Exploring the ESOL-PSIT relation: Interpellation, resistance and resilience

This article explores the relationship between English for Speakers of Other Languages (ESOL) and Public Service Interpreting and Translation (PSIT) and the problematic juxtaposition of the two in Government discourse. I analyse the discursive construction of the ESOL/PSIT subject at different levels and problematise language support provisions as acts of interpellation, drawing on quantitative and qualitative data collected in conjunction with associates at Multilingual Manchester that shed light on lived experience of provisions. I conclude that any city-region language strategy needs to address three interrelated issues: the problem of universalising approaches to subject formation, the supply and organisation of provisions; the politicisation of ‘resilience’. Crown Copyright 2018 Published by Elsevier Ltd. All rights reserved. ‘Whatever the forms of exile, language is what one keeps for oneself’ Jacques Derrida (1997/2000)


Introduction
The relationship between language, immigrant assimilation and integration in Britain has been the subject of considerable scholarly and public debate since the early 2000s. It has been shaped by critiques of multiculturalism, changes in the nature and extent of linguistic and cultural diversity in particular geographical areas, and changes to the funding of public services. For Resnyansky (2016), these debates have been to some extent stymied by views of language that appeal more to common sense (folk) ideas and shaped by 'metaphors and uncritically adopted assumptions' (p. 2053), thereby forming a questionable basis for policy making.
(English) language acquisition or English for Speakers of Other Languages (ESOL) and Public Service Interpreting and Translation (PSIT), may be conceptualised as vehicles for immigrant assimilation and integration, where assimilation is conceived as a form of absorption into a new society of residence and integration is a two-way process involving migrant and receiving society (Strang et al., 2018;Berry, 1997). Their role as distinct, yet interconnected practices and the political formations in which they operate are under-theorised in academic research, and also in relation to new drivers of policy in Britain, such as the notion of community resilience and the shifting of risk to local government and households. Even in my home discipline of translation and interpreting studies, the tendency to examine PSIT in discrete terms, as opposed to part of a wider ecology of language support provisions has to some extent limited how we understand their contribution to service access and wider social participation.
A dichotomous and under-theorised approach to the relation between ESOL and PSIT emerges from academic and state discourses in which the two are juxtaposed. Aspinall and Hashem (2011) for example, discuss the two in such terms claiming that government emphasis had moved away at the time of their writing from PSIT towards ESOL. However, they fail to provide a clear point of reference for the shift. To my knowledge, no written policy covering all PSIT provisions at state level has ever existed in Britain, and the lack of cross-portfolio policy making (e.g. Ozolins, 2010) on such matters is well known. 1 In general terms, the debate is characterised by neglect of key questions about how subjects are 'interpellated' in their limited language proficiency and the extent to which approaches that appear to universalise the immigrant subject-position have political utility when it comes to the organisation of language support provisions at the local level. In what follows I firstly explore the antecedents to the juxtaposition of PSIT and ESOL at Government and local authority political discourse before recasting the relation through the lens of 'interpellation' and examining examples of the lived experience of both in the city of Manchester. I argue that if language provisions are to strategically support the development of resilient communities, their conceptualisation and organisation needs to take account of how individuals actually engage with them and, perhaps more importantly, why they may resist them as top-down social interventions.
The discussion draws on research carried out under the Open World Research Initiative funded by the Arts and Humanities Research Council, UK, by a team of researchers, including the author, at Multilingual Manchester (MLM) 2 for which ethical approval was granted by the University of Manchester. These include a survey of a cohort of Entry 3/Level 1 ESOL learners (n ¼ 340) with Manchester Adult Education Service, and interviews with eight self-selecting respondents following the survey and with tutors. This article focuses on the PSIT-related findings of the survey; the full findings of the survey are available at MLM (Vasey et al., 2018). The discussion also draws on individual and focus group interviews which were conducted in parallel to the survey at various locations offering informal ESOL learning in the city, involving speakers of different languages: Farsi, Kurdish Sorani, Spanish, Arabic (Syria/Algeria), Urdu (Pakistan), Cantonese Chinese, Spanish (Spain/ Venezuela), Polish, French (Cameroon/DRC), Swahili, and Shona. Respondents (n ¼ 50) were aged between 18 and 60 þ and the amount of time spent in the UK varied between 6 months and 40 years. Although respondents were not asked directly about their immigration status, it became clear from the discussions that the sample included asylum seekers, refugees and immigrants from countries in the European Union. Participants attending informal ESOL provisions tended to have a lower overall level of proficiency compared to the survey sample, which meant that interpreter mediation was made available. Recruitment was achieved through contacts at local NGOs, and chain-referral sampling, which led to a good range of language and cultural backgrounds but limitations are acknowledged in terms of the representativeness within groups in terms of age range and gender.

ESOL-PSIT: a zero sum game vs continuum-based approach
It is commonly held that as ESOL provision is taken up reliance on professional PSIT provision diminishes, thereby creating a zero sum game. However, this is an oversimplification of what in reality are complex practices within complex organisational frameworks. For instance, the profile of English language learners is hugely variable, as indeed are levels of motivation and capacity for learning (e.g. Finn, 2010), making the tailoring of ESOL offer particularly challenging for course providers. 3 Relatedly, there is an assumption that provisions will be taken up where available; however, research shows that uncertainty over immigration status can undermine efforts to engage with learning opportunities (see Kelly, 2004).
Myriad organisational and structural issues also impact PSIT provisions, and there is generally poor public understanding of their function. In Britain, translation and interpreting are unregulated professions 4 and opportunities for specialist and language-specific training and education are still very limited, leading to variable service quality across domains and settings. 5 PSIT provisions are promoted as an instrumental language right in the legal setting by virtue of inter alia EU Directive 2010/64/EU on the right to interpretation and translation in criminal proceedings; however, even in this setting, criticisms over outsourcing by the Ministry of Justice 6 mean that the concept of 'fair access' to services needs to be critically appraised.
To understand the relation in all its complexity, concepts such as community, social capital and belonging have analytical import, requiring engagement with a broad range of disciplines in the humanities and social and political sciences. For example, research on language and integration in the social sciences has indicated that a continuum-based approach to L2 acquisition and interpreting and translation provisions more accurately reflects lived experience than considering PSIT as prior to ESOL (e.g. Resnyansky, 2016;Strang, 2004, 2008); to my knowledge, however, there have been no systematic investigation of how such a continuum operates or how it is experienced.

Political antecedents of the zero-sum game: PSIT
PSIT has long been blamed for slowing integration (e.g. Easton, 2006;HM Government, 2012) -however integration is defined -and perpetuating a long-term need for, and even dependence on such services. Such social commentaries may be read as part of the broader critique of multiculturalism, but they are not commonly supported by compelling evidence and tend to adopt a limited perspective on what constitutes translation. Individuals can find themselves in situations in which their 'truncated multilingualism' (Blommaert et al., 2005) and the lack of access to codes that enjoy power in certain spaces (Canagarajah, 2013, 204), cumulatively or even punctually, may be the difference between personal safety and harm; however, the vulnerability that derives from limited language proficiency is not commonly brought into the public debates on service provisions.
Recent history shows successive UK governments from across the political spectrum grappling with the nature and scope of interpreting and translation provisions, leading them to send mixed, if not incoherent advice to local authorities. For instance, in 2008 the New Labour government, through the Department for Communities and Local Government, produced Guidance for Local Authorities on Translation of Publications in which the following appears: Opinion is divided as to whether translation is a barrier to integration, or whether it is a stepping stone to better language skills. Our position is that it depends on the individual: where migrants from the past are still relying on community languages, then translations from English are likely to extend their reliance on their mother tongue; where new migrants do not speak English then clearly they need initial information in appropriate languages. Local authorities will judge what is bestbut our working assumption is that heading for the translators should not be an automatic first step in all cases (2008, p. 11).
The guidance stemmed in part from large quantities of written translated materials being unused and a desire to target resources more effectively. However, the comments in the guidance by the then Minister for Communities and Local Government, Hazel Blears, fail to clarify, as she observes: 'Of course, there will always be some circumstances in which translation is appropriatefor example, to enable particular individuals to access essential services like healthcare or participate in local democracy. That is why local authorities are responsible for making the decisions which reflect the needs of their community. ' (2008, p. 5).
What is striking about the text cited above and Blears' comments is the absence of reference to public services' obligations in relation to equality of access and the lack of evidence base to corroborate the claim about extending reliance on the 'mother tongue', a problematic term in itself. Blears also conflates translation and interpreting, and her mention of provisions supporting participation in local democracy is a source of further confusion owing to its vagueness; it risks perpetuating the myth that State-sponsored PSIT provisions are available for everyday encounters -the opposite to Blears' intended message.
The Coalition Government that followed New Labour in 2010 continued to assess PSIT provisions and juxtapose them with ESOL, as illustrated by following quote from a text promoting the launch in 2013 of a £6-million competition fund to support English language learning: In the past there has been the tendency to pay for interpreters, or translate documents into foreign languages, for those with no or extremely poor language skills. This undermines community integration and encourages segregation. 7 The approach is further promoted in the 2018 Green Paper on the Integrated Communities Strategy. (37) in the context of the rationale for boosting English language skills: Low levels of proficiency also create costs for providers of local services, such as local authorities and health providers, which have to pay for translation of information and may impact on others in the family, including children, who have to act as translators for relations or friends who cannot speak English.
The narrative is one of a threat to moral and social order; in addition to the claim about children having to act as translators, the Secretary of State, Sajid Javid's own experiences interpreting for his mother as a small child -recounted in his introduction to the 2018 Green Paper -appear to confirm it as a norm in contemporary society. In terms of subject formation, the discourse constructs the PSIT service-using subject as socially passive, disengaged and marginalised. While this description may apply in some cases, it is presented as an unhelpful generalisation and may be taken as an example of what Spolsky (2004) terms policy at the level of beliefs, or policy as discourse (following Ball, 1993;Bonacina-Pugh, 2012).
At the city level, by contrast, an arguably more coherent discourse emerges. For example, in 2016, a report to the Communities and Equalities Scrutiny Committee at Manchester City Council on the Council's in-house translation and interpreting service (M-Four Translations) highlights the Council's legal responsibilities in relation to a variety of equality measures. The Council's Communication Support Service (of which M-Four translations is part) 'supports equal access to services and information for all Manchester residents regardless of their race, culture or disability' (p. 3), as a clear reflection of its duties under equality legislation. It provides services for both internal Council services (in 2015-16 this chiefly concerned the Children and Families service) and external services (in 2015-16 these chiefly concerned health and social care providers, and the National Driving Improvement Scheme).
The data source shows that although M-Four is one of many providers of translation and interpreting in the city, the source and overall volume of requests from particular services help to counter perceptions that provisions are sought for everyday encounters. The Children and Families service, for example, addresses a broad spectrum of issues including domestic violence and homelessness, that is, service encounters that have less to do with 'integration' understood in terms of empowerment to socially contribute and achieve potential (as indicated by the Home Office in 2005, cf. Phillimore, 2011), and more to do with crisis management.
In sum, examining the different discursive levels it is possible to discern significant differences in approach to subject and policy formation. At State level, attempts to universalise the PSIT service-using subject are evidenced, but close critical reading of pronouncements by administrations from across the political spectrum shows inconsistency in presenting the scope and function of PSIT provisions. This inconsistency risks reinforcing stereotypes and underplaying the importance of professional provisions in the discharge of statutory services. Furthermore, emphasising the link between PSIT and integration tout court detracts from their wider social benefits (e.g. public safety, security, and individual well-being); the example of the local narrative in the Manchester City Council reports shows a much more targeted focus on social need and crisis management.

Political antecedents of the zero sum game: ESOL
Like PSIT, ESOL in England has developed unevenly in terms of its professional status (cf. Macdonald, 2013) and there is still no national ESOL strategy for England at the time of writing. Han, Starkey and Green (2010, 64) describe the 'sudden politicisation of what had been relatively low status courses [in English language learning]' from the early 2000s. As others observe, such politicisation was driven in large part as a response to social unrest in certain towns in the north of England, following which poor English language was cited (erroneously according to some) as a source of social disruption and even a threat to national security (cf. Blackledge, 2006;Cooke, 2009).
Since the early 2000s, ESOL provisions have been marked by various ideological agendas, with Cooke and Simpson (2009, 1) describing it as having become 'a receptacle for policy on skills education, employability and citizenship' (see also Cooke, 2009). Viewed as a 'skill for life', ESOL provisions were centralised in the early 2000s (Baynham and Simpson, 2010), but demand was such that centralisation was deemed unsustainable, leading to New Labour's New Approach to ESOL (HM Government, 2009) and its coordination by local councils and authorities (Simpson, 2012). ESOL provisions have also been impacted by the changing funding landscape and eligibility criteria. A recent House of Commons briefing paper suggests that funding declined by around 60% in real terms between 2009/10 and 2015/16, with available places falling from 180,000 to 100,000 per annum in the same period (Foster andBolton, 2017, 3, cited in Vasey et al., 2018). Further, the Casey Review 2016 highlighted shortfalls in funding for pre-entry and entry levels (p. 97).
Additional funding for ESOL was secured in 2013 through a £20 million funding scheme targeting Muslim women in an attempt to counteract what was described as 'passive tolerance' of separate communities and extremism. This discourse continues the post-9/11 shift away from an emphasis on diversity and tolerance to policies that promote integration and social cohesion. Crucially, it has been accompanied by a further ideological shift towards community-based provisions in ESOL (cf. 2018 Green Paper Integrated Communities Strategy) as part of a community resilience-based approach to public services.
In general terms, community resilience concerns individuals' 'ability to cope with adversity over the long term and manage risk' (Platts-Fowler and Robinson, 2016, 764), but it is commonly associated in the contemporary period with austerity politics. Certain groups of immigrants may be described as having already developed high levels of resilience by virtue of the challenges posed by conflict and trauma in their life trajectories; however, individual repertoires of resilience have their foundations in very different forms of capital (social, cultural and economic) to the local population, thereby requiring attention by policy makers to a broader range of variables and life experiences, which have consequences for L2 acquisition.

ESOL provisions in the city of Manchester
The majority of ESOL provision in Manchester is delivered by Manchester Adult Education Service (MAES) (in terms of both the number of classes and enrolments) and the Manchester College, via Skills Funding Agency funding, with a small amount delivered by the Workers' Educational Association (WEA), and the Voluntary and Community Sector. The Manchester ESOL Strategy, launched in 2016, is designed to support the Manchester Strategy 2016-2025 which focuses on delivering a 'progressive and equitable city' and a 'highly skilled city'. The strategy is also linked to the Manchester Work and Skills Strategy 2016-2021 (approved in January 2016), which is structured around three themes: Business and Enterprise, Resident Skills and Addressing Inequality. There is a clear emphasis on assimilation through employment and skills training, and a highly instrumental approach to language learning, although investment in community-based, and largely volunteer-led initiatives led by Talk English under the auspices of MAES show that a broad range of learner needs are being addressed.
Of particular interest to the discussion here is the contrast between local city-level ESOL narratives into which the Manchester ESOL Strategy feeds, and Government narratives discussed above. Researchers at Multilingual Manchester, for example, observe that the city council narrative positions ESOL as part of its celebration of equality and diversity and, crucially, highlight its potential to support resilience-building, that is, the development of social capital and skills that enable active participation (Vasey et al., 2018, 6). It is also striking that publicly available MAES reports to various city council committees make no mention of PSIT provisions (whether its own service, M-Four Translations, or others). This suggests that within Manchester City Council, the ESOL-PSIT relation is not juxtaposed as it is in Government discourse, suggesting that practice at the local level is shaped by a continuum-based understanding of service user engagement with the two types of provision.

Theorising language provisions as acts of interpellation
The discussion above highlights tensions between local and national level discourse about the purpose and positioning of ESOL and PSIT in the city's ecology of language support provisions. The discursive formation of both the ESOL and PSIT subject can serve as a first stage in formulating a political response (entextualised in policy 8 ), allocating resources and creating the social structures necessary for their realisation. However, a broad-brush approach papers over factors that can impact on the effectiveness of the policy intervention, among which one might cite examples of poor (digital) literacy levels among some ESOL learners and mismatched dialect pairing between interpreters and service users.
Choice is also key here. Cronin (2006, 53), for example, talks about the individual's 'right to exercise autonomous forms of translation' (i.e., putting the immigrant in control of the translation process), and contrasts autonomous forms with heteronomous forms through which others control the translation process. Both ESOL and PSIT may beand indeed areconceived as forms of linguistic opportunity, involving different types of translation, but their successful implementation depends on the nature of individual choice afforded by their provision. Whether -and for what gainboth forms of language support can be considered as interpellative acts and what implications this has for individual choice and social outcomes is explored in the next section.

Interpellation
The concept derives from Althusser's (1971) doctrine of interpellation, which is often described as the way in which ideologies lead us to apprehend cultural values and internalise them. According to Butler (1997, 106), Althusser's concept '[offers] a way to account for a subject who comes into being as a consequence of language, yet always within its terms'. Althusser exemplifies the concept through the vignette of a police officer who verbally 'hails' a subject: 'Hey you there!', and the subject turns round. In the act of turning, the subject is deemed to accept the 'terms by which he or she is hailed' (Butler, 1997); in this case, the terms mean to accept guilt 'to gain a purchase on identity' (Butler, 1997, 110). Butler (1997, 107) describes the turning round as a 'strange sort of middle ground', asserting that '[a]lthough there would be no turning around without first having been hailed, neither would there be a turning around without some readiness to turn'. She adds (p.109) 'the turn toward the law is not necessitated by the hailing; it is compelling, in a less than logical sense, because it promises identity'. In the case of limited language proficiency, the promise of identity is further complicated because of the language in which the hailing takes place, and because of the fluid nature of identity.

Interpellation and translation studies
In translation studies, Cronin (2006) is one of the few scholars to construe PSIT in interpellative terms, although he takes inspiration from Stuart Hall's concept of 'articulation' (Hall and du Gay, 1996) rather than Althusser. Cronin's concern is with how new arrivals with limited language proficiency in the majority language respond to their new linguistic situation, thereby placing emphasis on intercultural contact understood in terms of the minority-majority relation. Hall's concept of articulation, which assumes 'things are connected as much by their difference as by their similarity', for Cronin is 'central to the core definitions of translation itself ' (2006, p. 63).
Addressing someone in a language they understand encourages the person to 'turn'; in Cronin's terms, this makes it 'easier to invest in the subject-position of intercultural contact' (p. 63). In this case, the readiness to turn is not predicated on the idea of 'guilt' as Butler (1997) suggests in relation to Althusser's example, but what replaces it is not immediately obvious. 'Friendship' might be considered plausible, but this would not necessarily lead to the desired type of investment in the subject-position. In other words, there is an underlying risk that hailing in another's language may be taken as a sign that the relation will always operate thus, serving to disincentivize language learning. It also risks normalising the subordinate subject-position of the immigrant subject, diminishing agency and reinforcing stereotypes. On the other hand, it can be said to create a coherent subject for the purpose of meaningful initial interaction.
An alternative reason for the readiness to turn is potentially captured by the term 'hospitality' as a form of care-oriented cosmopolitics (following Derrida, 1997Derrida, /2000Watson, 2014). State-sponsored PSIT provisions may be construed as an 8 Policy is understood here in terms of a written statement setting out the parameters of provisions and detailing their organisation and delivery. example of unconditional hospitality since access to services is facilitated without any expectation of a return, thereby overlapping with Derrida's view of hospitality as a fundamentally ethical act. But there is arguably a return of some kind that is underplayed in discourse on PSIT provisions; that is, each interpreted encounter goes beyond a simple transactional exchange (e.g. a medical diagnosis) an entails some form of learning (about institutional structures, societal values, etc); in short, it is often as much pedagogical as it is transactional (Tipton, 2018), which is why it can be considered an interpellative act. The readiness to 'turn' therefore because someone speaks your language can therefore be considered as a moment of resilience building, of becoming a host in Derrida's terms, since it is not predicated on 'responsibility to a debt or motivated by a feeling of guilt' (Floriani and Schramm, 2010, 217); in turning, the individual is empowered to cross a threshold into a social space as an individual with an identity and some sort of stake in that space.
But what is to be gained from an investment in intercultural contact and why would an individual embrace this subjectposition? Cronin is careful to suggest that investment in the position remains a choice and that there may be certain situations rather than others in which it is beneficial. He cites Benhabib (2002) on the importance of commitment, accountability and deepening attachment in the context of democratic citizenship, as one example of the benefit of investing. There are also very tangible and practical consequences for the individual and wider society, for example in healthcare contexts, where timely PSIT provisions have been proven to prevent hospital readmissions and the need for more costly healthcare interventions (e.g. Lindholm et al., 2012). However, interpellating the individual through an offer of professional translation requires of that individual that they recognise their level of vulnerability and consciously choose to hand over control of translation to someone else. For many reasons, they may choose to resist as I discuss later.
A complicating factor in PSIT is that the formation of the PSIT subject starts from an assumption of zero, or almost zero level of proficiency in the majority language; while this may be true in some cases, it does not apply to all service users. In the survey of Entry 3/Level 1 learners, 77.5% (n ¼ 336) reported having received some tuition in English language prior to arrival (Vasey et al., 2018). Further, anecdotal reports of a domestic violence support service in Manchester (Tipton, 2018) strongly suggest that many women under-estimate and under-report proficiency levels, and can typically understand more than they can communicate on entering the service. This does not make a case against PSIT provisions, but it is significant because it suggests that the strategic organisation of provisions and their implementation ideally needs to take account of the interplay between English proficiency and ability and desire to exercise autonomy as a translated/ing being. Whether it is practically possible to do so is another question. Angermeyer (2015, 8) makes a similar point drawing on observations at a small claims court in New York City. The court is seen to enforce a monolingual norm in relation to language choice, according to which litigants are required to either speak English or speak no English at all, and therefore speak via an interpreter. Although such a norm might appear rooted in the logic of bureaucratic efficiency, the fact that there are many who seek to resist this norm suggests that in some circumstances it creates a form of structural violence that suppresses rather than promotes the individual's voice.

ESOL as an interpellative act?
The politicisation of ESOL discussed earlier points to its possible construal as an interpellative act. In other words, Statesponsored L2 acquisition is often couched in terms of a deliberate intervention designed to support certain social actions and therefore a certain social order (e.g. finding work, maintaining employment, engaging in other educational opportunities, handling everyday interactions), as the Manchester ESOL Strategy illustrates. Further, as Simpson (2015) observes, the connection made between multilingualism and social disorder is usually couched by politicians in liberal terms, i.e., for the avoidance of economic and social marginalisation. A key question, in light of Cronin's argument, is the extent to which ESOL as an interpellative act has the potential to encourage investment in the subject-position of intercultural contact and, if so, how and in what ways? Does it create more of a readiness to turn towards the agent that hails than PSIT provisions? What sort of subject-position is entailed through ESOL provisions?
Couching the question in these terms is risks reproducing the binary evident in the ESOL-PSIT relation presented as a zero sum game. Nevertheless, it does open up important questions about the assumed direct relation between 'hailing' and take up of provisions. In constructing the ESOL subject, the individual who apparently fails to 'turn' and take up provision is cast as lazy and, in extreme cases, as a threat to social order. At the same time, the inconsistency of provisions and frequent limited choice for individuals regarding which classes to attend inevitably causes a retreat into (a) familiar linguistic community/ies and, rather than build resilience, may encourage resistance and limit motivation for intercultural contact later down the line. In such cases, structural deficiencies lead to a breakdown of trust: individuals are hailed in their difference only to be failed in turning. Alexander et al. (2007) draw a similar conclusion in their research on English language, citizenship, nationhood and belonging. A central aspect of their argument is that English is commonly viewed in symbolic terms as a 'cultural boundary marker' that simultaneously defines minority ethnic communities and excludes them from the 're-imagined' national community. The authors point to the dual emphasis on the utilitarian and symbolic approach to language promoted through the (then) new citizenship test: English language would not only support practical matters such as employability, but also be the vehicle through which individuals would be able to pursue modernity. For these authors, the 'culturing of language' that inheres in such a dualism is problematic when mapped onto a particular vision of national identity. Furthermore, they claim that 'top down' approaches to community do not account for structural and personal barriers that can impact on an individual's ability to access the full range of rights that come with citizenship. As a counterpoint, they propose the notion of 'personal communities', which captures the notion of community 'from below' and its 'private performances' (p. 786).
The focus on community was an unanticipated theme in Alexander et al.'s (2007) interviews, which were initially designed to probe experience of accessing services through interpreters (50 individuals in two major urban conurbations in England, across five languages). The main form of interpreting support reported by respondents came from their 'personal communities' (i.e. non-professional interpreting). Across the sample, individuals reported that professional language provisions were not always able to meet their linguistic needs. These findings may be viewed as a failure of the kind of interpellation described above and serve as a point of comparison for data analysis in this study.

Lived experience of language support provisions: examples from the data
The survey results of the Entry 3/Level 1 learners confirmed the validity of adopting a continuum-based approach to PSIT and ESOL provisions. The fact that three quarters of those who reported using professional interpreters in the 12 months prior to the survey indicated that this occurred five times or less in each setting provides good evidence to support the punctual rather than long-term dependence on provisions. However, the survey data are limited in terms of the insight they yield into the social (and interactional) experiences of living in multilingual communities and working with interpreters, and the extent to which this shapes their individual motivation to progress with L2 acquisition and request an interpreter to access services. In the section that follows, I firstly provide a snapshot (3.1) of interpreter usage revealed through the survey of Entry 3/Level 1 learners across different learning centres of the Manchester Adult Education Service, and in 3.2 I draw on examples from the interview and focus group data to probe certain themes in more depth.

Survey responses: a snapshot of interpreter usage
The survey questionnaire (see Vasey et al., 2018) primarily elicited information on accessing ESOL courses and perceptions of L2 acquisition. It also included a short section on interpreter use in the 12 months prior to the survey being taken. The level of study (Entry 3/Level 1) provided a suitable point at which to gather such data as it was hypothesised that the students would be approaching a tipping point in their L2 development at which interpreter mediationwhether professional or nonprofessional -was less needed. A small majority of respondents (56%, n ¼ 333) reported using an interpreter in the past year, which confirmed the relevance of the questions in this section to the respondents in the sample.
The extent of the hybrid use of professional and informal interpreters is striking in the sample with more than four-fifths of those reporting that they had used an interpreter in the past year also reporting that family and friends had acted as interpreters for them. In all settings (GP, Hospital, Lawyer, Police, Job Centre, Council, School) at least half of respondents used both professional and informal interpreters and this rose to nearly three-quarters of respondents in the hospital setting. However, most of those who had used professional interpreters reported using them five times or less in each setting over the past year, which is indicative of highly punctual use of such services.
Nearly three-quarters (72%, n ¼ 234) of respondents reported their reliance on interpreters had decreased, although for this question the drop in response rate is noteworthy. Whether this is due to questionnaire fatigue -it was question 30 of 31or lack of perceived relevance is open to question. However, reasons cited for the decrease this include greater confidence in using English (12), increased knowledge of English (10), and practice (6). Amongst those who felt their need for an interpreter had not decreased, increased health needs were cited as the most common reason (5), with continued assistance from family members following close behind (4). A small number (2) still felt their English was not sufficient to express their needs.
More than 1 in 7 of respondents felt that they had been refused an interpreter when they needed one. However, the additional comments reveal that some (7) respondents had understood this question to be asking whether they had ever turned down an interpreter; it is therefore likely that this figure is slightly inflated. Respondents reported being denied access to an interpreter in hospitals (12), and GP surgeries (7), settings where complex and technical language is most likely to be used. Some of the service denials emerge as a result of ad hoc evaluations of English proficiency carried out by the service provider. Whether this is done simply to expedite the appointment and/or is based on the healthcare practitioner's preference for direct over interpreter-mediated communication is open to question.

Interview and focus group findings
The interview and focus group data included 36 respondents who were attending informal ESOL provisions in the city and 14 who were retired or working full time. The data also include 8 interviews with individuals who completed the survey and agreed to do a one-to-one interview. The data present a very mixed picture in terms of take-up of professional interpreting services and a surprisingly low take-up of interpreting provisions generally speaking -surprising given the limited language proficiency evidenced among some of the language groups interviewed. Assumptions that respondents would be likely to report 'clustered' use of professional interpreters to facilitate key interactions in the early phases after arrival were also not evidenced. However, demand for professional services in the city of Manchester remains high and is growing: healthcare statistics on interpreter requests in recent years shows this to be the case (see Gaiser and Matras, 2016). It is clear that the overall city picture is therefore more complex than the findings of this research suggest. In the sections that follow, examples from the qualitative data shed some light on individual experience and how this impacts on take-up of PSIT provisions. Given the unanticipated finding of low take-up, the section focuses on the reasons given for resisting provisions and the impact on subsequent institutional interactions.

Mismatched provisions and the problem of 'epistemic certainty'
Angermeyer's (2015) study of interpreter mediation in a small claims court in New York City mentioned earlier, draws attention to issues of scheduling and hiring of interpreters that are also reflected in the data sets in this study. He observes that 'the courts presuppose that litigants speak language varieties that are clearly distinct and identifiable ' (p. 22), and that interpreter scheduling is typically not sensitive to regional or non-standard varieties. In this study, three female focus group participants (in different groups), for example, recall requesting a Farsi-speaking interpreter to access GP services and a Darispeaker being sent: Farsi-speaking Participant 1 (back translation into English): 'I can understand what the interpreter was saying, by and large, but when they go on to communicate in English to the GP or professional I know that they are making mistakes.'

Farsi-speaking Participant 2 (back translation into English):
'It got to the point where I'd had that many different interpreters [for GP appointments], and things being lost in translation and the doctor was like 'why do you think it is that?' 'I don't understand what is going on with the condition'. I could tell that the interpreters were not interpreting my words accurately'.
These extracts -communicated via an interpreter who shared the language and cultural background of the participantshighlight the challenges of accessing a service through an interpreter who does not share the same language variety. They also point to evaluations of quality and accuracy that seem striking given the ostensibly poor command of English demonstrated by these particular focus group participants. It is noteworthy that in the first extract the interpreter's communication from English -although less than ideal -was tolerated, whereas into English the lack of control over the message led to a more critical assessment by the service user. There is a possibility that the lack of shared language variety generated an immediate bias in the service user, which, despite the individual not being in a position to verify the interpretation, led to the negative evaluation. The examples suggest that if provisions are only viewed from the structural perspective of supply, the contradictory and, in some cases, harmful practices that occur on the ground can remain obscured, practices that can exacerbate rather than mitigate vulnerability.
The examples also highlight the problem of what might be termed 'epistemic certainty' in interpreter-mediated encounters. This concerns claims of poor interpreting by the service user, as in: 'I know they are making mistakes'. Although L2 speakers are often able to understand more than they can confidently produce in early phases of language acquisition, the notion of epistemic certainty in relation to the perceived accuracy of interpretation merits some attention, even if only in speculative terms. In such circumstances, it is possible that intuition and crude measurements play a role, such as whether the length of turn is matched in the source and interpreted versions: social workers in Tipton's study (2012) reported using such techniques to make a general assessment of interpreter performance.
Negative evaluations of interpreter mediation and instances of potentially misplaced epistemic certainty are observed across groups in the data set (French, Kurdish, Farsi, Urdu, Shona, Swahili speaking participants), and in relation to different settings (e.g. asylum, court, and healthcare). What is of interest to the discussion here, however, is not the verification of these claims, since this is beyond the scope of the project, rather the extent and nature of such perceptions, and how they impact on subsequent decisions in relation to language support provisions. In other words, does the experience of an apparently poor (professionally) interpreter-mediated service encounter lead to a rejection of professional services in other encounters and/or greater reliance on family/friends?
The question helps to critically evaluate reported declines in the numbers of people accessing services via a professional interpreter; any decline cannot easily be attributed to improvements in language proficiency alone, although the survey data show this occurs in some cases (see the analysis in Vasey et al., 2018: 39). To investigate this question, further analysis of the drivers behind interpreter requests is needed because the relationship between decisions made 'upstream' of an interpretermediated encounter by the service user and perceptions generated during encounters appear to be highly variable and call into question some of the general assumptions that underpin the continuum-based approach.

Factors that influence interpreter requests
Many intersecting issues appear to bear on the decision to request a (professional) interpreter as identified in the data, among which the following are present in more than one language and cultural group: language ideologies, prior direct and indirect experience, attitudinal legacies, credibility in the asylum system, self assessment of proficiency levels, awareness of the availability of interpreter provisions, health and wellbeing, interpreter power and control over interaction, motivation to use limited English as proof to the self of ability and/or to avoid burdening services unnecessarily. What emerges in the examples highlighted for discussion below is the level of structural violence occasioned by practices on the ground that are typically obscured at the macro-level of the organisation of provisions. Several individuals, for example, articulated the extent to which language ideologies had influenced decisions to request an interpreter at an asylum tribunal. As one female from Cameroon observes: In my country we speak English. It is like a 'forced language', you know...so...coming to UK I was advised, you know, there is word and times of British, the way they say their words. I was not confident enough, so I said, ok, let me go for an interpreter to interpret for me. Getting them, you know, what's they supposed to do to interpret for me.is not saying it, so I have to be the one speaking English, and the judge now said they think eyes on me that maybe I was lying, that I did not speak English and I have to tell them 'I have English but, you know, British English.the words and the times that they use, I may not use it', that's why I said I needed an interpreter'.
In this extract, the epistemic certainty demonstrated in relation to interpreter performance prompts an action by the claimant to break away from Angermeyer's (2015, 8) concept mentioned earlier, namely the monolingual norm of language choice, i.e., the norm of speaking English or speaking no English at all in the institutional setting. The assertion of language rights in this case shows the problem of accommodating the subaltern voice in the asylum tribunal court: the communication-related vulnerability felt by the claimant as a result of colonial practices is in effect reinforced through the judge's response to the respondent's decision to disrupt the monolingual norm of language choice and communicate directly in English. The respondent reported that this experience led to the decision to communicate directly in English in subsequent institutional interactions. However, the extract above, which has been reproduced verbatim, provides an indication of the potential problems that decision might have caused for the individual in this and other institutional settings.
In other cases reported in relation to the asylum system by focus group and interview participants, direct and indirect experience of language ideologies and practices impacted on the decision to request an interpreter. In another example, a female from Cameroon explicitly stated to her solicitor 'I don't need a translator', based on observations of interpretermediated interactions with friends in which interpreters claimed to speak a wide range of African languages but apparently did not have a sufficiently high level of proficiency to perform as an interpreter in all of them. These comments can be seen as another example of structural violence present in some PSIT practices: limited language combinations are available in certification schemes and quality monitoring, while it does happen in some settings, is not carried out systematically due to resource constraints. 9 Variable workflow and income generation in one language combination is also likely to lead individuals to offer more language combinations than they are qualified to do.
Another example from the data highlights a different form of structural violence: a Kurdish-speaking focus group participant mentioned frustration at having his participation in a Home Office interview controlled at certain points by the interpreter, who was reported as advising on the amount of information conveyed and the type. At one point in the interview the interpreter is reported as saying 'I know the rest', thereby treating the claimant as a generic other. Although this had not led the individual to reject professional interpreting services in subsequent interactions, it did lead him to question what professional interpreting means and highlighted his vulnerability as a non-English speaker in a high-stakes encounter.
In other participant testimony the decision not to request an interpreter was motivated by reasons of personal pride. In one case, a male respondent from Syria who had worked as an English teacher before coming to the UK reported deliberately limiting requests for a professional interpreter to several initial institutional encounters due to difficulties in understanding the local Manchester accent. Personal pride also motivated the decision in the cases of a female from mainland Spain and female from Poland who both felt an obligation not to burden services -in both cases the individuals' respective partners provided informal interpreting support if needed.

Conflation of issues: gender, monolingual interactions, performance
In some of the data, negative impressions of professional interpreter mediation were reported in ways that juxtaposed several issues, making it difficult to discern the actual source of the problem. In all three examples below, reported in English, the negative perception of professional interpreting impacted on subsequent decisions to attempt direct communication with limited English proficiency.
Example 1 Female Arabic-speaker: 'I need to go to GP, so when I speak with that woman -I understand a little bit in Englishok? - It is unclear whether the problem lay with the fact of having a male interpreter or whether the same outcome would have arisen had a female interpreter been supplied and asked similar questions.
Example 2 Male Swahili speaker: 'I was in the court with someone and there was this old lady who kind of translated in our language, but she was old, but when she was translatingas before we were sayingit was a totally different statement, totally different'.
Here again, it is unclear whether the fact of having a female/male pairing (the friend in this case was male) or indeed the age of the female interpreter was perceived as a problem and/or whether the main issue concerned the lack of shared language variety.

Example 3
Male French/Lingala speaker reflects on an asylum interview at the Home Office: 'I hope this doesn't sound bad but [the interpreter] was someone of white skin colour from Belgium or France if I remember. You could see that they could not relate at all to my case. They couldn't understand anything at all.'cos I might be wrong but I believe the interpreter has to relate to you to some extent, yes.to them it was just a case of ticking boxes, just doing a job'. This is an example of an intuitive response to the gender/race of the professional interpreter in this case and a possible misreading of the impartial stance taken by the interpreter: a trained interpreter may come across as lacking in empathy if s/ he actively promotes impartiality through body language, tone of voice and even in the choice of expression. The reflections were not accompanied with examples of perceived inaccuracies in the interpretation itself: reference to 'they couldn't understand anything at all' seems therefore to be a comment on the interpreter's lack of shared experience rather than comprehension difficulties. The reported response to the Home Office rejection letter clearly shows the individual laid the blame at the interpreter's feet. Later in the interview the respondent talked of professional interpreters using language that was perceived as too formal and as not providing sufficient explanation to facilitate understanding in encounters, but the comment was not made in relation to the asylum setting specifically.

Established communities
Approximately one fifth of those interviewed belong to immigrant groups that have a sizeable and long standing presence in the city of Manchester: Chinese (Cantonese), Urdu, and Polish speakers (see MLM, 2013). Among these groups, the use of interpreters was reported as some of the lowest; yet, proficiency in English was also reported as some of the lowest, raising additional questions about the assumptions that underpin the ESOL-PSIT relation.
What emerges among all three groups is the ability to rely on in-group support for almost all aspects of daily life. Several Cantonese-speaking participants described going into the restaurant business relatively soon after first arriving, but having the support of Cantonese-speaking solicitors to set up the business, handle property purchases, etc. Gradually, through limited interaction with customers ('The only ways of learning English in England was to read the menu') and then through their children's ability to speak English, their English language proficiency improved. The discussion highlighted abortive attempts to engage socially with the local English-speaking population: incompatible drinking and eating habits precluded some interactions, and racism and discrimination reported across the group often led to a retreat into monolingual interactions within the community group. Recourse to interpreter mediation was reported as low among those who had been in the UK for twenty years or more, largely due to the lack of available provisions at the points they were most needed: 'Well, we could only use the best of our English and luckily the children's medical attention was not that greatit was either a flu or they weren't eating properly, so we could get by explaining that to the doctor'.
Although racism and discrimination were not reported among the Polish and Urdu-speaking respondents, similar experiences of monolingual intra-community interactions were cited as the reasons for not requesting interpreter mediation and, in some cases, for not taking up ESOL provisions. Word-of-mouth information on where to open a bank account at a branch where a Polish speaker worked, for example, and where to find work were reported as widespread among respondents' social networks. Work and shifts were cited as the common barrier to taking up ESOL classes. All of the four Polishspeaking respondents had experience of almost exclusively Polish-speaking workplaces and all lived with Polish speakers.
In the case of Urdu, the respondents included young people 17-19 age group and adults in their 40s-50s. Among the older adults, negative experiences of professional interpreter mediation in Home Office encounters were also, leading to reliance on family in other institutional interactions. The young people reported developing proficiency through school reasonably quickly thanks to additional support in English. However, one respondent who came forward to be interviewed following the Entry 3/Level 1 survey highlights the impact of belonging to a sizeable minority language group in the city and the impact this can have on English language proficiency in the family over the longer term: 'Actually the problem in Manchester is the Urdu speaking community is more than the English speaking and at Manchester College [X campus]. There are a lot of Urdu speaking people and my daughter is speaking Urdu with Pakistani people. I told her 'you will not improve your English if you are talking with the Pakistani people in Urdu, please find English people and make communication and your language will improve. Otherwise you will pass your time and you will never improve your English'.

Discussion
The survey data and interviews show that engagement with formal ESOL provisions can be sporadic and non-linear due to family and work commitments, with some learners returning to ESOL classes many years after they attended their last one. Respondents across the groups often claim that voluntary and community organisation-led English conversation classes are at too low a level to be useful and some highlight mixed experiences of Jobcentre ESOL provision, such as a lack of differentiation within the group of learners. In this regard, some reported having to repeat courses to comply with job-seeker allowance requirements in cases where progression to higher level courses was not possible in the local area. This is an important finding in terms of critically appraising ESOL as an interpellative act; although it stands to reason that targeted and timely ESOL provision is desired as soon as possible after arrival (taking account of eligibility criteria), structural deficiencies can preclude progression and impact on motivation to engage over time.
The validity of a continuum-based approach to ESOL and PSIT suggested inter alia by Ager and Strang (2008) is further supported here, but the data show that movement between the two is less frequent than perhaps anticipated and does not follow a discernible pattern. These authors highlight efforts to improve the intercultural communicative competence of key organisations, pointing to the availability of translated literature and interpreting services ('heteronomous translation' in Cronin's terms). They therefore point to the 'interpellative' role of key public organisations and in so doing draw attention to interlingual and intercultural aspects of 'hailing'in such cases, it is no longer the newcomer that is seen to invest in the subject-position of intercultural contact, but also the indigenous professional, thereby lending support to views of integration as a two-way street. Cronin's (2006) macro-level focus on how to engage newly arrived immigrants through language provisions tends to assume that such provisions occur in optimum conditions (e.g. language matching, gender-appropriate services, adequately trained interpreters, etc.). The examples discussed here help to show that behind the notion of service supply a wide range of practices occur, suggesting that greater emphasis needs to be placed on aligning macro and micro perspectives when examining the local context of practice. In some cases, the inability of services to meet intended goals can be a motivating factor for individuals to develop self-resilience and progress in L2 learning through their own means. In others, vulnerability may be increased as they turn away from formal support structures.
The testimonies from the interviews and focus groups point to the thorny issue of agency in the interpellative acts of language support provisions. The question emerges about not what policy should be in place but how to build structures that allow different forms of agency to emerge. The analysis here shows the duality of structure (Giddens, 1984) to be enabling and constraining: structures set up ostensibly to support social participation, can sometimes cause violence to the individual. There is at present no acknowledgement or readiness to acknowledge this possibility among institutions tasked with providing access to services through interpreters and translators. And even if there is some form of recognition, there are currently no mechanisms for thinking through the consequences or offering solutions.
While both ESOL and PSIT may be conceived in interpellative terms, the voices of those who engage with provisions tell important stories about resilience building from below that occur often as unintended consequences. It is striking that none of the personal narratives from the focus groups and interviews mention 'citizenship' or 'integration'. Any retreat away from ESOL and PSIT provisionsas mentioned aboveoften appears in this sample of respondents as an unintended consequence and evidenced through: 1) the lack of availability of ESOL provisions at appropriate level and venue; 2) the possibility for some groups to avoid institutional contact and manage key encounters (opening a bank account, finding accommodation, employment) due to the availability of language-specific professional services; 3) frequently reported problems or perceived problems when accessing services through professional interpreters.

Conclusion
This article explored the discursive construction of the ESOL and PSIT subject and the political formations in which the relationship between the two are currently embedded, the aim being to interrogate the oversimplified approach to the relationship in which they are commonly presented in terms of a zero sum game. Prevailing political drivers of policy at the local level reveal particular orientations to risk and care, for example, through the concept of resilient communities, and are marked by the neoliberal shift to individuals taking responsibility for themselves. This shift has not been systematically incorporated in conceptualisations of immigrant assimilation and integration, but emerges as a significant point of tension in this study in terms of how the frameworks for resilience building are created for both newly arrived and indigenous/settled populations.
The data suggest that state-level provisions of ESOL and PSIT are entangled in various practices (professional/volunteer) at the level of implementation that raises questions for the nature of agency enabled for immigrants engaging with both. Although a continuum-based approach to immigrant engagement with ESOL and PSIT is helpful, it assumes that the offer of both is consistent and consistently available, assumptions that this study has empirically challenged.
The decision to resist certain types of provision at critical moments is ostensibly influenced by a complex set of intersecting issues. The study shows that the perception of provisions-as-practice can be particularly impactful on the lived experience of individuals. It also shows how those who 'turn' because of the promise of identify that is rooted in a certain conceptualisation of community and subject formation cannot be blamed for resisting provisions if that conceptualisation is incompatible with the dominant discourse-as-policy. The idea that individuals' everyday language practices are often distinct from state planning in relation to language policy is not new (e.g. Pennycook, 2013), but the tensions cannot be ignored especially in situations of extreme vulnerability. The study reported on here suggests that self-reliance and resilience emerge in some cases due to a failure of interpellative acts which can push individuals to adopt folk remedies through a retreat into community groups and greater risk taking.
Engaging key stakeholders in developing language-sensitive approaches to community resilience requires cognizance of the lived experience of the practices that impact on individuals' life chances, and willingness to subject professional/nonprofessional language support provisions to the level of critical scrutiny that is arguably absent at this time of writing. This type of dialogue is essential to the development of coherent city-region language strategies of the future.

Funding
This work was supported by the Arts and Humanities Research Council of Great Britain under the Open World Research Initiative, Grant Number: AH/N004647/1.

Declaration of interest
None.