Language Ecology in Cyprus , Sweden and Estonia : Bilingual Russian-Speaking Families in Multicultural Settings

We investigated language transmission in Russian-speaking families in multilingual settings in Cyprus, Estonia and Sweden. What they have in common is their Russian-language background and the minority status of their native language. In Cyprus and Sweden, participants mainly come from immigrant and mixedmarriage communities, while in Estonia they live in a bilingual society, where Estonian is a prestigious language and Russian has low status. To investigate the complex contexts of the informants’ language choices, the language ecology theory was chosen as a theoretical framework. Particular attention was paid to similarities and differences in the three country groups under investigation. Written questionnaires and oral sociolinguistic interviews were used for data collection among Russian-speaking informants in the three countries. We asked whether Russian as the first language was transmitted to the second generation and why. The attitudes towards bilingualism and Russian language transmission (including the change of these attitudes over time) – depending on the parents’ success in bringing up children bilingually – seemed to matter. Parental language choice is one of the main factors contributing to successful transmission. A lot depends on whether there was a desire for integration with the dominant language community, for staying isolated and only preserving the home language, or for having a balanced bilingual/multilingual approach and positive attitude towards both majority and minority languages. The socio-economic status, level of education and mother’s employment status played crucial roles in language transmission and attitudes. The linguistic repertoire of the father (minority, majority, or mixed) also had an effect.


Introduction
In this study we investigated the linguistic and sociolinguistic profiles of immigrant and minority communities in the three countries.In particular, our research was focused on the home languages of the members of these communities, and whether Russian as an L1 was maintained and transferred to the second and third generations and how.The study examined "bilingual" Russian-speaking families in Cyprus (which shares the Eastern Orthodox religion with Russia), in Estonia (part of the USSR until 1991) and Sweden (a more neutral country).In Cyprus, as well as in Sweden, Russian is spoken by a small minority group -immigrants or members of mixed marriage families -and may come under threat of extinction in the future in the host countries.In Estonia, Russian is a minority language, the former sociolinguistically dominant language, and still used as the L1 among almost one-third of the country's population.
The focus of this study was on the socio-linguistic ecology (Haugen, 1972(Haugen, /2001) ) of the Russian-speaking communities in Cyprus, Estonia and Sweden.More precisely, our exploratory study investigated Russian language maintenance and transmission, as well as the linguistic and cultural identities and family language policies (hereafter FLP) of Russian-speaking female informants in multilingual settings of the three countries Cyprus, Estonia and Sweden (King & Fogle, 2013).
According to Lo Bianco (2008), language transmission by parents is crucial for language maintenance and recovery.The success of language transmission depends on daily language use at home, attitudes towards language use and preservation, and efforts to create opportunities and incentives for language use in and outside the home (Laleko, 2013).As suggested by Spolsky (2012), language maintenance, particularly heritage language (hereafter HL) maintenance, is not solely a result of the influences of the linguistic environment, but largely depends on the orientations among the speakers themselves towards maintenance and transmission to the next generation.
In line with Kasuya (1998) and King et al. (2008), we implemented the sociolinguistic approach to FLP, which has become a new framework for examining parental language ideologies reflecting broader societal attitudes, ideologies and child language development.The data of the study consisted of parental interviews, qualitative observations and naturalistic recordings for better understanding of how parental language ideologies informed the application and realisation of FLP over time and their impact on child-language development, adding to our understanding of the relationships between FLP and child language outcomes (King, 2016).It is important to view the family as a dynamic system, taking into consideration a more diverse range of family types, languages and contexts.In line with previous research by Purkarthofer & Muni Toke (2016) and Busch (2016), the aim of the current study was to investigate FLP in multilingual, transcultural families, the parental expectations and strategies used for the construction of safe spaces for language transmission, challenges and support through such institutions as kindergartens and schools, and agents of normalisation.

Previous research
According to the ecological theory of human development (Bronfenbrenner, 1977;Gardiner & Kosmitzki, 2002;Timofeeva & Wold, 2012), the development of an individual is affected by his/her environment, interaction between individual micro-systems and macro-societal settings.Due to increased migration worldwide, there is a growth of immigration and cross-linguistic/crosscultural marriages.The linguistic repertoire of such families consists of both minority and majority languages.There is great variation concerning the initial language choice, motivation, emotional relationship, actual use of the HL and its transmission within the family, as well as cultural identity, which can lead to either active or passive, balanced or unbalanced bilingual development of the children (De Houwer, 2007;Pearson, 2007;Timofeeva & Wold, 2012).
Previous research has shown that parental choice of bilingual childrearing and motivation to transmit the HL depend on various reasons such as the need for communication with the extended family in L1 countries, a feeling of satisfaction and personal fulfilment in the society, emotion-related factors, and contact with other bilingual and bicultural families (Varro, 1988;Okita, 2002;Pavlenko, 2004).A lot of immigrant mothers do not have a job or any kind of achievement in the host country and they try to invest into their children and HL development.
Some families might not have enough motivation to maintain the L1 due to the fear of isolation and delay in the development of the majority language; consequently, they tend more to integrate with L2 society (Yamamoto, 2001).It is obvious that bilingual childrearing requires a lot of effort from the parents and it could be quite challenging (Okita, 2002;Timofeeva & Wold, 2012).It is important to pay attention to language proficiency in both the L1 and L2, and to quality and quantity of input as well as consistency of linguistic interactions and discourse strategies in the family (Döpke, 1992;Lanza, 1997;Kasuya, 1998;Takeuchi, 2006).Interactional styles can be with and without code-switching depending on whether the parents use explicit or implicit strategies.The first one is focused only on the use of the L1, while the second one allows the use of both minority 1 and majority languages with an emphasis on comprehension and an overall comfortable atmosphere of the communication (Goodz, 1989(Goodz, , 1994;;Kasuya, 1998).
It is essential not to neglect the role of the father in the family.The presence or absence of the father, his knowledge and use of minority and/or majority language, can affect the interaction within mother-child dyads and distance from mother-child unit (Lanza, 1997;Okita, 2002).Social and psychological factors, attitudes towards minority language and culture, the social and educational environment, the surroundings of the child, nuclear family language use and social networks play crucial roles in the process of minority language maintenance and transmission (Varro, 1988;Yamamoto, 2001;Okita, 2002).Extra-curricular activities, L1 weekend classes, visits to home countries of mothers, communication with grandparents and relatives can help to support the minority language development (Varro, 1988;Tannenbaum & Berkovich, 2005;Caldas, 2006;Takeuchi, 2006).

The sociolinguistic situation in Cyprus, Sweden and Estonia
Language ideologies are an important part of the historical dispositions that constitute multilingual behaviour.Language ideologies are intimately linked with language policy and planning, and particularly with implicit family language policies.Therefore, investigating the sociohistorical situation yields information about the attitudes and beliefs concerning speakers' choices of multilingual modes.In the following sections, we present the sociolinguistic background to our study, providing some brief descriptions of the historical contexts of the Russianspeaking community formation in the three settings.

Russian community in Cyprus
Since the 2000s, it appears that Russian has been emerging as a new lingua franca in the former USSR republics and abroad (Pavlenko, 2006).There is an increased valorisation of Russian in Cyprus due to tourist flow, immigration, international marriages, cultural and religious ties, military and political cooperation, investments and transnational corporations (Kuznetsov, 2010;Filippov, 2010).Russian is functioning as lingua franca in Cyprus and it is perceived as a commodity (Eracleous, 2015).
According to a Cypriot government census (2011), there are approximately 11 thousand people of Russian origin residing in Cyprus, nearly 50 percent of them living in Limassol.Most of them arrived in Cyprus in the 1990s due to Soviet and post-Soviet immigration.There are Russian schools, Orthodox churches, Russian-language television and a Russian radio service.Immigrants from the former USSR are a multilingual, multi-ethnic, and multicultural group of people.In Cyprus, they are all perceived as Russian, regardless of their ethnic roots, nationality, citizenship or multiple identities.The emerging role of Russian in Cyprus as one of the dominant and preferred foreign languages has led to some changes in the educational process and policy, as more and more Cypriots start learning Russian.The official and dominant language of Sweden is Swedish.Yet there is considerable linguistic diversity; over 150 languages are spoken in Sweden and many children grow up multilingual.Russians have never been a large immigrant group in the country.Russian, unlike Finnish, Turkish, Arabic, Assyrian, Persian, Spanish, Kurdish or Somali, is not a major immigrant language in Sweden.As stated in Parkval (2015, p. 276), today there are approximately 30,000 Russians in Sweden (0.30% of the population) and their number is increasing (ibid).It is difficult to count the number of Russians since many residents are listed as coming from the former Soviet Union.The Russian immigrant community is very heterogeneous in Sweden.

Russian-speaking community in Estonia
In Estonia, the collapse of the Soviet Union left a large Russian-speaking minority, which shares a number of common features, but for whom the ethnolinguistic outcomes are remarkably different (Ehala & Zabrodskaja, 2014;Zabrodskaja, 2015).After Estonia regained its independence in 1991, standard Estonian became the single official language, which was to be used and accepted at all levels of society, according to the Language Act of 1995/2011. 2Competence in Estonian became heavily connected to access to higher education and professional career opportunities.Education in Russian is provided at the basic school level only.

Aims, scope and methodology of the study
The main goal of this paper is to describe and theoretically interpret the Russian/language ecology in Cyprus, Sweden and Estonia.This general goal is divided into three research questions: RQ1: Are there differences and/or similarities among the countries in terms of Russian-language use, maintenance and transmission?RQ2: Which factors affect FLP in Cyprus, Estonia and Sweden?
RQ3: Does minority vs. immigrant context affect the linguistic and cultural identity of the heritage speakers?
To address these questions both quantitative and qualitative methods were used, and we introduce them next in detail.

Methodology
This study implemented parental written questionnaires with the focus on general background, socio-economic status, language proficiency, language maintenance, use and transmission, as well as the linguistic and cultural identity of the participants (Otwinowska-Kasztelanic & Karpava, 2015).The questionnaire had general and specific, open-ended and multiple-choice questions.The informants had ample time to fill it in.The participants were recruited using the snowball technique as it was the most efficient way to access members of the Russian communities in Cyprus, Estonia and Sweden.Oral data were collected with the help of semi-structured interviews (Ringblom, Zabrodskaja & Karpava, 2015).The interviews took place only after the participants had completed the questionnaires.The participants were asked to reflect on their life experience in L2 countries, and to share perceptions of their diverse language and cultural identities.They provided in-depth information on Russian language use, maintenance and transmission to the next generation.Both probing and interpretative questions for deep inquiries and clarification were used (Kvale, 1996).The researchers had an interview guide in order to follow the structure of the interview and avoid deviation from the topic of the research.The duration of the interviews varied from 30 minutes to 2 hours.The focus was on FLP activities in 62 multilingual families in the three different cultural and linguistic environments of Cyprus, Estonia and Sweden.The grounded theory research method was implemented regarding data collection, coding strategies, analysis and interpretation (Glaser & Strauss, 1967;Glaser, 1978;Corbin & Strauss, 1990;Willig, 2008).The participants were audiorecorded; then the oral data was transcribed, translated into English and thematically coded.

Participants
The data was collected from 62 Russian-speaking mothers residing in Cyprus ( 27), Sweden (20) and Estonia (15).Most of the Estonian participants (94%) were born in Estonia; the Cypriot participants -mainly in Russia (61%), but also in Belarus (19%), Moldova (4%) or Ukraine (16%), while the Swedish participants -in Russia (20%), Ukraine (25%), Belarus (4%), Azerbaijan (10%), Kyrgyzstan (2%), Poland (4%) and other republics of the former Soviet Union (35%).The age range of the participants of the three countries was from 20 to 65 years old.It should be noted that the participants from Estonia were younger than the participants from Cyprus and Sweden; see Table 1.The participants differed with respect to the mean length of residence in the host country.The mean length of residence in Estonia was over 30 years, in Cyprus 10 years, and in Sweden around 15 years.The Estonian participants had spent all of their lives in Estonia (100%).The participants in all three countries had high level of education, see Table 2.

Results
The analysis of the data, both questionnaires and interviews, showed that the Cypriot participants identified themselves mainly with Russian/the home language, whereas the Estonian and Swedish participants -mostly with both Russian and Estonian/Swedish languages, see Table 3.
Language identity is shaped under the influence of various factors.It is obvious that living in different countries, geo-political, sociolinguistic and sociocultural settings L1 Russian speakers have distinct linguistic behaviour, attitudes and perceptions.In Cyprus, the respondents have predominantly L1 Russian monolingual rather than multilingual or bilingual language identity, in Sweden -more bilingual or multilingual, while in Estonia -bilingual and only to a lesser extent L1 Russian monolingual.Examples (1) and ( 2) are excerpts from interviews with Cyprus participants regarding their use of Russian and Greek in the host country.
( Some Sweden informants, on the other hand, were oriented towards integration with Swedish society and were engaged in literacy practices in Swedish as well. They reported a higher proportion of sending emails, SMS, reading books, listening to the radio and watching This mother has very limited opportunities of using Swedish and thus sees her child as an extra opportunity of practicing Swedish with him/her.It was found that the Cypriot participants identified themselves mainly with Russian culture; the Estonian participants -either with Estonian or mixed (Estonian-Russian) culture, while Swedish participants -mostly with both L1 and L2 country cultures.It should be noted that only in immigrant settings, Cyprus and Sweden, were there evasive answers concerning cultural identity see Table 4.
Cultural identity is closely related to language identity.Only in Estonia, one third of the population under investigation have L2 majority cultural identity; in Sweden -mainly mixed (Russian-Swedish), while in Cyprus -predominantly monolingual L1 Russian cultural identity.Cypriot participants prefer to use Russian, the minority language, in order to talk about their families, while the Estonian and the Swedish participants -either the majority or minority language.English seems to play an important role in all three countries see Table 5.
The analysis of the data showed that the Cypriot participants used Russian (38%), English (15%) or Greek (37%) in order to talk to their partners.The Estonian participants spoke more Russian (60%) than Estonian (40%), while the Sweden participants -more Swedish (38%) or Russian (37%) than English (8%) or Ukrainian (7%).Immigrants in Cyprus and Sweden, members of mixedmarriage families, have a nearly balanced distribution of L1 and L2 usage.This can be explained by the fact that their partners mainly do not know Russian, thus English is used quite often as a bridge for communication, especially during the first years of residence in L2 country.
The Cypriot participants choose mainly Russian rather than Greek or English in order to talk to their children; the Estonian participants -more Russian or Estonian, while the Swedish participants -either Russian or Swedish, see Table 6.Despite the fact that L1 Russian mothers tend to integrate into L2 society and realise the importance of English as an international language for communication, the data shows that the participants in all three countries   Interestingly, the Cypriot participants use more English than Russian or Greek in order to communicate at work; the Estonian participants -Russian or Estonian, while the Swedish participants -Swedish, see Table 8.Language choice in the professional sphere mainly depends on the economic and political situation in the country, practical reasons and the needs of the employer rather than on subjective and emotion related factors.One of the crucial aspects of this paper is that it compares language ecology in minority and immigrant settings.It was found that the Cypriot participants had come to Cyprus for various reasons, such as family (44%), finance (18%), politics (16%), work (15%) and other reasons (7%).The Estonian participants had lived in Estonia all their lives, while the Swedish participants had come to Sweden due to family (63%), work (17%), politics (16%) and other reasons (4%).The analysis of the data showed that the Cypriot participants have friends who mainly speak different languages; the Estonian participants -those who speak different languages and those who speak both languages, while the Swedish participants -mainly those who speak different languages, see Table 9.There is a similarity between Cyprus and Sweden as Russian women migrated to these countries mainly in order to create a mixed-marriage family and/or to find better career opportunities in a more stable socio-economical/political environment.
In Cyprus (88.88%) and in Sweden (100%), Russian speakers consider it necessary to learn the majority language, compared to 60% in Estonia.For immigrants, the L2 is an essential 'survival tool' in an L2 country.Without knowledge of the L2, it is not possible to function in L2 society (e.g. to find a decent job, to communicate with your colleagues and friends, to raise children).In Estonia, the situation is different, as everybody knows two languages, Russian and Estonian, though nowadays after the dissolution of the   USSR only the latter is the official language, with a high level of prestige and a vital role in the society.
Russian speakers in Cyprus position themselves as full members of Cypriot society, with equal rights, or claim that they belong to both Cypriot and Russian societies.The Estonian and the Swedish participants mostly feel that they are full members of Estonian/Swedish societies, with equal rights, see Table 10.However, in all three countries, there is also a tendency for alienation and isolation, as some of the participants, with the highest percentage in Sweden, stated that they belong to neither society.Nearly one third of the respondents in Cyprus and Estonia, feel that there is no integration with L2 society.The way they perceive themselves in the L2 country is reflected in their linguistic behaviour, their stance towards their heritage language and family language policy.
The analysis of the questionnaires and interviews revealed that both the Cypriot (82%) and the Swedish (78%) societies are perceived to be tolerant of nonnative speakers, whereas the Estonian society -is mostly intolerant (58%) of non-native speakers of Estonian.
The data shows that in Cyprus and Sweden, L1 Russian immigrants have favourable conditions in order to use their native language in their social network, develop Russian-speaking communities, maintain and transmit the HL to their children.In Estonia, with an overall negative attitude towards Russia, the Russian language and culture, it is difficult to revitalize and maintain the minority language.
However, even though the self-reported data depicted a positive, non-discriminative atmosphere in Sweden, it was found that the participants in this country had the highest percentage of those who had been told to stop speaking their native language (40%) in comparison to Cyprus (20%) and Estonia (24%).Besides, the Swedish participants had the highest percentage of those who had heard about discrimination against multilingualism in the host country (35%) in comparison to Estonia (20%) and Cyprus (11.11%).Even though all three countries under investigation are members of the European Union that promotes 'unity in diversity', being a non-native speaker in an L2 country, either in a minority or an immigrant setting, seems to be quite a challenge.Russophone residents in the host countries need to find their way of how to adapt to and integrate into L2 societies, to overcome potential conflicts between majority and minority communities, to preserve and develop their heritage language, culture and religion, to secure their well-being, social organization and education.
Code-switching seems to be one of the strategies of communication adopted by Russophone inhabitants, though it is a more frequent phenomenon in Cyprus (60%) and Sweden (58%) than in Estonia (36%).Living in the host country presupposes everyday interaction in private and professional spheres, neighbourhoods and service sectors.The participants in Cyprus (77.79%),Estonia (53.24%) and Sweden (60%) mainly never feel that their neighbours have bad attitudes towards them because they speak another language.Most of the participants in all three countries (Cyprus: 74.07%; Estonia: 53.34%; Sweden: 60%) stated that professionals (doctors, nurses, teachers and clerks) never had bad attitudes towards them because they speak other languages.It is obvious though that the percentages in each country are different, with the highest in Cyprus and the lowest in Estonia, which allows us to assume that in Estonia, L1 Russian mothers experience more negativity due to their L1 background in comparison to Cyprus and Sweden.Therefore, their task of HL maintenance and transmission becomes even more challenging.
It is not surprising that nearly all of the children of the Cypriot participants (96.29%) speak and comprehend Russian, which is not the case in Estonia (46.68%) and Sweden (55%).In Cyprus, nearly all of the children can read and write in Russian (88%), while in Estonia (24%) and in Sweden (38%), the percentage is lower.The Cypriot participants use only Russian at home (100%), while in Estonia (60%) and Sweden (60%) only about half of the participants do so.The sociolinguistic, economic and political situation in the country, overall atmosphere in the society and attitudes towards the minority community, government support and effective language policy do affect the development of heritage language literacy.The data shows that L1 Russian residents in the countries under investigation have different opportunities in order develop the Russian literacy skills of their children.
The motivation of the mothers to maintain L1 Russian should be facilitated not only at the family level, but also outside, in L2 society, by teachers and professionals.Our research project revealed that the Cypriot participants have never been advised at school to stop speaking Russian with their children (100%), whereas in Estonia and Sweden this was true of only 60% and 64% of the participants, respectively.In Cyprus, nearly all of the participants (99%) do not think that their child is discriminated against at school because s/he speaks Russian, which is not the case in Estonia (44%) or Sweden (62%).Definitely, more support of the HL would benefit the younger generation of the Russian-speaking population in the host countries, especially in Estonia and Sweden.The Cypriot participants mainly often or very often (80%) tried to teach their children the Russian language, while in Estonia most of the parents (58%) were hesitant to answer; in Sweden, many participants tended to avoid answering (30%), while the other participants said they taught their children Russian often or very often (30%), sometimes (22%), or never (18%).The participants in all three countries were not unanimous as to whether they insisted that their children use Russian at home and outside.In Cyprus, the most frequent answers were never (38%) or often (46%); in Estonia (45%) and Sweden (42%) there were mainly evasive answers.Family language policy depends on the language policy and tolerance in the country as well as on motivation and willingness of the parents to put effort into their children's development regarding majority and minority languages and literacy skills.

Discussion
Our study investigated Russian-language maintenance and transmission in minority settings outside Russia, contributing to the field of language ecology.The first research question concerned the differences and similarities among the countries in terms of Russian language use, maintenance and transmission.Among the three countries under investigation, Cyprus had the most favourable conditions for HL maintenance and transmission as the status of the Russian language in Cyprus is high.Estonia was the least favourable due to the socio-political situation in the country, the low status of Russian in the country and the general negative attitude towards Russia and Russian-speaking countries.In Sweden, there was a general tolerance of multilingualism, although more support was needed to maintain the Russian language.The socio-political situation in the target country was reflected in the linguistic behaviour and preferences of the participants.
The second research question concerned the factors that affected FLP in Cyprus, Estonia and Sweden.The level of education of the participants, length of residence, as well as their socio-economic and professional status in L2 country affected their willingness and opportunities for use, maintenance and transmission of their HL to the next generation.As far as CS is concerned, the participants tended to mix languages or code-switch at work; at home, they either mixed languages or used the OPOL strategy (one parent, one language).Our data show that not all parents followed the OPOL strategy.Mixed language and cultural identity led to mixing languages, while the preference for the L1 culture triggered the use of the L1.Instrumental and integrative motivation was an important factor for language use by parents and children.However, each family seemed to be unique.Thus, FLP depended on the individual life trajectories, socio-economic background, exposure to linguistic and cultural resources, and the preferences and practices of each family.The interaction between external factors (e.g.language policy and attitudes of the host societies) and internal factors (e.g.ideology, identity, individual motivation, agency and cultural awareness) shaped FLP in the immigrant and minority communities.
Regarding the third research question, the status of the minority/immigrant language in the target countries, and the willingness to integrate or stay isolated affected language and culture identity.The lack of job opportunities prevented some of the participants from integrating into L2 society (especially in Cyprus: an inclusive society with multilingual practices), which could lead to the increased use of the L1.However, most of the participants tried to learn the L2 due to integrative and instrumental motivation, but not all felt that they had fully integrated into the L2 society; some of them had mixed identities.Overall, the Russian women in Cyprus (as well as in Estonia) were well-educated; most of them had university degrees.Although they had had good jobs in their Russian-speaking countries before coming to Cyprus (the same was also true for the Sweden participants), they were mainly unemployed housewives in the host country.
Dominant societal discourses affected the formation of the linguistic and cultural identities of the Russianspeakers through their lifespan development in the host countries.Their identities were negotiated in such contexts as home, work and school (for children).It also seemed that the participants in minority environments (Estonia) were not willing to discuss issues of discrimination.They might not have felt secure in the country, with its dominant Estonian and minority, low-status, low-prestige Russian situation.
Overall, the informants from the three countries identified themselves with Russian, but the majority of them used three languages in various domains.In Cyprus, the participants mostly used Russian with their families and children, in contrast with Sweden and Estonia, where the majority and minority languages were often used interchangeably.Most of the participants felt the need to learn the mainstream language for various reasons, mainly to integrate into the dominant society.Opinions varied regarding how they felt in the country.Some of them believed that they were fully integrated into the mainstream society, and some not.
It seems that the population of the host countries is tolerant of people who speak languages other than the official one.Only a few of the participants had been told to stop using their languages.Most of the participants had never heard about discrimination (though many Estonian participants preferred to leave this question unanswered).The participants had never experienced discrimination at work, nor did they experience any bad attitudes from their neighbours based on the linguistic factor.Only in a few cases had they experienced bad attitudes towards them from doctors, nurses or teachers.Nearly all of the children had never been discriminated against at school for speaking Russian.Few participants had been advised to stop speaking Russian at home.
Although bilingualism was usually the goal of the parents for their children, the participants often switched codes at home and their work places.Nearly all of their children could read and write in Russian.Some of them refused to use Russian.Siblings tended to speak the mainstream language among themselves and Russian with their parents and relatives.Not all of the participants were satisfied with their children's level of Russian (cf."kitchen Russian", Pavlenko & Malt, 2011).The participants either used the OPOL approach or mixed both languages while communicating with their children, while a very small group of participants used only Russian.
In Cyprus, nearly all of the children attended extracurricular classes and activities, mainly language courses or sport activities.Greek, English and Russian were used there.Parents tried to teach their children Russian.Most of the children attended classes where they learned how to read and write in Russian.In Sweden and Estonia, many families were concerned with their children's proficiency in the majority language and tended to focus only on it, rather than teaching them Russian.Some mothers tended to apply a translation strategy when speaking Russian to their children, thus using the dominant language as a cognitive base for teaching them Russian.In the long run, this strategy did not seem to be very effective since the children stopped seeing any need to learn Russian when they realised that they could communicate with their mothers in the socio-linguistically more powerful language.
Another factor that could lead to the child's monolingualism in the mainstream language in all three countries was when parents in inter-marriage couples did not clearly separate the spheres of their two languages, not making it clear to the child which was the mother's and which was the father's language.In Sweden, the country that provided mother tongue instruction in Russian, the parents mentioned several implementation problems with the provided instruction, regarding the quality and quantity of it.In Estonia, the Russian-speaking parents were concerned that the maintenance of the Russian language among the youngest generations was not assured via the high school system.The children often saw the heritage language as useless, both educationally and economically.The majority of the participants never insisted that their children use Russian at home or outside, but they often insisted that their children take part in activities related to Russian culture.Overall, most parents felt that it was their responsibility to bring up their children bilingually.

Conclusions
Our results show both differences and similarities among Russian-speakers in the three countries, not only in their family language practices, but also in their attitudes towards Russian-language literacy.This was an exploratory study, comparing three different settings, and involving minority groups and immigration.The comparative analysis of the data in the three countries showed that there were similarities between Cyprus and Sweden with respect to the immigration environment, reasons for moving to the host country and the level of education, though there were also some differences.Most of the participants in Cyprus were unemployed, whereas in Sweden they were middle-class workers.The participants in these countries were members of mixed-marriage families and used the OPOL strategy with their children.They identified themselves with both cultures.Both countries seemed to be tolerant of the minority language.However, though Sweden is tolerant of multilingualism, the tolerance of the country did not automatically lead to minority language transmission.In addition, extracurricular activities were mostly in Swedish in Sweden and the parents chose Swedish schools and preschools for their children.The Russian-speaking informants born in Estonia lived in minority group neighbourhoods.Only in Estonia did bilingual and bicultural mothers live in an officially monolingual country where an official ideology of one nation -one language prevailed.Only in Estonia did the participants provide evasive answers concerning questions regarding discrimination.
Many parents spoke Greek/Swedish/Estonian for various pragmatic and practical reasons that need to be explored further.Some parents perceived the mainstream language as more advantageous from the point of view of the child's future or encountered too many obstacles in their efforts to speak Russian.The preliminary results indicate that the integration of the mothers into the societies and their bilingualism may have had a negative effect on the Russian language transmission to the children.Working full time and spending less time with the children after the age of eighteen months, when the child enters day care, was another factor."Not enough input" was often the answer given.Overall, it was found that multilingualism and the maintenance of the Russian language and culture were usually encouraged, and parents often chose the OPOL approach at home.However, not all of the efforts resulted in successful home language transmission.Intergenerational language transmission was not onesided: children's personalities, language environments and real-life experiences shaped their own linguistic practices and choices.Both perspectives must be taken into consideration if we want to explore why some children refuse to speak Russian.
The fact that young children tended to have a preference for the majority language is not new.Children's language choices influenced the language choices of their parents, which in turn changed the language patterns.The parents switched to the dominant language to accommodate to the language choice of the children, which led to passive bilingualism, sometimes resulting in the child's monolingualism and/or changing attitudes towards a bilingual upbringing and language transmission (cf.Festinger, 1957 and the theory of cognitive dissonance).Several parents mentioned that the literature on bilingualism they had been exposed to described bilingual acquisition as a natural process, which they felt was far from the case in real life.In our further comparative studies, we need to examine in more detail situations in which parents intend to raise their children as bilingual but do not succeed, in order to fully understand the complexity of this phenomenon.

Notes
1 The term minority language as it is used here does not mean that Russian has official minority status in the three countries and may thus be used interchangeably with the term HL. 2 https://www.riigiteataja.ee/en/eli/512012016001/consolide.
2.2.2.The Russian community in Sweden and the linguistic situation in the country

Table 1 :
Age of the participants.

Table 2 :
Level of education of the participants.

Table 3 :
Language identity of the participants.

Table 4 :
Cultural identity of the participants.

Table 5 :
Which languages do you use to talk about your family?

Table 6 :
Which languages do you use to talk to your children?Cypriot participants tend to communicate more in Russian than Greek or English with their friends; the Estonian participants -either Russian or Estonian rather than English, while the Sweden participants -either Russian or Swedish rather than English, see Table7.The social network of the speakers influences their linguistic repertoire that more or less consists of L1 >L2 > English.

Table 7 :
Which languages do you use to talk to your friends?

Table 8 :
Which languages do you use to communicate at work?

Table 9 :
What kind of friends do you have in L2 country?

Table 10 :
How do you feel in the L2 country?