Supporting speakers of community languages: a case study of policy and practice in primary schools

ABSTRACT This article explores policy and practice in relation to support for speakers of community languages in Northern Ireland primary schools against the backdrop of the broader UK context, with reference also to the Republic of Ireland and wider European and international experiences. After an initial discussion of the educational, social and political context pertaining to Northern Ireland, we examine language and education policy as they relate to community languages, drawing out the issues that are common across the UK and pointing up those that are specific to or have particular resonances in Northern Ireland. The discussion is informed by ethnographic fieldwork in Belfast primary schools with contrasting socio-economic profiles. Our findings show that in addition to familiar challenges, a number of key factors are critical in Northern Ireland, including the recent surge in numbers of newcomer pupils, the lack of a statutory language component in primary schools, the economic dynamic in the education system, the divided nature of Northern Ireland society and its education system, the legacy of the conflict and socio-economic factors. We end by making suggestions for a way forward in policy terms.


Introduction
Community languages 1 are a vibrant component of everyday life in the UK and are supported by a network of supplementary schools, a proactive National Resource Centre for Supplementary Education, as well as a plethora of grassroots initiatives and a number of legal protections in terms of access to services (e.g. medical or legal). However, they are often undervalued in terms of cultural capital, community cohesion and economic prosperity (Ayres-Bennett & Carruthers, 2018) and operate in a complex dynamic with questions of ethnicity, religion, socio-economic status and gender (Modood & May, 2001). Indeed, as we shall see, research into community languages highlights a number of problematic issues, particularly in the field of education, where language is one component of 'multiculturalism', itself a highly contested construct in terms of the education system (ibid.). In these islands, only Scotland and the Republic of Ireland (henceforth RoI) have positive policies in place. In Scotland (Scottish Government, 2012), community languages are included in the '2' of the '1 + 2' policy which is based on the European Union model whereby every child learns two languages in addition to their 'mother tongue' (which in Scotland's case, is usually either English or Gaelic). There is particular support for Urdu and Chinese, although the implementation is challenging, not least due to the lack of a cohort of trained teachers (Foley et al., 2018). The RoI published its Languages Connect strategy in 2017, which includes community languages alongside modern foreign languages and is in the early stages of implementation (DfES, 2017). This strategy foregrounds inclusiveness and the multiple benefits of supporting the languages of the 'new Irish' in a context where immigration has increased sharply in recent years and now represents 11% of the population (DfES, 2017, Foreward).
The Belfast or Good Friday Agreement (Northern Ireland Office, 1998, henceforth GFA) acknowledged Northern Ireland's (henceforth NI) linguistic diversity in the section entitled 'Rights, Safeguards and Equality of Opportunity' (our italics): All participants recognise the importance of respect, understanding and tolerance in relation to linguistic diversity, including in Northern Ireland, the Irish language, Ulster-Scots and the languages of the various ethnic communities, all of which are part of the cultural wealth of the island of Ireland.
However, in practice, the policy priority has been Irish and Ulster Scots; while there have been some positive initiatives in relation to community languages, there has been no policy intervention. In this paper, our focus is on support for speakers of community languages in the NI primary education context against the backdrop of the wider UK situation; reference will be made also to RoI and other European or international settings. We will draw on a wide range of policy, strategy and resource documents and on data from ethnographic fieldwork in two Belfast primary schools which are in sharp contrast in socio-economic terms.
The first section ('Community languages in the wider UK context: problematic issues') will outline briefly the major issues raised by research on support for speakers of community languages at primary level in the rest of the UK. The section on 'Community languages in NI' will set out the NI policy context in which community languages sit and will outline our methodology. The central section of the article -'From policy to practice. Evidence from fieldwork' -will focus on the findings that emerge from our ethnographic fieldwork. We will ask if and how the issues raised in relation to other parts of the UK might be mirrored in NI and we will discuss the ways in which the distinctive features of the NI context are reflected in practice within our two schools. The penultimate section will focus on 'Socio-economic factors' and we will conclude with a section that identifies ways in which our findings could act as a catalyst for future research and policymaking in this field.

Community languages in the wider UK context: problematic issues
Over the past few decades, the question of educational support for speakers of community languages has received considerable scholarly attention. Because of the absence of a policy in most of the UK, language learning takes place largely outside mainstream schooling, through supplementary or community schools which are often supported on a voluntary basis through grassroots level community organisations, charities or particular religious groupings. Initial research on community languages centred mainly on various aspects of bi(multi)lingual literacy, including the management of linguistic diversity in the school sector, the ways in which mainstream schools can benefit from children's multilingualism, and the role that supplementary schooling plays in their identity formation and linguistic repertoires (Creese, 2009). Recent studies focus on how diverse actors in the school environment such as educators, newcomer parents, pupils and their English-speaking peers position themselves vis-à-vis policy discourses as a means of opening up ideological and implementational spaces for the use or non-use of community languages (Wei, 2018).
The 'language barrier': community languages as a problem 2 In English-dominated multilingual settings such as the UK, any lack of English is often considered as a 'deficit' and speaking a community language is viewed as a problematic 'barrier' to the acquisition of the host language, leading to sink-or-swim type immersion programmes as a quick solution for the integration of newcomers (Capstick & Delaney, 2016). 3 Contemporary research advocates numerous functional and cognitive benefits of using community languages in multilingual classrooms, where newcomers' bilingualism could be an asset for the whole class (García & Wei, 2014). In practice, however, many schools continue to perceive pupils' competence in home languages as a disadvantage, often placing intelligent young multilinguals in 'low ability classes' (Ainscow et al., 2016). Home language competence is often considered a 'resource' only when pupils are proficient in English and will be able to engage in 'language brokering', 4 a practice which is not always viewed positively in terms of children's wellbeing (Bauer, 2016).

Monolingualism inside and outside the classroom
Contemporary research in England suggests that the recognition and promotion of community languages in the classroom depend to a great extent on "individual teachers and schools' interpretation of policy" (Bailey & Marsden, 2017, p. 284). Classroom practices are often governed by an 'English only' monolingual ethos whereby children "suppress many of the linguistic and cultural competences they possess", causing a rapid home language shift during the early years of schooling (Kenner & Ruby, 2012, p. 118). Mehmedbegović (2011), researching multilingual classroom settings in Cardiff and London, also notes that some educators exhibit reluctance towards the use of community languages in the classroom. Such ideology frequently renders the non-English speakers prone to marginalisation, with "L1 being categorised as the 'non-legitimate' language" (Saxena, 2009, p. 168). Moraru (2019, p. 6) underlines similar trends in her research on second-generation British-Arab migrants in Cardiff, where English and Welsh possess a greater degree of capital but "languages associated with the phenomenon of immigration are rendered illegitimate". Moreover, newcomers' multilingual practices may also cause instances of negotiation and/or conflict among peers because they represent skills that their monolingual peers would not possess (Evans & Liu, 2018).
If such scenarios are not managed wisely by educators, it may lead to alienation or exclusion of immigrant children in the school settings (Manzoni & Rolfe, 2019).

Language hierarchies in mainstream education
There is evidence that European modern foreign languages are valued over community languages (Collen, 2020) and that in terms of social, symbolic and cultural capital, most immigrant languages are at the bottom of the power structure, with the only exception being Chinese (mainly Mandarin), which has increased its status from a community language to a modern foreign language over the past two decades. This is due to the government's intention to "double the number of Chinese learners, particularly in schools and among young people, by 2020" to secure both its cultural and economic ties with China (Tinsley & Board, 2014, p. 4) and to investment in the Mandarin Excellence Programme (DfE, 2018). Even in the Scottish '1 + 2' language policy, the main focus has been on European languages and Gaelic, with some promotion of Chinese and Urdu . In Wales, the '+1' in the 'bilingual (Welsh and English) + 1' strategy is a European foreign language in the majority of cases.

Issues of training, resources and confidence amongst educators
At primary level, questions of teacher training and CPD in relation to teaching in the multilingual classroom are difficult to separate out from the area of EAL. Although initially taught separately, EAL is no longer considered a subject of specialisation and is widely perceived across all administrations as a "generalist skill desired of all teachers" (Creese, 2004, p. 190). Some local authorities and/or individual schools in England with a high level of diversity may also appoint more specialised consultants to work predominantly on mainstream teachers' professional development from a whole school perspective, rather than just the provision of specialist teaching support for EAL learners (NALDIC, 2020). In practice, mainstream teachers often lack systematic preparation in relation to bi/multilingualism during their initial teacher training (Foley et al., 2018) and for some classroom teachers, as Skinner (2010, p. 75) argues, "there is a definite sense of having to learn on the job".
Along with EAL support, there is a range of mentoring programmes and several toolkits available, offering advice on welcoming newcomer children and practical suggestions for teaching (British Council, 2020;EEF, 2020). Indeed the DfE (England) recommends use of the 'Belonging Toolkit' published by the children's charity Coram (Coram Life Education, 2020). Similar toolkits are available in Wales and Scotland (Education Scotland, 2020;Welsh Government, 2016). The approach is variable but looking across the piece, there is a focus on EAL and on celebrating cultural diversity; promotion of home languages often takes the form of 'language of the week', or classroom learning of greetings or particular words. One of the most supportive toolkits is the National Union of Teachers' Welcoming Refugee Children to your School which advocates strongly for collaboration with supplementary schools (NEU, 2019). However, the extent to which good practice is implemented on the ground depends very much on the individual school's policies and the level of interest from their employees (Manzoni & Rolfe, 2019). Overall, training, resources and implementation of good practice in the multilingual classroom are highly variable.

Background
Among the immigrant communities, the Indian and Chinese are the longest established ethnic minorities in NI, having been settled since the 1930s and the 1960s respectively (Holder, 2003). Even though transnational migration to NI slowed down during the 'Troubles' , it regained momentum during the initial years of the current century because of migration from the eight new member states of the European Union. The 2011 Census reveals that, apart from English, Polish was the most spoken mother tongue in NI, with over 17,000 speakers, followed by Lithuanian with 6,250 speakers. Portuguese, Slovak and Cantonese and Mandarin are the other migrant languages, with more than 2,000 speakers each; Filipino, Latvian, Malayalam (India) and Hungarian are also widely spoken (NISRA, 2012). School settings are thus increasingly diverse, as is evident from several government reports (see EA, 2018;Kernaghan et al., 2015). For instance, the NISMP records a 72% increase among immigrant students between 2007 and 2012 (Kerr, 2014) and the Department of Education's (henceforth DENI) recent statistics on school enrolment (2020) state that there are almost 16,000 newcomer pupils attending NI primary and secondary schools. The term 'newcomer pupil', as DENI (2009, p. iii) defines it, refers to a child or young person who has "enrolled in a school but who does not have satisfactory language skills to participate fully in the school curriculum and does not have a language in common with the teacher"; currently, newcomers represent 4.9% of the total primary and secondary school population. This linguistic diversity in the NI education sector is nonetheless much lower than in England and lower than the other devolved administrations: according to the DfE's (2020) School Census data, 21.3% of primary and 17.1% of secondary school pupils in England are classified as EAL. In Wales, EAL pupils represent around 7.5% of the total primary and secondary school population (Welsh Government, 2020), while the corresponding percentage for primary and secondary EAL pupils in Scotland is 6.7% (Scottish Government, 2019).

Policy context
NI's education system differs in important ways from England, Scotland and Wales and some of the differences have important cultural resonances that are directly and indirectly relevant to questions of language learning. At primary level, most schools are either 'maintained' (largely Catholic) or 'controlled' (largely Protestant), both of which are state-funded; there are also integrated schools and Irish-medium schools (both state-funded) and a small number of private schools. As a result, most children attend a primary school that is firmly rooted in one of the two largest communities -CNR ('catholic-nationalist-republican') or PUL ('protestant-unionist-loyalist'). This core division is of course operating in a post-conflict situation in NI, where the legacy of the conflict still permeates many aspects of everyday life.
There has been no official language policy in NI but the recent re-formation of the executive (2020) has involved a commitment to legislate to support Irish. The North-South bodies created after the GFA include a language body and there are additional 'soft' support mechanisms in place for Irish and Ulster Scots. Irish-medium education enjoys some protection in the form of a requirement to encourage and facilitate its development (Carruthers & Ó Mainnín, 2018). As elsewhere in the UK, under a 'policy of selfreliance' (Wei, 2018), maintenance of community languages depends to a large extent on supplementary schools which in turn depend on parental interest and grassroots-level community endeavours. Currently in NI there are two Russian, one Arabic, one Punjabi, one Chinese and around twelve Polish supplementary schools and some mainstream secondary schools facilitate entry for GCSEs in the languages concerned.
It is also important to note that in terms of mainstream education, NI is the only part of the UK where there is no statutory provision for Primary Languages (Ayres-Bennett & Carruthers, 2019). Although some individual schools are extremely proactive, provision is patchy and lacks financial support (Jones, 2015). As in the Australian context (where language learning at primary level is only compulsory in the state of Victoria), content, time allocation and prioritisation are variable and dependent on a host of factors such as school type, status, funding models and parental support (Cruickshank et al., 2020). Moreover, since Education is devolved, wider factors can influence indirectly the position of languages. For example, DENI comments frequently on the oversupply of schools in NI relative to the size of the population and thus the financial inefficiency of the system, arising from the need for multi-sector provision and the rural nature of much of NI (Meredith, 2019).
Maintaining and developing proficiency in community languages, as McMonagle and McDermott (2014) note, is extremely important for identity and cohesion within communities. In the context of education, the NI government's only policy towards newcomer pupils is contained in Every School a Good School: Supporting Newcomer Pupils (DENI, 2009) and is focused strongly on rapid English acquisition for newcomer children so that they can access the curriculum and integrate in the classroom (Skinner, 2010). The policy states that its primary concern is to support "the newcomer students in their acquisition of the language of instruction, whether it be English or Irish, in an inclusive manner, to enable them to access the curriculum in particular and partake in every aspect of school life" (DENI, 2009, p. IV); indeed the term 'language barrier' is used multiple times. Acquisition of the language of instruction primarily takes place through immersion in the classroom and often employs 'buddy schemes', where a newcomer pupil is paired with a native anglophone classmate (McDermott, 2011). This policy has also allocated funding to schools per newcomer child for at least the first three years of his/her school life so that schools can provide specialised support and build the expertise of their teaching staff, although it has been argued that this funding is not adequately ring-fenced or supervised (Jones, 2015). Additionally, DENI has funded the setting up of a Regional Support Service in the Education Authority (EA) called the Intercultural Education Service (IES) to strengthen and improve support to newcomer pupils and their parents.
DENI is currently undertaking a review of policy in relation to newcomer pupils and it is clear from the initial pre-consultation exercise, involving focus groups with various stakeholders, that questions around the need to value community languages are now surfacing: parents have expressed concern about children losing their home language; they would like to see their children gaining a qualification in their home language; and they are concerned that there is some confusion on the part of the authorities between SEN classification and lack of English language skills (DENI, 2018). As a result of these concerns, one of the questions in DENI's current consultation is: "in your experience, are Newcomer pupils given opportunities to maintain their home language(s) in school? What are the difficulties / challenges in facilitating this?" (DENI, 2019). Moreover the consultation document includes a section on 'Practice in other UK and European Jurisdictions' which summarises elements of best practice elsewhere and states: "use of a Newcomer pupil's mother tongue should be encouraged as the rate of new language acquisition has been found to be consistently faster when the mother tongue is also used within the learning environment" (DENI, 2018, p. 17). This apparent evolution in educational policy perspective is aligned very strongly with a key document recommended by the EA and published by the former Inclusion and Diversity Service (now IES) in support of newcomer pupils which was developed originally with the support of a North-South body, SCoTENS, and the Departments of Education in RoI and NI, i.e. the Toolkit for Diversity in the Primary School (IDS, 2016). The toolkit is available on the website of the EA and represents a major advance on the 2009 policy documents. Its purpose, as a senior adviser in the Education Authority puts it, is to "draw together advice into one resource for schools to cover intercultural awareness, welcome and communication with home, assessment and planning and curriculum access". 5 It opens with a section on the importance of maintaining home languages, giving multiple examples at every level of how schools and teachers can use the home languages of children in their classroom teaching for everyone's benefit, both educational and personal. It also suggests that the school should encourage parents to use the home language with their children. This Toolkit is in fact one of the most comprehensive and sophisticated of those available in the UK: requests to use the Toolkit have come from Canada and five European countries. Promotion of good practice is also evident in the resources for 'Schools of Sanctuary', where welcome and inclusion in contexts of cultural diversity are promoted, with increasing numbers of NI schools gaining this status (EA, 2020).
In short, although there are common contextual and policy factors with the rest of the UK, NI is different in several important ways, all of which have the potential to influence policy and practice at school level in relation to languages. First, there is a fundamental divide between maintained and controlled primary schools, most of which are populated largely by members of one of the two indigenous communities. Second, NI has a less diverse pupil population than elsewhere in the UK, albeit one where diversity is increasing rapidly; it is still in the process of evolving best practice in relation to community languages. Third, there is no statutory provision for primary languages in NI and all the evidence suggests that practice is patchy. Fourth, multi-sector provision against a strongly rural backdrop places particular financial constraints on the system which may have implications for school-level decisions around priorities. Fifth, the language landscape in NI's educational context involves indigenous languages as well as foreign and community languages, with a growing Irish-medium sector, particularly at primary level.

Fieldwork methodology
This study adopts an ethnographic approach to fieldwork (observation, interviews and focus groups), with the aim of allowing us to understand how the everyday language practices of various policy actors interact with macro-level policies. Data was collected 6 in two schools with very different socio-economic profiles. School A is situated in an area of high social deprivation which suffered greatly during the NI conflict. It has a recent newcomer population and a Free School Meals Entitlement 7 (FSME) percentage of almost 90%. School B is situated in one of the most affluent parts of the city and has a long-established tradition of educating children from international families. It has a FSME of less than 5%. The purpose of the fieldwork is not to make systemic generalisations on the basis of research in two schools but rather to use detailed ethnographic fieldwork with multiple actors in a school setting in order to gain close access to everyday practice in two socially-contrasting state schools.
Observational data were collected throughout the 2018-2019 school year. During the first phase (seven extended visits), the fieldworker (Anik Nandi) observed language practices in relation to newcomer students in various school contexts such as in the classroom, in the playground or when they were having a meal. This involved newcomer children's interactions with educators, classmates and with siblings or peers who speak the same language, as well as teachers' approaches to newcomer languages in the classroom. A second observational element of fieldwork involved organisation of a storytelling event for each school, where a professional storyteller worked with children to encourage them to tell a story in their home language. The objective was to gather observational data on the reaction of the children to their peers' newcomer languages through an enjoyable event which nonetheless gives a small insight into children's language ideologies because it values multilingual repertoires shaped by and representing history, culture, values and beliefs . All observational data was recorded in a notebook: there was no video or audio recording in the context of working with young children.
One-on-one audio-recorded semi-structured interviews of around 40 minutes were undertaken with the two school Principals, both of whom were receptive to the nature of the research being undertaken. A series of questions focusing on school policy, classroom practice and support for newcomer pupils formed the springboard for discussion. The Principals and the newcomer coordinator acted as gatekeepers for access to other staff with experience of teaching newcomers for participation in focus groups. Audiorecorded focus groups of around one hour were held in the schools to discuss policy and practice with a group of teachers and classroom assistants in each school: five took part in School A and seven in School B. We also intended to organise a parents' focus group (one hour) for each school in order to understand their perspectives on teaching/learning of their home languages. This was advertised through the school with the aim of obtaining volunteers. It is important to note here that whereas both parents from four families in School B participated in the focus group, it was impossible to enlist any parents from School A, despite several attempts with the strong support of the Principal (we shall return to this in section 'Socio-economic factors'). The School B parents' focus group was audio-recorded and held in Queen's University. In the case of focus groups, numbers are relatively small and there is always scope for unrepresentative views having undue influence and/or for statements about practice to be contradicted by reality. For these reasons, interview and focus group data were consolidated and validated through observational data as far as possible, and an audit involving basic details of respondents' language trajectories and usage, giving a holistic picture of both ideologies and everyday practices from all actors, i.e. Principals, teachers, parents and pupils. 'Thematic analysis' (Clarke & Braun, 2017) was deployed as the medium of data interpretation. This article draws particularly on classroom observation, the interviews with Principals as well as the focus groups with both teachers and parents.
From policy to practice: evidence from fieldwork Common issues with the rest of the UK An overall analysis of our data reveals several similar issues to those highlighted by research in the rest of the UK.
The 'language barrier': community languages as a problem The identification of community languages as a problem (Ruíz, 1984) and English immersion as the solution is widespread in our data. When asked about the school's language education policies towards newcomer pupils, School A's Principal clearly identifies home languages as inhibitors to the acquisition of English, without which pupils face a "language barrier". S/he states that the school intends to offer maximum exposure to English for pupils, so as to "submerge" them in the language of "our country":

Extract 1: School A (our italics in all extracts)
Principal A: … each teacher has an area of access on our online system to resources for our newcomer pupils so that when they enter the school, they're able to function as much as possible without the language barrier. We build on pupils' learning the language of our country which is English.
[ … ] we are trying to submerge our children in as much English as possible without trying to use their home language as we don't want them to be totally reliant on that and inhibit them from accessing the English.
Although the formulation of this discursive agenda is unquestionably well-intended and aimed at helping pupils to integrate and to access the curriculum, English is firmly positioned as the language with social, symbolic and cultural capital (Bourdieu, 1991), while home languages are viewed as a barrier.
In the context of School B, although about twenty different home languages are spoken in the school, the Principal acknowledges that a large majority of newcomer students come with high levels of English competence. For those without fluent English, School B also retains an explicit immersion policy and introduces a range of language management strategies involving peer-learning, tutorials with educators, one-to-one language support through classroom assistants (henceforth CA) when necessary and 'buddying': Extract 2: School B: Principal B: Nearly all the children when they come, speak English fluently. If they don't speak English with fluency, [ … ] we have a period in which they have to be immersed in the language [ … ] We'll just be giving them as much opportunities as possible to mix with children who have mastery of English already.
In short, in our data, community languages are never described in terms of their benefits; they are normally described as a barrier to success.
Monolingualism inside and outside the classroom A corollary of the English immersion policy is that there is little or no discussion of the multilingualism of the pupils and the relevance of this at certain points to classroom discussion. For instance, observational fieldnotes taken from a numeracy class on 20/11/ 2018 in School B notes that the teacher corrects a pupil's mathematics on the board when, in response to the task of giving two (different) numbers between 1 and 5 that add up to make 6, the pupil offered /٤/. The teacher corrected /٤/which is Arabic for 4to '3', assuming that it was a mistaken attempt to write '3'. This scenario meant that the pupil's numerical knowledge in their home language was overlooked, as was the benefit of retaining knowledge of the meaning of the symbol in Arabic. Indeed some forms of language management can lead to a gradual loss of their acquired knowledge of children's respective heritage cultures (Wei, 2018).
Teacher 3 in the focus group argues that newcomers' competence in home language(s) is a hindrance to developing their proficiency in English as they often do not speak English in the home domain: Extract 4: School B Teacher 3: I have a little boy in my class, and they speak Greek at home all the time. I do think there's need for them to speak more English because with his reading, even his pronunciations or some other words he knows, he says it in a funny way, but it's just his language.
Teacher 2: Err, children can get caught in between the two languages and not be proficient in either.
Teacher 5: But sometimes what happens, like I mean, they may be able to speak and write in it (English), but they can't read in it. Because nobody is teaching them the reading at home. They are being taught to read in English at school. But unless the parents and teachers help them in English, and not in their home language, they're not going to have that … Teacher 3: … proficiency. Or else they don't speak English at home and still they are reading in English, but they are not good at it. So that is difficult.
Various aspects of Ruiz's 'language-as-problem' orientation become prominent in the above extract where major reservations are expressed about parents speaking their home language(s) inside the family. There is an unfortunate and erroneous connection made between parents' bilingual practice and academic difficulties (cf. Ruíz, 1984, p. 19).
Language hierarchies between English, modern foreign and community languages As noted above (section 'Language hierarchies in mainstream education'), teaching/ learning of European languages is generally considered as a resource and is often valued over the maintenance of community languages. For instance, School B is proud to teach French not only as part of the school curriculum in P6 and P7, but also through a privately funded after-school French club: Extract: 5 School B Principal B: … the language that we would teach in the school would be French. So, there's a French club that also operates in the school … […] … that happens once a week throughout the year. The children get into primary 6 and 7, we actually pay a teacher, French teacher to come in and do an hour of language, each week with each class. So, by the time the children move on to post-primary school, they've already had two years of French an hour a week.
Interestingly, having French as a home language is seen much less favourably, especially if it is perceived as the source of any educational issues with the child. For instance, during the parents' focus group, a multilingual family comprising a Spanish-speaking father and a French-speaking mother was interviewed (they follow a 'one parent one language' policy). Although the child is proficient in both Spanish and French, s/he went through a 'silent period' in the early school years and the parents were advised by the school to speak "only in English all the time" at home to compensate for their child's English language deficiency: Extract 6: School B Father: We had an interview with the principal, and he recommended us that "forget about the Spanish and French and just speak to him, only in English all the time".
In this type of context, the monolingual exterior may actually penetrate the heritage language-speaking family's interior home space and affect everyday language practices. In this case, the parents chose not to follow the advice, further underscoring bottomup discourses of resistance towards external language governance.

Issues of training, resources and confidence amongst educators in multilingual classrooms:
Teacher training in NI is spread across several institutions and raises similar issues to those discussed above in 'Issues of training, resources and confidence amongst educators'. Our discussions both with teachers and with staff at the EA would suggest that most current programmes contain training in teaching in a multilingual environment and CPD training is also on offer. However, the extent to which more established staff avail of CPD in this area is highly variable, as is the way in which bi-and multilingualism is promoted in individual schools (Jones et al., 2017). Both our fieldwork and our discussions with the EA would suggest that valuing and promoting home languages is probably the element of good practice that is least frequently implemented, despite its presence in training and in the Toolkit. Policy and practice in individual schools are highly dependent on the view of the school leadership and can therefore be patchy.

Practice in NI: context-specific factors
As mentioned above (section on 'Community languages in NI'), the policy context in NI differs in some significant ways from elsewhere in the UK. The combination of a lack of statutory requirement in terms of what the school is required to deliver (and on which it is inspected) and the financial pressures within which decisions around prioritisation are taken, is evident in the following extract from the Principal of School A who describes languages as a luxury: Extract: 7 School A Principal A: It's difficult because there are huge pressures put on us, as a school, to teach the core subjects and we are only inspected on that and because our inspection process is so heavily focused on our outcomes, our results and what we are producing in terms of the mandatory things that we have to do. Unfortunately, things like a second language is kind of a luxury thing to be able to take on, and if you don't have a staff member that is maybe willing to do that or take that on.
This point of view is strongly aligned with the findings of one of the latest large-scale studies in this area (Jones et al., 2017) which concludes that although the majority of school principals value language learning, "not all school principals and teachers see additional language learning as a key priority, given its absence as a statutory element of the Northern Ireland Curriculum" (p. 2). While the factors mentioned by Principal A relating to the focus of inspections are common across the UK, the non-statutory status of primary languages is unique to NI and further increases the chances of languages being viewed as a luxury rather than a necessity. Moreover, as is clear from Extract 7 above, where language learning is mentioned by Principals or teachers in a primary context, or where a lack of confidence or training on the part of teachers is discussed, this is entirely in relation to French, Spanish or Irish. In terms of provision, our two schools represent two ends of the spectrum: School B employs a teacher of French for certain classes and runs a fee-paying after-school club; the other (School A) cannot contemplate the luxury of teaching languages.
In terms of teaching in a multilingual classroom, our teacher and CA focus groups told us clearly that they are familiar with the Toolkit and that they use it frequently in developing teaching materials. It was also clear that bilingual CAs were deployed to support children and parents: Extract: 8 School A Principal A: So, we have a classroom assistant that is from Poland, but of the other languages, we had a parent who had come in from a country … .[…] … quite close to Poland and we had a parent come in with a concern and because there is a common, sort of some common phrases in that other language, she was able to understand what this parent needed and we were able to then, help that child.
In School B, there was excellent knowledge of which languages were spoken at home and strong implementation of policies such as buddying and effective use of the newcomer co-ordinator and bilingual classroom assistants: Extract: 9 School B Principal B: … we would ask that classroom assistant to take a number of sessions in a oneto-one with the child with some of the work that the child is doing. We would get that classroom assistant some basic training in the sort of activities that she should be doing. That training would be provided by our newcomer coordinator. She would take the classroom assistant aside and talk her through the Toolkit and the resources … However, in neither school did we find evidence to suggest that the Toolkit's recommendations are being implemented in relation to the maintenance of and support for home languages. None of our discussions with Principals, teachers and parents, and none of our classroom observation would suggest that community languages are viewed as a 'resource' (in Ruiz's terms), either for the school or for the individual pupil, in terms of his/her development, wellbeing or acquisition of English (extracts 1, 8 and 9) beyond, for example, a Welcome notice. Principals and teachers were very focused on what they saw as the major priorities for the pupils' wellbeing, integration and progression, i.e. access to the curriculum and good educational outcomes in the core skills of literacy and numeracy.
It is important to say that there is clear evidence of the promotion of cultural, ethnic and religious diversity (in the sense of respecting different religious backgrounds such as Christian, Muslim, Hindu, Jewish). In schools such as School B, where there is an established and largely middle-class diverse population, there are regular events such as international food festivals and the marking of important cultural or religious dates such as the Chinese New Year, Diwali and Eid. Such events are also beginning to happen in School A and sit well with the statutory requirement in NI primary schools for 'PDMU' (Personal Development and Mutual Understanding), where one theme involves teachers enabling students to develop knowledge and understanding of 'Similarities and Differences' (CCEA, 2019). Nonetheless, knowledge or promotion of community languages does not feature as part of this celebration of diversity and is not mentioned in the PDMU guidelines which cite culture, heritage, traditions, food, clothes, symbols, celebrations, gender, disability, hobbies, race, religion, sporting interests, abilities and work. Moreover, it has been argued elsewhere (in the Australian context: Arber, 2008) that this type of event can be tokenistic or strongly owned by particular communities rather than fully inclusive. In any case, all the evidence suggests that in practice, respect for diversity means all types except linguistic diversity. The 'blind spot' in educational terms is that promotion of the value of community languages and intensive English language learning are seen, despite the clear recommendations in the Toolkit, as mutually incompatible. For example, in School B, which implements many of the positive recommendations such as buddying and has an impressive record of promoting cultural diversity, e xtracts 2 and 10 suggest that the benefits of maintaining the home language have not been taken on board: Extract: 10 School B Teacher 3: … But what we find is then the children will go home and very often not speak any English at home or go away for the whole summer and not speak any English at all and then come back in and have to cope with the things! I feel for the children.
Our fieldwork would suggest that policy and practice are firmly focused on a twopronged strategy of 'respect for diversity ' (i.e. cultural, ethnic, religious) and 'English immersion for the purposes of accessing the curriculum', with community languages losing out in both cases.
Finally, the legacy of the conflict in NI, and the sharp binary divide in the education system, can make linguistic diversity more challenging for some schools, particularly those in areas of high social deprivation, such as School A. For the duration of the conflict, outsiders of any description would not have settled in such communities which were often dominated by particular actors, in this case, the UDA and UFF, with loyalist murals still prevalent on the walls. In other words, not only would there have been few if any pupils from migrant communities, but also, the school community would have been rooted exclusively in one of the two NI communities, in this case, PUL. For many parts of Belfast, including the area where School A is based, the latter has not changed despite the peace. There is strong evidence that housing is no more mixed in terms of PUL and CNR communities in many areas than it was during the conflict (Capener, 2017). The changes brought through recent migration are therefore all the more dramatic as they are taking place in an otherwise highly monocultural and monolingual community. Moves to embrace diversity in any shape or form in schools such as School A require great sensitivity. The following teacher discussion demonstrates how newcomer students' language practices are continuously being negotiated on the ground as their monolingual counterparts feel daunted or even offended by their multilingual repertoire: Extract: 11 School A Teacher 5: And we would also find the children from here become very easily offended, you know, no matter what is spoken there in a different language the children who are at that table are very, sort of, offended by them … that's my experience you know. At times, I thought that the children got a little bit paranoid when they* talked mainly in their own language, like what are they saying … makes sense? That's maybe where the bit of offense comes from not that they mind that they're speaking in their own language, they just get a bit paranoid as if what's that. (* = two newcomer pupils) Such challenges are much rarer in schools such as School B which is situated in one of the few longstanding areas of mixed-community housing in NI terms and has always had at least some element of PUL-CNR mixing in the pupil population.

Socio-economic factors
There are further differences between the two schools that are related specifically to socio-economic factors although clearly, we cannot extrapolate upwards to a more general pattern. Nonetheless, the differences resonate with wider patterns that suggest socio-economic distinctions in terms of access to, time invested in, and attitudes towards the value of languages both in the UK (Coffey, 2018;Collen, 2020) and internationally (Cruickshank et al., 2020); type of school (independent, selective etc.), parents' SES (measured by levels of FSME) and questions of cultural heritage have all been shown to play an important role.
First, in a context of no statutory languages provision at primary level, attitudes to teaching a language are quite different in our two schools. For the Principal of School B, the teaching of French in Primary 6 and 7 is an important preparation for the transfer to secondary school (which is overwhelming to selective grammar schools) and the initiative is supported by parents. Many of those grammar schools will offer three or four languages at KS3 (in some cases a language is compulsory at KS4) along with the possibility of taking a GCSE in a community language, so language learning at primary school is the first part of an ongoing trajectory. As Coffey (2018), drawing on Bourdieu, shows in a different context, questions of economic and social capital come strongly into play here. Many families attending the school are regular travellers to Europe and beyond; they aspire to international mobility for leisure and even potentially for employment in the future and language learning is highly compatible with that world view. In School A, by contrast, where no languages are taught, the pupils only rarely transfer into the grammar school sector and do not come from families who take regular holidays abroad. The focus in School A is on core provision and on issues such as ensuring attendance.
Second, levels of parental engagement with their children's learning are very different in the two schools, with the latest ETI inspection highlighting 'outstanding links and partnerships with the parents' in School B which is categorised overall as 'outstanding'. Many pupils in School B have high levels of economic and social capital, including linguistic capital in the form of being bilingual speakers of a foreign language and English. This is not the case for School A which has a much more recent migrant community, where many of the parents are not speakers of English and have other obstacles to close engagement with the school such as childcare. These differences are reflected in the fact that whereas we recruited parents from School B easily for fieldwork focus groups, we tried several times, ultimately unsuccessfully, to convene a parents' group in School A (similar difficulties in engaging parents from lower SES groups were encountered by Cruickshank et al., 2020). Levels of parental empowerment in an educational context are also related to questions of economic and social capital. Thus in School B, for example, when the Principal suggested an 'English only' approach at home where a child's silent period was long, the parents felt sufficiently empowered through their own knowledge about bi/multilingualism to override this recommendation (extract 6).
Third, the opportunities for introducing a linguistic dimension into other celebrations of diversity are much greater in the socio-economically affluent school than in the school with high levels of FSME. Parents are often in a sufficiently comfortable socio-economic position to give time and/or resources to participate in cultural activities which are much rarer in School A. As part of our engagement with School B, we incorporated a linguistic dimension into one of their major cultural events, i.e. International Day, through storytelling that showcased the multilingualism of the pupils. Bi-or multilingual parents were also encouraged to tell a story in their own languages, provoking stories in Irish, Hindi and Arabic. Feedback was extremely positive, with a large majority of respondents agreeing that linguistic diversity should be valued and promoted in the school sector. The possibility of incorporating a linguistic component into such events was a case of augmenting an already-existing celebration of cultural diversity by working with a group of well-disposed parents, an open-minded Principal and a group of children who, although lacking in confidence at the beginning of the exercise, were able to develop their skills in a safe environment. It would have been much more difficult at this point in time to attempt to incorporate community languages into an event in school A: such events do not yet form part of the tradition of the school, newcomer children are a much more recent phenomenon and it is much more difficult to get parents involved in extra-curricular cultural events of any type. Support for community languages in NI: research and policy issues for the future The position of community languages in NI has much in common with GB and indeed with other international anglophone contexts such as Australia (Cruickshank et al., 2020). However, it is clearly also tightly connected to several other cultural, educational and social issues that have particular resonances in the NI context. Moreover, the recent surge in numbers of newcomer pupils is such that the need for effective policy and practice has come sharply into focus. The documentation attached to the current consultation in this area suggests strongly that DENI is aware that current policy and practice is not entirely fit for purpose. There is an opportunity to build on the high-quality resources that have been developed and promoted by the EA, notably the Toolkits, and to draw on knowledge of good practice in international contexts (notably in continental Europe) that is evident in the DENI consultation documentation.
A number of broader policy developments would reinforce the value of community languages, several of which are mentioned in Ayres-Bennett and Carruthers (2019). A statutory primary languages programme for Northern Ireland, as exists in GB and most of the rest of Europe, would significantly raise the profile of languages, although it is important not to create a hierarchy that might place community languages at the bottom of the scale. For this reason, consideration should be given to building an integrated strategy across the different groups of languagesforeign, community and indigenousin order to connect the different sectors (including building links between mainstream and supplementary schools) and promote language learning and multilingualism across the board. Although Irish is treated differently in RoI (as an official language with longstanding statutory provision), the Languages Connect strategy is a good model of an attempt to integrate foreign and community languages in a context where there has also been major growth in the number of newcomer pupils. Similarly, there are now opportunities to learn from both the benefits and challenges of the Scottish '1 + 2' strategy which integrates Gaelic, foreign and community languages. Finally, it has to be acknowledged that the most historically entrenched problems in NI society also bear on the position of community languages and resolving these will be a longer-term political objective, notably the community divisions that cut through the education system and housing. These divisions continue to act as a barrier to progressive government and have a disproportionately negative impact on opportunities for children in communities where levels of social deprivation are high.

Notes on contributors
Janice Carruthers is Professor of French Linguistics at Queen's University Belfast and the Arts and Humanities Research Council (AHRC) Priority Area Leadership Fellow for Modern Languages. Her research is situated mainly in sociolinguistics, including temporality, variation and orality in French, as well as language policy in both France and the UK.
Anik Nandi was Postdoctoral Research Fellow at Queen's University Belfast, on the AHRC-funded Priority Area Leadership Fellowship in Modern Languages, held by Janice Carruthers; he is currently based at the Seminario de Sociolingüística, Real Academia Galega, A Coruña, Spain. He specialises in interdisciplinary approaches to study language policy and literacy practices in multilingual societies.