16.1 Introduction

Over the course of the twentieth and twenty-first centuries, the United Kingdom has seen sustained migration from South Asia, predominantly from former British colonies.Footnote ¹ Today South Asians represent just under 9 per cent of the total population, the largest ethnic minority in the UK.Footnote ² The population increased significantly during the twentieth century and has stabilised in recent years. South Asians have settled primarily in urban centres across England, including in London, Birmingham, Leicester and Bradford.

As a result of this sustained migration, South Asian languages – languages spoken on the South Asian subcontinent – have become a very significant part of the linguistic ecology of Britain. Dozens of South Asian languages are spoken across the country, but five, along with their sub-varieties, represent the largest of these: Bengali, Gujarati, Hindi, Panjabi and Urdu. The vast majority of South Asians in Britain speak one or another of these languages, all of which belong to the Indo-Aryan branch of the Indo-European family of languages. In the 2011 census,Footnote ³ these five South Asian languages were the most commonly spoken languages in the UK after English and Polish (Figure 16.1). By the 2021 census, their rank (but not numbers) had dropped due to a substantial net increase in European migration, particularly from Romania, Portugal, Spain and Italy. Several other South Asian languages are also well represented in the UK in somewhat smaller numbers. These include Tamil (a Dravidian language spoken in South India and Sri Lanka), Sylheti (under-reported, as frequently reported as and grouped with Bengali, as in the census data in Figure 16.1), Pashto (one of the four main regional languages of Pakistan) and Malayalam (another Dravidian language).

Figure 16.1 Main language other than English, as reported in the UK censuses of 2011 and 2021.

16.2 History and National Demographics

Contact between South Asian language speakers and English speakers has a long history, beginning with the arrival of the English in India at the end of the sixteenth century, continuing throughout the period of British colonialism in the Indian subcontinent, and in the postcolonial period through the migration of South Asians to Britain. Immigration to the UK, too, has a long history. In the 1920s, Sikhs, some of them soldiers who had fought for Britain in World War I, peddled goods door-to-door in many British cities (Agnihotri Reference Agnihotri1987; Mahandru Reference Mahandru, Alladina and Edwards1991). They encountered racism and exclusion, and notably formed bonds with the Jewish community, who shared this experience. There were also Bangla migrations in the 1920s and 1930s. But the major movement of South Asian populations into Britain came in two later waves: first in the 1950s and 1960s, when post-war British labour shortages led to policies encouraging migration from South Asian and the Caribbean, and second in the late 1960s and 1970s from East Africa (Kenya and Uganda), when policies of Africanisation forced South Asians to leave. Among this second wave were the majority of Gujarati-speaking migrants (Dave Reference Dave, Alladina and Edwards1991). East African migrants were often well educated and had professional and commercial occupations, in contrast to earlier migrants, especially from the Indian and Pakistani Panjab and from what is now Bangladesh, who were mainly from rural areas and had more basic education.

The number of South Asians in Britain has grown dramatically since the 1960s, increasing from to 4.4 per cent (2.6 million) in 2001 to 6.9 per cent (4.4 million) in 2011, and to 8.2 per cent (5.5 million) in the 2021 census. In 2001, just under half of this group were listed under the census ethnic category ‘Indian’; by the 2011 and 2021 censuses, this proportion had dropped to a third, with proportional increases in the categories of ‘Pakistani’, ‘Bangladeshi’ and ‘other Asian’. The majority of the top three Asian categories – Indian, Pakistani, and Bangladeshi – are now represented by individuals born in the UK. The South Asian population continues to grow at a faster rate than that of the white majority.

South Asian communities in Britain share some characteristics but differ markedly in other respects. It is common for later generations to remain actively bilingual, partly because of the substantial size and sustained in-migration in many of these communities. This means that language maintenance can be heavily affected by the size of the community, the status of the heritage language, and the relative proportion of recent migrants from South Asia who speak the South Asian language in question. Stable bilingualism in a community can in turn be an important source of continuing influence on the English spoken by later generations, so in communities where we see longer-term maintenance of the South Asian language, we sometimes see continuing development of a local ethnically marked variety of British English too (see Sharma and Wormald, this volume). Many of these observations, regarding community size, patterns of in-migration, and the nature of social networks between first and later generation migrants, apply equally to other community languages in the UK, not just those of South Asian origin.

16.3 Urban Demographics

South Asians have settled principally in the large conurbations of England. The largest concentrations are in Outer and Inner London, followed by the West Midlands, West Yorkshire, Greater Manchester and Lancashire. There are significant communities in Bedfordshire and Berkshire, in the home counties north of London, and in the smaller industrial towns of Warwickshire, to the south of Birmingham. In Scotland, the main concentrations of Pakistanis and Indians live in the west of Scotland, particularly in Glasgow. In Wales, there are significant numbers of all three main South Asian communities in Cardiff, but lower numbers elsewhere.

Panjabi speakers, of both Indian and Pakistani ethnicity taken together, represent the largest group of South Asian language speakers in Britain. The largest proportions of Panjabi speakers are found in the Inner London borough of Newham and in the Outer London boroughs of Hounslow, Ealing, Greenwich, Waltham Forest and Redbridge. Outside London, there are sizeable communities of Panjabi speakers in Birmingham, Coventry and Wolverhampton in the West Midlands, Leicester in the East Midlands, and Bradford and Leeds in West Yorkshire.

The largest Bangla/Sylheti speech community lives in Tower Hamlets in London (81,400 in the 2021 census).Footnote ⁴ Outside London, the largest Bangladeshi communities are in Birmingham, Oldham in Greater Manchester, Luton in Bedfordshire, and Bradford. Pakistani Panjabis have settled in many of the same areas in Inner and Outer London, Birmingham, Bradford, and Greater Manchester, especially Oldham and Rochdale. The Gujarati speech community is concentrated in Leicester, with other sizeable populations in Brent and Harrow in Outer London, Bolton and Blackburn in Lancashire, and Kirklees in West Yorkshire. The Hindi-speaking community is much the smallest of the main five Indo-Aryan languages. In London, the main areas of Hindi speakers are Newham, Brent, Ealing and Hounslow. Unlike other South Asian language groups, the Hindi speech community is more middle class and more dispersed across the country and across neighbourhoods, due to different migration and employment patterns.

The census data on the geographic distribution of Tamil speakers in the UK shows that, like other British South Asians, Tamils are an overwhelmingly urban population, with 70 per cent of Tamil speakers in England and Wales living in London (Jones Reference Jones2020). The Tamil community is smaller relative to other South Asian groups in London and is concentrated in South East London boroughs such as Bexley, Bromley and Croydon, as well as in the West (Harrow, Ealing, Kingston-upon-Thames). Some of the diversity in settlement reflects differences between the Sri Lankan Tamil population in the UK and the Indian Tamil population (Sankaran Reference Sankaran2020).

16.4 South Asian Language Speakers in London

As noted, London is the largest hub of South Asian settlement in the UK. According to the 2021 census, 36 per cent of the South Asian community lives in London (20.8% of the city’s total population), with major concentrations in West and East London. South Asian languages are concentrated in different neighbourhoods of London: Gujarati in North London (Wembley), Bengali in East London (Tower Hamlets), and Panjabi in West London (Southall and Hounslow).

Classifying languages and estimating speaker numbers accurately has historically been a challenge in the UK. In the case of London, for example, there was initially a practice of collecting such statistics through the now disbanded Inner London Education Authority (ILEA). There followed a period of little uniformity or consistency in local educational authority (LEA) statistics, with use of divergent terms for home languages (e.g. variable use of ‘Panjabi-Pakistani’, ‘Panjabi-Sikh’, ‘Panjabi-Hindi’, ‘Panjabi-Mirpuri’, ‘Mirpuri’, ‘Panjabi (Urdu script)’, ‘Urdu’). More recently, improved LEA statistics have emerged, with a more detailed picture of home language use.

The largest detailed survey of London language use to date – one that goes beyond the single ‘main language’ question in the census, and that took place before the census had any language questions in England – was conducted by Baker and Eversley (Reference Baker and Eversley2000). They surveyed 850,000 children in London schools and asked what the primary home language was; 71.6 per cent of children reported that English was their main home language (608,500). The next four languages in order of number of responses were all South Asian languages: Bengali (including Sylheti), Panjabi, Gujarati and Hindi/Urdu (ranging from 40,000–26,000). In Baker and Eversley’s data, French ranked 15th (1,800) and Polish ranked 25th (1,500). Using large-scale language mapping via geographical information systems, the project estimated that Panjabi was the second most used language after English, with between 143,600 and 155,700 speakers. This was followed by Gujarati (138,000–149,600 speakers). Hindi/Urdu, controversially classified as one language, was fourth, with a high estimate of 136,500 speakers, closely followed by Bangla-Sylheti, with somewhere between 120,000 and 136,300 speakers. Three other South Asian languages figured in the top forty: Tamil (17th), Sinhala (37th) and Pashto (39th).

Twenty years later, and with more direct census data, the figures look quite different for the top six languages. According to the 2021 census, out of 8.8 million London residents – both Inner and Outer London boroughs – 78 per cent (6.8 million) report English as their main language, followed by Romanian (1.9%, or 159,000), Spanish (1.4%), Polish (1.3%), Bengali (1.2%), Portuguese (1.1%) Gujarati (1%), Italian (1%) and Panjabi (0.9%). Though the two surveys are not directly comparable, in proportional terms several European languages have clearly expanded substantially, while South Asian languages have remained among the most widely attested in the capital. This is primarily due to the continued high presence of South Asians in Britain as well as chain migration of relatives after settlement. Census figures derive mainly from these first-generation South Asians; later generations of British Asians tend to report their main language as English, even if bilingual. This is one of many problems with the use of a single ‘main language’ question in the census.

16.5 Language Status in South Asian Communities

The official quantitative picture of languages in the UK is very limited. Until 2011, the national census gathered no language-related information at all, and it subsequently introduced just a single question asking for a respondent’s main language, a monolingual framing that was retained in the 2021 census. Other borough- or council-level information data can be informative but has not always been consistently gathered over time and across regions. These figures can erase many aspects of the sociolinguistic complexity of South Asian multilingual repertoires, cultures and communities. In this section, therefore, we consider the more qualitative backdrop to South Asian language use in the UK.

Several South Asian communities are strongly associated with a high-prestige South Asian language, even though the majority of community members may speak another language. The majority of Bangladeshis in Britain come from the Sylhet province in the north-east of Bangladesh and speak Sylheti as their home language, rather than standard Bangla. Bangla is nevertheless often the language that is promoted more in terms of prestige and literacy (Lawson and Sachdev Reference Lawson and Sachdev2004) and reported in the census. The Pakistani community is strongly associated with Urdu as a language of literacy and as a high-prestige lingua franca that is used across the entire northern part of the Indian subcontinent. However, like Bangla, Urdu is often not the primary home language for Pakistanis in Britain, nor indeed in Pakistan. A substantial number of Pakistani immigrants to Britain came from the Mirpur province and speak Mirpuri, a variety of the Potwari language that shares features with Punjabi and Pahari. Some Pakistani migrants speak Pashto as their home language. Many Pakistani households will involve a layered form of ‘triglossia’, with Mirpuri, Panjabi and Urdu layered across social situations and social networks.

This diversity shows that (i) South Asian languages in the UK are not in an equally subordinate role relative to English, but rather form a complex internal hierarchy of prestige and (ii) it is a mistake to attempt to map one language to each South Asian community.

Panjabis are an excellent example of how fruitless an attempt at a one-to-one mapping would be. The geographical area called the Panjab (East and West) in South Asia predates the partition and postcolonial creation of India and Pakistan. As a result, the ethnolinguistic group described as ‘Panjabis’ are subcategorised into Panjabi Sikhs, Panjabi Hindus and Panjabi Muslims based on their religious affiliations, all very large demographic groups in South Asia. The region is united by a common language, Panjabi, which forms a bond between the three groups now divided by religious and national boundaries. In the UK too, these three groups are large and maintain a linguistic bond.

Another interesting complexity across the wider Panjab region – and consequently across the three main Panjabi cultures in London and in the UK more widely – is their distinct ideologies regarding their choice of language of prestige for education and government. Panjabi Sikhs consider Panjabi their mother tongue, but for historical reasons also the symbolic language of their culture, and so it has been formalised as their language in education and politics, and in the language of literacy via the Gurmukhi script. Hindu Panjabis and Muslim Panjabis, by contrast, also use the language as their mother tongue but generally ascribe it the status of a colloquial language, associated with their vernacular home culture, and appropriate for such uses as folk music on festive occasions. For well-known historical reasons, these two communities choose different languages as their language of prestige for educational and other official functions: (Sanskritised) Hindi and Urdu respectively. Some Sikh and Hindu Panjabis do have an interest and competence in two literacy traditions, but for the most part, this three-way division of religious and national identity gives rise to distinct educational and prestige ideologies in the Panjabi-speaking communities of South Asia and the UK. At the same time, the language facilitates strong bonds of friendship across religious subcommunities and among Panjabi-speaking second-generation British Asians from different cultural traditions (Sharma and Rampton Reference Sharma and Rampton2015).

The Tamil community similarly defies a simple national or religious cultural profile, with a large Sri Lankan Tamil population, many of whom historically sought asylum from the civil war in Sri Lanka (see Sankaran Reference Sankaran2020), as well as an Indian Tamil group, and additional migrants from the sizeable Indian Christian community in South India.

16.6 South Asian Languages in Education

These layered systems of ethnic, national and religious identity, along with associated language ideologies and priorities in different communities, and changing national education policies, together form a very complex situation of South Asian language teaching in Britain’s schools.

South Asian languages are an important dimension of modern languages in the British educational context. Approximately 1.7 million students in England are currently reported to have English as an Additional Language (EAL), with London, Slough, Luton and Leicester, all cities with large South Asian populations, having some of the highest proportions (Cushing, Georgiou and Karatsareas Reference Cushing, Georgiou and Karatsareas2021). In some boroughs or LEAs, children who speak a South Asian language at home form the majority of incoming pupils. In Birmingham, Leicester and Bradford, they have at times represented a quarter of the total state school population. I first offer a brief review of the historical presence of South Asian languages in British education, and then return to the role of language ideologies and market forces in language choice in education.

Britain has a long tradition of teaching the languages of the Indian subcontinent. The expansion and consolidation of the Indian empire necessitated the training of British civil servants and army personnel in the languages of colonial India. Early Oriental Studies departments at the universities of Cambridge, London and Oxford offered courses in Hindi–Urdu–Hindustani, guided primarily by instrumental rather than integrative motivation, but nevertheless resulting in pedagogic and philological research in South Asian languages.

The arrival of a large number of children from the Indian subcontinent and other regions in inner-city British schools in the 1960s attracted a variety of rushed educational responses by language planners in LEAs, the erstwhile Department for Education and Skills (DES) and the Home Office. Language debates moved away from linguistic diversity to English-only stances in the early education of these children. Their educational context was perceived as a problem, both linguistic and cultural. Their parents were frequently trapped by socio-economic pressures and hostile socio-cultural experiences, and so were easily persuaded that their heritage languages would interfere with the children’s acquisition of English as a second language. The mother tongue was perceived as a handicap and an assumed lack of competence in English as a disadvantage. Taking a narrowly monolingual view, the Plowden Report (Reference Plowden1967:69) said that ‘it was absolutely essential to overcome the language barrier’, notably presupposing that a second language was a barrier. Ironically, it was in this period that linguistic research was starting to show that the home language can provide crucial scaffolding for the acquisition of further literacy, numeracy and cognitive skills (Cummins Reference Cummins2000).

The Bullock Report (Reference Bullock1975) was the first major positive assessment of the pedagogical and socio-cultural significance of the mother tongues of ethnic minority children. It recognised these as ‘an asset … to be nurtured’ by schools (1975:249; see also Rampton et al., this volume). A series of studies followed that supported bilingual education. For example, one of the conclusions of the DES- sponsored Linguistic Minorities Project (1979–83) was that ‘as far as Panjabi was concerned, the clear superiority in performance of the experimental (Panjabi-speaking) group was accompanied by indications of a transfer of higher-level ability to more complex tasks in English’ (Fitzpatrick Reference Fitzpatrick1987:109). Another study in Bedford found that Panjabi children had achieved a higher level of literacy in English than in Panjabi but were orally more fluent and accurate in Panjabi than in English (Tosi Reference Tosi1980).

During this time, minority communities began to establish community language and culture classes, and these positive stances to language maintenance were taken up by many professional teachers’ organisations and by some LEAs. Nevertheless, the Swann Report (Reference Swann1985), an enquiry into the education of children from ethnic minority groups, continued to adopt the long-standing monolingual argument that mother tongue use in primary schools would be a barrier in the acquisition of English. This view has persisted to the present day, despite research showing evidence that South Asian home languages do not interfere with English literacy development (e.g. Stuart-Smith and Martin Reference Stuart-Smith and Martin1999). The report also established the continuing view that minority ethnic communities are responsible for the teaching of their languages. South Asian communities depend largely on a combination of home use and complementary schools – also referred to as supplementary schools, Saturday schools, heritage language schools or community language schools (Cushing et al. Reference Cushing, Georgiou and Karatsareas2021) – for heritage language maintenance, particularly in the crucial early childhood years.

Swann recommended subsuming South Asian languages within the modern foreign languages curriculum, rather than within the model for indigenous minority languages of the UK (Welsh, Irish, Scottish Gaelic), which have more potential for integration into the primary or secondary curriculum. This put the home languages of non-autochthonous Britons in competition with high-prestige European languages rather than treating them as heritage languages, and contributed to a dramatic loss of modern language skills across the UK over the past two decades (see Rampton et al., this volume). The later Kingman (Reference Kingman1988) and Cox (Reference Cox1989) reports used rhetoric around speakers’ rights to their heritage languages but continued to endorse Swann’s recommendation.

The rise of a market-based approach to language learning has led to further shifts in education. Language planning policies in the school curriculum have transformed ‘mother tongues’ into ‘foreign’ languages, driven by an ideology not to foster or support the cognitive advantages of early bilingualism but to teach these languages for academic and commercial pursuits. After a reform of GCSE and A-Level qualifications in 2015, a number of smaller or ‘lesser-taught’ languages were discontinued, with a focus on Arabic, Modern Greek, Gujarati, Bengali, Japanese, Modern and Biblical Hebrew, Panjabi, Polish, Portuguese, Turkish and Urdu (Long, Danechi and Loft Reference Long, Danechi and Loft2020). And regular intervention is needed to make the case for retaining national-level qualifications for lesser-studied languages.

Today, the choice of languages at GCSE and A-Level is determined to a large part by which languages are offered by schools, which in turn is influenced by prevailing language ideologies that tend not to favour South Asian languages as highly as European languages (and recently also Arabic and Chinese), even in communities where the main minority languages are South Asian.

Table 16.1 shows entry rates for a number of community languages in the GCSE ‘Other Modern Languages’ category. The first thing to note in comparison to Figure 16.1, is that entry levels do not correspond to community size: Arabic and Chinese appear to be languages with global appeal, chosen beyond their community numbers, whereas most South Asian languages have significantly lower uptake, despite larger communities, suggesting that they are seen as home languages and not languages for educational or professional advancement. Second, Table 16.1 shows that this discrepancy is widening: Arabic and Chinese have increased over the last decade, while Panjabi, Gujarati and Bengali have dropped in numbers, sometimes by more than 50 per cent. Urdu groups with Arabic and Chinese, being seen as a high-prestige lingua franca with global appeal.

Table 16.2 delves further into these language ideologies. The data are from one of very few detailed examinations of GCSE language statistics beyond the three main European languages (Vidal Rodeiro Reference Vidal Rodeiro2009). Although the data are not recent, the detail of interest is in the internal contrasts here.

Note, firstly, that the level of uptake in 2007 was much higher overall than four years later, from 2011 onwards (Table 16.1). As seen earlier in Table 16.1, Table 16.2 shows again that Urdu, the language with highest prestige in the South Asian regional context, attracts approximately five times more students than the next most popular South Asian languages at GCSE, yet it has never been the Asian language with the most reported speakers in the UK. The same was noted earlier for Arabic, also visible here. These languages are selected for their religious and cultural status; many South Asian schoolchildren currently choose Arabic or Urdu over their home language at GCSE level. This is made clear in Table 16.2: 95 per cent of GCSE students of Bengali had the language as their mother tongue (this of course treats Sylheti speakers as Bengali speakers, a layer of ideological erasure), in contrast to only 58 per cent for Arabic, 43 per cent for Urdu, and of course less than 1 per cent for the main GCSE languages of French, German and Spanish. Although this might suggest that a high proportion of Bengali speakers are taking their language at GCSE level, it actually only means that few other people take Bengali. In fact, only 45 per cent of Bengali speakers took Bengali at GCSE (and only 15% of Panjabi speakers), in contrast to 46 per cent of Urdu speakers, the highest of all the South Asian languages. In sum, the numbers are a stark illustration of language ideology in action: high uptake of Urdu by Urdu speakers and by others, in contrast to lower uptake of Bengali by Bengali speakers and by others. Panjabi and Gujarati pattern like Bengali, being treated as only community-relevant and tending to lose out to higher-prestige languages.

Current institutional provision for South Asian languages thus ranges from voluntary community language classes managed entirely by the communities with minimal financial support from the local authorities to community language classes offered as part of the mainstream curriculum, for example as GCSE or A-Level subjects, but almost exclusively after the primary-school years when languages are best acquired. South Asian language organisations and South Asian media, along with some provision of heritage language subjects at university level, also continue to support learners with an interest in their heritage languages.

16.7 Language Maintenance, Mixing, and Shift

The overwhelming majority of sociolinguistic research on the languages of these communities has focused on new British Asian English varieties (see Sharma and Wormald, this volume). Much more research is needed on the sociolinguistics of South Asian heritage language use. This section reviews some of the insights gained from a small set of studies to date.

Although South Asian communities see relatively healthy levels of heritage language maintenance and use, they face the almost inevitable decline in usage seen in such communities. This occurs both due to general integration into the majority language over generations and because certain languages are associated with higher value, as discussed above. Patterns of language maintenance, language mixing and language shift are remarkably similar across South Asian language communities. For example, smaller communities see more language shift both towards English and towards larger South Asian languages.

In early work on language maintenance, Romaine (Reference Romaine, Dabène, Flasquier and Lyons1983) reported a classic pattern of contact bilingualism among Panjabi schoolchildren in Birmingham, namely a gradual community-wide shift to English, particularly through interaction between siblings in the home. The Linguistic Minorities Project (1985) showed that the use of Panjabi, both oral and written, and knowledge of Panjabi was fairly solid in the adult British Panjabi population, but that English was becoming the dominant language for children. Nearly two generations later, Sharma (Reference Sharma2017) gathered detailed language use questionnaires with seventy-five people across three generations in the Panjabi community of Southall, in West London, and found a similar pattern in the next generations: almost all seventy-five participants across three generations had basic or advanced knowledge of Panjabi, but with some decline in the proportion of Panjabi used in their daily repertoires. So while competence was fairly intact, use was slightly in decline. This difference, along with the widespread use of a ‘mixed code’, was similar to what earlier studies showed (Agnihotri Reference Agnihotri1987). Sharma’s (Reference Sharma2017) findings also parallel Saxena’s (Reference Saxena1995) findings for Panjabi Hindus in Southall twenty-five years earlier, with domain-specific language roles. Saxena found Panjabi to be predominantly associated with the family domain, Panjabi and English with community, English with education and employment, and Hindi with Hindu religious contexts. He described this as a double overlapping diglossic situation: ‘Hindi is the High language and Panjabi the low at the community level; and English is the High language and both Panjabi and Hindi are the low languages at the wider societal level’ (Saxena Reference Saxena1995: 220). Notably, he concluded that English had not encroached enough on the family domain to threaten Hindi and Panjabi, and Sharma’s later findings bear this prediction out; Reynolds (Reference Reynolds, Cotterill and Ife2001) came to a similar conclusion for the Panjabi community in Sheffield.

In sum, despite shifts in use, decline in knowledge of Panjabi has not been precipitous over fifty years, at least in the West London community, possibly due to community size, sustained migration, and language teaching via gurdwaras and other community forums. Sharma found that the social network detail that correlated most strongly with level of Panjabi use was the proportion of India-born ties in the person’s network. In terms of casual interaction across friendship groups, for example between youngsters in school, or even between two middle-aged Sikh and Muslim friends who grew up in West London (Sharma and Rampton Reference Sharma and Rampton2015), the community shows evidence of extensive use of vernacular mixing of multiple Panjabi and English styles (Rampton Reference Rampton1995). The project also found an important role for consumption of Indian film, music and other media.

The Linguistic Minorities Project (1985) found similarly robust levels of competence in Gujarati, but again with a trend towards code-mixing and reduced use in younger generations. Later research found a similar positive community orientation to language maintenance: Choudhry and Verma (Reference Choudhry and Verma1994), looking at 109 adults and 141 children in the Gujarati communities in London, Leeds, Bradford and Leicester (Hindu and Muslim), found that the majority felt that ethnicity without mother tongue was not sustainable. Gujarati remained dominant cross-generationally in the home domain. Some shift to a mixed code with English among later generations was evident, but this and later work (Sneddon Reference Sneddon, Martin-Jones and Jones2000) found that strong family networks and close links to the region of origin helped sustain a degree of language vitality. The notable shift was in Gujarati literacy skills among younger community members (also found by Bhatt and Martin-Jones Reference Bhatt, Martin-Jones and Fairclough1992), a situation reflected in most South Asian language communities.

The majority of Bangla speakers in the UK are Bangladeshi Muslims, but with some presence of Hindu Bangla speakers from West Bengal in India. As a result, much of our knowledge of Bengali use in the UK is through research on Sylheti-speaking communities. As with the other languages, studies over a long period of time have shown robust use of Bangla in the community (Husain Reference Husain, Alladina and Edwards1991; Mesoudi, Magid and Hussain Reference Mesoudi, Magid and Hussain2016), but with the familiar dynamic of English mixing and preference among younger children, for example in a study in Leeds (Verma et al. Reference Verma, Mukherjee, Khanna and Agnihotri2001), and a preference for English in the British-born generations’ sibling interactions (also found in Blackledge Reference Blackledge, Martin-Jones and Jones2000, a study based in Birmingham). Literacy practices reflected an orientation to both Arabic, for religious use, and Bangla (Gregory and Williams Reference Gregory and Williams2000).

As noted, the Tamil community in the UK and particularly in London is somewhat different for a number of reasons. First, it is a smaller community, and so language maintenance can be compromised by small numbers of speakers, fewer opportunities for language use, and fewer resources for community language teaching. Second, there are two different primary sources of Tamil speakers, a larger group of Sri Lankan Tamil speakers and a smaller group of South Indian Tamil speakers. Sri Lankan Tamils came to the UK in distinct waves: first with middle-class professional migration in the mid twentieth century, then in larger post-independence numbers, especially after Sinhala was made the official language in Sri Lanka in 1956, and finally a particularly large number of refugees after the onset of civil war in 1983. Sri Lankan Tamils are therefore a very diverse group in terms of nationality, caste, class, education, religion and reasons for migration (Sankaran Reference Sankaran2020). Recent studies indicate fairly similar language maintenance to other South Asian languages, though interestingly, Canagarajah (Reference Canagarajah2013) suggests that the South Indian dialect spoken in Tamil movies may influence the Tamil spoken by younger Sri Lankan Tamils in the diaspora. Smaller or more dispersed language communities (e.g. Nepali, Malayalam, Sindhi, Pahari, Konkani, Balochi) can struggle more with language maintenance.

In contrast to more regionally circumscribed languages, Hindi and Urdu have a different profile in terms of maintenance. Information about Hindi use and acquisition in the UK, particularly as a mother tongue, is less clear as such speakers may come from diverse regions and there are consequently few clear Hindi-speaking neighbourhoods in the UK; it is also used by more middle-class migrants who have more diffuse British settlement patterns, and as a lingua franca among speakers of other languages. Census reporting tends to blur the distinction between Hindi and Urdu, which complicates the picture, but certainly Hindi/Urdu has a fairly stable presence in the UK for wider reasons of prestige, religion and popular culture, discussed earlier in relation to Tables 16.1 and 16.2.

Overall, sustained migration from South Asia, relatively large settlements in the UK, and a reduction in the early explicit hostility to migrant cultures has meant that South Asian speech communities have broadly maintained their languages in the family and the community, though with gradual inroads by English with each generation.