percent are born in Estonia (by 1989 census -38 %).According to the census of 2000 the percentage was 17 (it is important to note, that Belarusians, Ukrainians and other Russian-speaking inhabitants are not taken into consideration here).Almost 44% of Russian-speakers have some kind knowledge of Estonian (Estonian Statistics Office).In 1989 the percentage was 15.One can conclude that language situation has changed: there are more and more Estonian in everyday life.Thus, there are two groups of Russians in Estonia: indigenous Russians, or Estonian Russian, and non-indigenous Russians.Zabrodskaja's (2006 a) research on on Estonian object marking allow explicating the negotiation of novel object marking patterns through speech accommodation in the social network.The first results of the research indicated that the native Estonian speakers who communicate with Russians do show different patterns of variation from those Estonians who do not have such contacts.PDF created with pdfFactory trial version www.pdffactory.com Thus, increasing command of Estonian among second and third generation Russianspeakers, knowledge of Estonian as second language by young Russian-speaking children PDF created with pdfFactory trial version www.pdffactory.comand attitudes of Russian-speaking population towards Estonian may provide new developments of identities and new patterns of linguistic behaviour.The informants of the present study represent mostly the second and the third generation of Russian-speaking population living in Kohtla-Järve.They can be termed young Estonian Russians.They acquire Estonian mainly in the classroom.Young children in the process of learning two languages often use elements from both languages in the same utterance or stretch of conversation when they start speaking.This new generation of Russian-speakers becoming essentially Estonophone in contrast to their parents, who are Russophone.Direction of the dynamic aspects of code-switching seem to be moving to creating a new Estonian Russian language(variety).The children are being brought up in a dramatically different context than that of their parents: namely, that this generation is the first in which the dominant state language is Estonian.So, analyzing conversational patterns of Russian-Estonian code-switching, this study shows what happens to this generation of children's language use as they acquire their first (Russian) and second (Estonian) languages with the prestige values of each of the languages rearranged.
opinion that code-switching is possible even if learners have basic proficiency in second language.Numerous examples demonstrate how Russian-speaking children use different opportunities to incorporate what they know into their Russian speech.They are speakers of Estonian as second language.Comparison with the portrait as suggested by Franceschini (1998), the differences between immigrant and traditional minority context in Western Europe on the one hand and the Baltic countries, particularly Estonia, on the other become obvious.It PDF created with pdfFactory trial version www.pdffactory.commust be noted that the Baltic situation is utterly different from the rest of the post-Soviet context, too.Russian is one of the minority languages spoken in Estonia and it has had intensive contacts with Estonian, the majority language.By distinguishing themselves as an ethnic group, Russian-speakers draw a line between the two groups -notably, indigenous Russians, or Estonian Russians, and non-indigenous Russians.The pragmatic consequences of Russian-Estonian code-switching taking place among Kohtla-Järve nonindigenous Russian children are examined.At the present, Russians do not form a homogenous group; in fact, the Russians of Estonia are considered to be the most heterogeneous Russian population on the territory of the former Soviet Union.This paper focuses only on the conversational functions of code-switching.It demonstrates that the classical conversational functions are well represented in the data.

Table 1 .
Mother tongue of Kohtla-Järve enhabitants (males and females) PDF created with pdfFactory trial version www.pdffactory.comSuchnouns as jõulud 'Christmas', päkapikk 'gnome', piparid 'gingerbread', tore pidu 'good feast', uus aasta 'New Year' unquestionably have been acquired during the Estonian language lessons.They belong to Estonian context and culture.Lexical item piparid 'gingerbread' can be perceived as a good example of linguistic creativity.Code-switching can be used in situations where the certain speaker is connected with use of the Estonian language.Excerpt 6 describes the switches occurred in children's speech quit unexpectedly.After the camp the boys are going home.One of them is going to the library (see line 2).A new child interjects in a conversation in line 4.For him the term biblioteka 'library' comes from Estonian-related domain.He wants to know who is going there and in the question Estonian noun raamat 'book' is used.This may be caused by the fact that some librarians are dominant Estonian speakers.They tend to talk to children in Estonian or in Russian but still using a lot of Estonian words.
Expressive functionThe expressivity can make the speaker to use an appropriate word in the other language than mother tongue.Children wish to express different emotions from anger to sadness.Excerpt 8 contains an example of astonishment.Two boys are about to go home.They are packing their things into schoolbags.Suddenly one of them produces 25 crowns from his pocket.In line 1 the other boy begins to wonder.
does not astonish that the switch occurs.It seems for him to be quite natural.PDF created with pdfFactory trial version www.pdffactory.com5.1.7.PDF created with pdfFactory trial version www.pdffactory.com