“Talk in Tamil!” – Does Sri Lankan Tamil onward migration from Europe influence Tamil language maintenance in the UK?

1 Introduction

Migration has profound sociolinguistic consequences (Kerswill 2006). One such consequence being debated is the fate of languages that migrants bring along into the host communities (Cavallaro 2010). In recent years there has been emerging research that critiques the conceptualisation of international migration as a simple bipolar event – a move from country A to country B. “Globalisation and globally networked migration trajectories give rise to a plurality of migration pathways which subvert the linear model of origin-destination” (Puppa and King 2019: 1937). Increasingly migrants have chosen to move on to other countries instead of settling down in the first country or returning to their countries of origin. This has been termed by scholars in the migration literature as “onward migration” (Puppa and King 2019), “transit migration” (Collyer and de Haas 2012), “secondary migration” (Bang Nielsen 2004), “transmigration” (Ossman 2004) or “twice-migration” (Bhachu 1985). Although this migration pattern is observed to be on the rise, few empirical studies have described this phenomenon (Puppa and King 2019: 1937). Preliminary research suggests that among certain groups of new citizens in the EU, there may be a significant emerging trend toward relocation within the EU (Lindley and Van Hear 2007: 2). Lindley and Van Hear (2007: 2) define “new citizens” as people who had claimed asylum in the EU and who were granted refugee or some other form of humanitarian or temporary protection. They subsequently gained citizenship of their host state and became new citizens. Scholars claim that relocation movement forms a new strand in EU mobility that is poorly understood. In this article I explore the different migration trajectories of the Sri Lankan Tamils (SLTs), focusing on onward movement, i.e. the movement of SLT new citizens from the EU to the UK. Various patterns of migration have given rise to approximately 180,000 SLTs in the UK, including those born in the UK of SLT parents (Walton 2015: 962). Rutter (2015: 159), however, posits that the number is “likely to be about 200,000 persons”. I argue that it is only by examining the relationship between SLT migration patterns and the communicative resources SLTs bring with them can we have a better understanding of issues relating to diaspora re-grouping, immigrant integration and transnational relations (Lindley and Van Hear 2007: 2).

To do this, I use the well-established linguistic anthropological concept, “communicative repertoire” (Blommaert and Backus 2011; Blommaert and Rampton 2011: 6–7; Rymes 2012, 2014). This perspective shifts attention away from the idea of languages and speakers as localised and quantifiable speech communities to transnational communities, whose individual communicative repertoires include sets of language resources to draw from. These resources can vary from expert competence in a wide range of genres, registers and styles in a named language, to specific “bits” of language and literacy variables (Blommaert 2010: 8). In addition, one's communicative repertoire also includes other means of communication, such as gesture, dress, posture, knowledge of communicative routines, familiarity with types of food or drink, and mass media references including phrases, dance moves, and recognisable intonation patterns that circulate via actors, musicians, and other superstars (Rymes 2012). Thus, “communicative repertoire” has increasingly become a way to describe how individuals deploy their multiple language resources and other modes of communication in interactional contexts (Rymes 2014).

In the next few sections, I (i) address the debates around heritage language (HL) maintenance, focusing specifically on discussions around Tamil language varieties and Tamil maintenance in the SLT diaspora; (ii) describe the changes in SLT migratory trajectories to the UK; (iii) discuss how the SLTs' changing migration patterns link to a change in Tamil practices in the diaspora. I use data gathered from a research project carried out with the SLT diaspora in London between 2015 and 2018 to argue that the increase in SLT onward migration leads to a move toward Tamil language maintenance in the UK.

2 Language shift and maintenance

Fishman (1991) has defined language shift as a threat to native languages resulting from fewer users, while Mesthrie et al. (2001: 253) describe it as “the replacement of one language by another as the primary means of communication and socialisation within a community”. The studies on language shift have been extensive (Lee 2013; Pauwels 2005; Sofu 2009) and demonstrate that language shift amongst diaspora communities is not extraordinary in itself (Báez 2013: 28). What is significant, and which researchers have only just begun to consider, is how individualised trajectories of people's movement shapes sociolinguistic transformation (King and Haboud 2010; Wyman 2013). Research carried out with the SLT diaspora helps to illustrate this.

Canagarajah (2008) conducted a project which studied language shift in the SLT diaspora living in English-speaking host communities. His research provides an ethnographic perspective on the role of the family in preserving the HL, which Fishman and Tej's (2004) Graded Intergenerational Disruption Scale (GIDS) emphasizes as being crucial for the maintenance of HLs. Canagarajah highlighted that “community members are troubled by a distinct language shift toward English and a dramatic decline of Tamil in the diaspora in the USA, UK and Canada” and discussed how “some members of the community see the negligence of the family as the chief factor in the decline of Tamil” (2008: 145). While his study did not aim to test the GIDS model to investigate whether or not the family is the critical social unit to help a community preserve its HL, its objective was to explore the dilemmas and challenges for families in transmitting Tamil to their children. In doing so, Canagarajah found that there is very rapid language attrition in SLT diasporic communities, surpassing the typical immigrant pattern of language shift that spans three generations (2008: 151).

Interestingly, Canagarajah's (2012, 2013, 2019) analysis of HL use in contexts of language shift shows that SLT youth use a range of codes – Tamil, English and sometimes Sinhala when communicating – and employ language practices that use bits and pieces of these codes: (i) code-switching into Tamil from English; (ii) emblematic use of Tamil for symbolic reasons and performative acts; (iii) switching to the use of Sri Lankan English varieties; (iv) using words borrowed from Sinhala; (v) showing receptive competence in Tamil which enables them to respond in English; (vi) performing ritualized practices of communication where they can participate in communicative events through multimodal resources (Canagarajah 2013). Compatible with the “communicative repertoire” concept, he uses his findings to argue for a practice-based orientation to HL, as migrant families relate to HL differently and consider diverse non-traditional verbal resources and pragmatic strategies as indexing HL (2019: 13). Thus, his analysis supports the view that where there is language shift, HL has to change to accommodate new identities and repertoires of migrant communities (2019: 43).

Another account which explores migration and social integration of Tamil-speaking migrants is provided by Das (2016), who considers the different migration cohorts of Tamil speakers to Montréal, Québec during the late twentieth and early twenty-first centuries. She discusses the linguistic rivalries between the elite Tamil Brahmins from India and the non-elite Jaffna Tamils from Sri Lanka, who identify with different spoken and written varieties of Tamil to justify their self-segregation and alternative prospects for social mobility. Colloquial or Spoken Tamil is considered the more hybridised and impure register and Indian Tamil Brahmins are regarded as embodying the modernising virtues of this variety. The purist virtues and classical provenance of literary Tamil or Written Tamil, on the other hand, are considered to be personified in the Jaffna Tamil refugees who, in their search for a safe and secure homeland, try to reinforce the primordialist version of Tamil Eelam (proposed independent Tamil state) using this supposedly older and purer register. Das further draws attention to the different generations of Tamils: the first generation Indians and Sri Lankans who display a lesser competence in French and English than in Tamil and the more English-and French-savvy Canadian-raised or born second generation Tamils, who have variable expertise in speaking and writing in Tamil. Das observes that this second generation practice a lot of code-switching and code-mixing, where Tamil and French may be mixed with English to increase the former languages' prestige. However, they rarely mix Tamil and French together because to do so would invite the contempt of community leaders and government officials invested in maintaining the idea of separate Indian and Sri Lankan Tamil, and anglophone and francophone worlds. Das's research thus illustrates how Tamil-speaking migrants must navigate the linguistic rivalries shaping the politics of French- and English-speaking Canada and the Tamil-speaking diaspora. Her work highlights that Tamil language use and maintenance need to be understood within a context of complex multilingual repertoires and migration trajectories.

The above studies demonstrate how migration plays an important role in shaping communicative repertoires in the diaspora. Changes in diasporic migration patterns would undoubtedly have an impact on HL communicative practices, which makes it imperative that they are investigated. Currently there is “comparative neglect of […] older youth and adults […] in more heterogeneous, multilingual and transnational contexts” (Duff and Talmy 2011: 101) and very little research on the theme of onward migration (Lindley and Van Hear 2007: 1–2; Puppa and King 2019: 1948). Blommaert et al. (2017) thus recommend that more ethnographic work should be carried out within migrant communities where “a fresh uptake in the investigation of social situations that either involve or are a by-product of communicative interactions” can help identify the complexities which “we ought to study” (Blommaert et al. 2017: 352). This would involve more research on how onward migration processes could challenge dominant narratives about language shift.

3 Sri Lankan Tamil migratory trajectories to the UK

The SLT diaspora is said to be formed as a direct result of the political conflict in Sri Lanka (Sriskandarajah 2005: 494). In the years after Sri Lankan independence, discrimination against the Tamils became progressively worse as rivalry between the Tamil minority and the Sinhalese majority intensified, causing SLTs to migrate in large numbers (Canagarajah 2008: 153). Early waves of SLT migrants were mainly professionals, students and elites (Daniel 1996). This was followed by the largest exodus of Tamils from Sri Lanka, triggered by the onset of the civil war beginning in 1983.

The Sri Lankan civil war lasted 26 years and contributed to the formation of a substantial SLT diaspora of between 1 and 1.5 million people scattered across the globe (Brun and Van Hear 2012; David 2012; Venugopal 2006). Sriskandarajah posits that it is likely that one in every four SLTs now lives outside Sri Lanka (2005: 494).

According to the UNHCR (2001), 256,307 SLTs applied for asylum in Europe between 1980 and 1999; and between 1984 and 2004 a total of 49,545 asylum applications to the UK were made by Sri Lankans, nearly all of whom were likely to have been Tamils (Aspinall 2019: 54). Eventually, the SLT's increased presence in the UK led to a pattern of chain migration. MacDonald and MacDonald (1964: 82) define chain migration as “that movement in which prospective migrants learn of opportunities, are provided with transportation, and have initial accommodation and employment arranged by means of primary social relationships with previous migrants”. In other words, chain migration entails a tendency by foreigners from a certain city or region to migrate to the same areas that others from their city or region have settled in (Tovares and Kamwangamalu 2017: 216). Bloch's (2002) sociological study of different refugee groups in London showed that virtually all the Tamil respondents gave their reason for coming to Newham (London) as having kinship or friendship ties, or because of the presence of places of worship. David (2012: 377) also claims that “it is the strength of the transnational connections of these more settled Tamil migrants in London boroughs such as Newham, in the east of the city, or Brent in north-west London, which draws even greater numbers of Tamils, seeking asylum, searching for employment and for places of refuge”. This pattern of chain migration thus gradually led to an established SLT community in various locations in the UK. By 2002, it was estimated that 110,000 SLTs lived in the UK, including 60,000 refugees (Zunzer 2004).

However, with the increasing restrictions introduced in the mid-1980s by the British government, explicitly to curtail SLT asylum migration to the UK, SLT asylum migration to countries in continental Europe tended to gather momentum in the late 1980s and 1990s (Pirouet 2001). After living and forming families in European states for a decade or more, waiting for their appeal for refugee status to go through, SLTs eventually became recognised as refugees and acquired citizenship of EU member states. This enabled them to move within the EU, and a large number relocated to the UK (Lindley and Van Hear 2007: 3, 12). Thus, in the UK, there was a decline in asylum migration from Sri Lanka and an increase in onward migration.

Unfortunately, “the scale of this onward migration [from Europe to the UK] is unclear since new arrivals will be recorded as EU nationals rather than as Tamils, but anecdotally, according to the Tamil community leaders and some local government workers, it is significant” (Lindley and Van Hear 2007: 13). One notable feature of the cases investigated so far is the short period of time between securing a passport allowing mobility within the EU and the actual move to the UK, which seems to indicate that perhaps the idea of such a move was already well established (Lindley and Van Hear 2007: 17). Research has shown that this relocation to the UK has been mainly motivated by the “greater perceived economic opportunities, for reasons of education and language, to re-group with family and friends, and because of the greater critical mass of Tamils in parts of the UK and therefore the greater possibility to lead a familiar^[1] life” (Lindley and Van Hear 2007: 17). However, as mentioned previously, there are no statistics to indicate how many SLTs moved from the EU to the UK as onward migrants. One of the reasons for this is that the UK census collects information on country of birth but not nationality. Thus, it can be difficult to obtain information from census sources on, for instance, Sri Lankan-born French nationals living in the UK (Lindley and Van Hear 2007: 4).

4 Methodology

Many Tamils arriving in the UK “appear to have settled in areas near to the core areas of Tamil settlement in London” (Lindley and Van Hear 2007: 13). Thus, my fieldwork was mainly carried out in London, in the boroughs of Harrow, Ealing, Redbridge and Kingston-upon-Thames, where there are concentrated Tamil populations.

4.1 Participants

Participants were selected through snowball sampling from my own family and friendship networks. Since my origins are from Chennai, South India, the Tamil variety I speak is influenced by regionally dominant languages, such as English, Hindi and Telugu (Annamalai 2011), and is considered highly colloquial. This immediately alerted my participants to the fact that although I have connections to the SLT community through my partner's family, I do not directly belong to it. My position in this respect allowed me to preserve my “participant-as-observer” role (Gold 1958).

There were 42 participants altogether and they represented a range of migration histories (as spouses, students, refugees, elderly parents; arriving solo or with family from Sri Lanka and Europe). They also varied in terms of their length of stay in the UK; i.e. newcomers who had lived in the UK for less than 10 years, first generation SLTs who had been living in the UK for at least 25 years and second generation Tamils who had either been born in the UK or who had arrived as children.

Those who arrived in Western countries as adults were considered generation 1^[2] (N = 28). Those who arrived as children below the age of five or who were born in the West were considered generation 2 (N = 13). One participant was of generation 1.5 as she was born in Sri Lanka but migrated with her mother to join her father in the UK when she was 10 years old. The term 1.5 generation is defined in Rumbaut and Ima (1988) as children who migrated to their host country with their first generation parents and who have characteristics of both first and second generation migrants. Table 1 shows the distribution of my participants across the different generations. It also gives details as to the number of years they have been living in the UK and the migratory journeys they undertook. Highlighted in yellow are the participants who arrived as migrants to the UK from Europe.

Table 1:

Participant profiles.

5 Findings and discussion

Fishman and Tej's (2004) widely acclaimed GIDS, constructed to explain language maintenance as well as how the reversal of language shift may be achieved, provides a complex model of the different social domains implicated in the health of a language. Key to the model is the sixth of the eight stages of intervention, the “intergenerational and demographically concentrated home-family-neighbourhood efforts” (Fishman and Tej 2004: 427). Fishman claims that it is around this stage that all language maintenance efforts revolve (cited in Canagarajah 2008: 144) and for this reason, I've used my questionnaire to enquire closely into whether and to what extent my participants use Tamil with their families and communities.

The data in Table 2 (below) portrays the different language resources used by generation 1, generation 2 and generation 1.5 participants in different contexts: at the temple; with people of the GP (grandparent) generation, aunties/uncles, parents, siblings, partners, friends, children, grandchildren and SLT youths. The table also presents the language resources used in participants' early lives, at work, when counting, using new media, at their highest level of education, playing sports, and when angry.

Table 2:

Communicative repertoires.

Crucially, the table portrays how the participants draw on their communicative repertoires in these contexts. The cells are shaded from dark blue to light blue to mark whether only Tamil is used (dark blue); whether mostly Tamil is used (medium blue); whether Tamil is used equally together with other language resources (turquoise); or whether only a little Tamil is used (light blue). The cells shaded in yellow show the use of a European language resource, e.g. French, German, Swiss-German, Danish etc. Note that often Tamil is used alongside one of these European language resources. Please refer to the KEY in Table 2 for more information.

In addition, Table 2 attempts to capture the range of each individual's communicative repertoire and demonstrates how they say they deploy their language resources in each interactional context. My presentation of the data has been guided by Canagarajah's (2019) definition of heritage language as practice-based. He claims that in diaspora settings, traditional notions of heritage language as primordial, pure and territorialised lack support. He argues instead for languages to be conceived as mobile resources, which are always in contact and mutually shaping one another (2019: 10). Hence, Table 2 illustrates how the use of Tamil only makes sense in the context of each participant's other language resources.

6 Discussion and conclusion

This article considers HL maintenance in the context of participants' communicative repertoires and argues that there seems to be a move toward Tamil language maintenance in the UK as a result of onward migration from Europe. Canagarajah (2008: 145) claims that in English-speaking countries there is “a distinct language shift toward English and a dramatic decline of Tamil in the diaspora”, but my data does not indicate a pronounced language shift. I discuss how SLT migrants from Europe have been responsible for introducing European language resources into the community and that they are often used alongside Tamil and English in various communicative contexts. This resonates with Canagarajah's recent position regarding the very nature of HL. He claims that Tamil practice amongst diasporic youth is beginning to accommodate other verbal and multimodal repertoires and that “in such contexts, it is difficult to tease apart the HL from the other languages. In fact, a good argument can be made that it is such mixed languages that might be considered the HL for diaspora participants” (2019: 28).

I build on this new framing and show a greater use of Tamil resources amongst the younger generation in terms of the contexts they use Tamil in, as well as the different Tamil registers/varieties they employ, i.e. “pure” Tamil, the Sri Lankan Tamil variety used by parents, mixing Tamil with English and European language resources in both face to face and new media interactions, South Indian dialects and “Tamil cinema”-style banter with peers. I also indicate that there seems to be a change in attitude toward Tamil language use after the end of the civil war. Further research, however, is required to have deeper insight into how Tamil is situated and maintained amidst other communicative resources.

David (2012: 378) declares that “recent Tamil migrants from other parts of Europe – Germany, Holland, France and Switzerland […] are frequent travellers between Britain and Europe”. This constant movement of SLTs also facilitates the maintenance of Tamil in encouraging its use as a lingua franca in transnational contexts. Mobility between the UK and Europe has increased largely due to onward migration. Lindley and Van Hear (2007: 1) describe the “onward movement of new citizens of refugee backgrounds within the European Union [as] an apparently growing pattern in European mobility which has largely been overlooked”. It is therefore crucial that we take it into account when analysing current SLT language practices in the UK as onward migrants from Europe have now become a part of the UK SLT sociolinguistic landscape. Evolving migratory patterns in our ever-changing political climate inevitably affect how a community is internally organised. Whether and how Brexit will impact the SLT diaspora is thus a matter for future consideration.

The distinction between SLT onward migrants and migrants who have long been settled in the UK is not an artificial construct as it is shaped by participant understandings. The concept of “communicative repertoire” provides an ideal framework to appreciate these kinds of internal complexities within the SLT diaspora. Representing the wide range of migrants' linguistic resources and complex communicative practices in various interactional settings within the family and community helps to question the stereotyping of these communities as homogeneous groups (that they come from a limited number of countries and share more or less similar economic, social, cultural, religious, or linguistic backgrounds) (Budach and de Saint-Georges 2017: 63). Again, further research needs to be carried out to better understand the internal diversities that exist within the SLT diaspora, with a view toward demonstrating the extent of their integration into British society. Referring to the widely held belief that migrants will never be fully integrated into the local or “native” community (Vertovec 2010: 90), there is thus a growing need for studies to show how complex mobilities, transnational links and communicative repertoires can contribute to varying degrees of belonging within their own communities and the wider society.