CC BY
21
DOI 10.30842/alp2306573714314
Y. Matras
University of Manchester, Manchester, Great Britain
THE MULTILINGUAL MANCHESTER RESEARCH MODEL: AN INTEGRATED APPROACH TO URBAN LANGUAGE DIVERSITY
1. Introduction
In this contribution I outline a new approach to research on urban language diversity, with the potential to innovate methods of enquiry in a number of ways. First, the model involves external (non-academic) stakeholders in an iterative process of enquiry, seeking inspiration from the practical challenges that they face in areas such as heritage language maintenance and service provision to a multilingual population.
Next, it engages students, benefiting from the questions that they raise as well as from their experience of direct immersion with communities of practice in the local voluntary and public sectors. This latter experience is facilitated through a student volunteering and placement scheme that creates a setting for research where enquiry is reciprocal and researchers 'give back' knowledge and skills to local actors.
Finally, it adopts a holistic approach to language practices, one that incorporates a wide range of methodological tools and types of data and data analysis. These include conversation and structure oriented analysis of interaction in multilingual contexts, ethnographic observations and analysis of narratives and meta-discourses about language and identity, as well as investigation of language policy at macro-, meso-, and micro-levels and development of new digital tools to capture data.
This approach, which we have previously described as a 'non-linear' model of participatory research [Matras, Robertson 2017] on account of its dynamic reciprocity, its iterativity and responsiveness, is designed to capture a holistic picture of multilingual practices in the complex setting of the globalised city. It also aims to find an ethical way of accommodating to a new academic environment in which students
are regarded as clients and external stakeholders are viewed as target beneficiaries of research. At the higher level of university management, this environment is partly driven by metrics that are intended to measure performance in terms of student satisfaction and graduate employability, and the impact of research on the daily practice of non-academic audiences. Resource allocation for both teaching and research is often linked to such performance indicators. Our model looks for a way to embed research into this reality, while at the same time utilising it to re-define the relationship between academia and its environment as one that is mutually productive. It thus offers gains that are both of an academic and a social nature: On the academic side, it opens up new avenues of enquiry by testing ideas in new collaborative settings. On the social side, it seeks to re-negotiate the balance between expertise that is commissioned in order underpin policy, activism that mobilises knowledge for a pre-defined cause, and the kind of Intellectualism that manifests itself as informed commentary, independent of, but not indifferent to the possibility that critique may succeed in setting in motion change in practice.
2. The context: New directions in the study of urban multilingiialism
Research on urban language diversity owes much to the pioneering work of [Garcia, Fishman 1997], who present an inventory of minority language communities in a metropolis. The contributions to their volume on New York's 'multilingual apple' rely on a combination of observation and introspection. They address 'communities' as discrete entities with more or less unchallenged boundaries — well represented by the division into chapters each devoted to a different language. The series of large-scale surveys by [Extra, Yagmur 2004] and collaborators present inventories of a different kind (see also [Barni, Extra 2008]. Here, the aim was to provide an overview of language maintenance trends and practices through the lens of a uniform quantitative approach, designed to capture a scalar picture of attitudes to home languages among school pupils in various large European cities. Introspection is replaced by self-reporting, and the qualitative overview by a computational grid that measures language vitality. The compositional nature of the multilingual city as an assembly of communities thus gives way to the multilingual city as an integrated whole, with the school institution as its proxy and its language communities as discrete (digital) components.
Inventories of urban linguistic landscape, captured in the form of quantitative surveys that correlate language to location and outlet [Landry, Bourhis 1997], offer another measure of language vitality.
Heller [2011] introduces a new view of the city as the site of language policy that is localised as well as pluralised and is therefore in a sense 'post-national'. This coincides with a new focus on practices of local actors and networks of actors in individual institutions and sites [Cadier, Mar-Molinero 2012], reflecting a new understanding of 'policy' as enacted not just in the form of scripted documents, drafted and implemented by state agencies, but characterised by smaller units of practitioners and ad hoc practice-bound partnership constellations that operate at the micro-level [Liddicoat, Baldauf 2008; Bonacina-Pugh 2012; Davies, Ziegler 2015].
These developments have accompanied a conceptual shift from an understanding of pre-set categories and groups as determining practice, to a view of practice as embedded in space and responsive to permeating constellations where categories are fluid [Blommaert, Collins, Slembrouck 2005]. This shift has been labelled 'critical sociolinguis-tics' — a new approach to the study of language practices that questions stable correlations between linguistic variables and extra-linguistic descriptors and is instead set to capture the constant dynamics of change that are characteristic of the complexity of the multilingual city and its superdiversity [Blommaert, Rampton 2011; Blommaert 2013; Ar-naut et al. 2015]. Studies inspired by this conceptual framework investigate the multilingual metropolis through the prism of practices in which individuals engage as they move across places as part of their daily routines [Lamarre 2013; Pennycook, Otsuji 2015]; or through the lens of life histories [Stevenson 2017], intertwined with actors' self-reflection on practices and the surrounding public narratives on language and identity.
Pursuing the critical channel further, researchers have been turning the notion of 'translanguaging' from a designation of practice (where users resort simultaneously to a full range of elements from their wholesale repertoire of linguistic forms, cf. [Blackledge, Creese 2010], to a designation of method, where they question language boundaries as an analytical concept and look beyond linguistic structures at multi-modal aspects of communication in cross-cultural settings [Li Wei 2018]. The 'critical' aspect takes a literal sense when authors overtly question the usefulness of wide scale surveys that capture quantitative data on languages,
dismissing them as 'demo-linguistics' [King 2015; Pennycook, Otsuji 2015] that fail to capture actual practice.
My reflections are anchored in a trajectory of a sequel of works which together seek to develop a framework that is integrative and holistic: It is integrative, in that instead of negating earlier concepts, such as the search for quantitative indicators of language vitality, it tries to reconcile them wherever possible with newly emerging methods and theory. It is holistic, in that it aims at developing tools to describe various aspects of urban multilingualism: from individuals' functional management of a complex repertoire of linguistic structures as a pathway to an explanatory model of contact induced structural change [Matras 2009], through an interpretation of the city's language landscape as a map of repertoire management and spatial construction [Gaiser, Matras 2016b; Matras, Gaiser, Reershemius 2018], to an assessment of language policy and provisions in the city [Matras, Robertson 2015; Matras 2017], and the development and piloting of new research tools [Matras, Robertson, Jones 2016] and participatory research models [Matras, Robertson 2017]. Bringing together this range of enquiry strands, we are in a position to formulate a new epistemology of urban multilingualism: We ask how active engagement with practice communities informs and shapes our enquiry, and how, through the model of reciprocity of knowledge exchange, facilitated by the 'civic university', research makes an active contribution to shaping public narratives on language and in that way, potentially, also to shaping practice.
3. The setting: Language diversity in Manchester
Manchester is known as the world's first industrial city, which owes its growth to labour migrants first from neighbouring regions and subsequently from overseas. It is also the historical setting of radical movements in support of the abolition of slavery, reform of parliamentary representation and universal suffrage, nuclear disarmament, asylum for refugees, and other causes of equality and social justice. As industry declined in the late 1970s, the city began to re-brand itself. A concentrated regeneration effort was launched in the late 1980s and early 1990s, characterised by a model of partnership between the public and private sectors and community initiatives aiming to capitalise on the city's diversity to project a cosmopolitan image that blends prosperity with empowerment and social justice [Peck, Ward 2002; Williams
2003; Young et al. 2006]. As austerity measures were introduced in the aftermath of the 2008-2009 financial crisis, creating a hole in the budget of local authorities, the city embarked on a strategy that would see voluntary sector initiatives take on more responsibility for advice and support, helping in particular what is regarded as 'emerging communities' acquire so-called 'resilience capital', which in effect means avoiding or overcoming dependency on municipal agencies for basics such as access to housing and routes into employment.
Manchester's population is currently estimated at around 530,000, but the city serves as a commercial and cultural centre for a metropolitan area with a population five times larger. In the 2011 Census, 16.6% of households, representing around 85,000 residents, declared a 'main language other than English'. However, it is widely believed that this figure underestimates the number of multilingual households [Matras, Robertson 2015]. For one, there was lack of clarity as to whether the term 'main language' represents personal preference, proficiency, or frequency of use (see below). In addition, respondents who would attribute the same or similar importance to English as they do to another language did not have the opportunity to indicate this on the census. That, together with the realisation that the number of multilingual households has increased since 2011, puts the likely proportion of households that use languages other than or in addition to English at anywhere between 3040 %, or roughly 150,000-200,000 of the city's residents. Recent annual School Census data show that around 40 % of school pupils are identified as having a 'first language' other than English. Again, it is believed that this figure under-reports multilingualism, as it does not necessarily take into account children who speak English at home with one parent and another language with another parent. The realistic figure of pupils with a multilingual background is thought to be upwards of 50%.
Statistical sources on the number of languages spoken in the city also vary. Published data from the 2011 Census named around 70 individual 'main languages' that were reported by respondents, and grouped additional languages by region of origin. The complete list of responses provided to us by the Office for National Statistics contains a total of 286 languages for the city of Manchester (367 for Greater Manchester), of which, however, only 169 (230 for Greater Manchester) were listed by 5 respondents or more. The annual School Census for Manchester tends to report upwards of 150 different languages as pupils' 'first languages'. Interpreter requests in the health care sector show regular demand for around
120 languages. Languages with large numbers of speakers include Urdu, Chinese, Arabic, Polish, Panjabi, Bengali, Somali, Persian, and Kurdish. French and Portuguese are widespread among communities of both African and European origins. Greater Manchester has the country's highest speaker concentrations outside of London for a number of languages including Yiddish, Somali, Kurdish, and Romani.
Of the city's 32 wards (administrative units with an average population of around 16,000), 5 have more than 60 % of children who speak a language other than English in the home, and in another 8 the figure is more than 40 %. In over 20 Manchester schools, pupils who have a first language other than English make up more than 70 % of the school population. Around 45 Manchester schools identify more than 30 different first languages that are spoken among their pupils. The most frequent languages include Urdu, Arabic, Somali, Panjabi, Bengali, Polish, French, Yoruba, Portuguese, Chinese, Pashto and Kurdish.
Census figures from 2011 suggest that around 3.5% of the population, or roughly around 17,000, rated their proficiency in English as low or non-existent. In 2017, around 4,000 people were believed to be enrolled in various classes for English as Second or Other Language (ESOL), with an estimate of around 1,000 being on waiting lists. Public services in the city generally maintain provisions for interpreting and translation. The city council operates a translation and interpreting service with 11 contracted staff and around 200 freelance vendors, who respond annually to over 12,000 requests for interpreting and translation in more than 70 different languages. One of three major hospital trusts in the city, Manchester University NHS Foundation Trust (MFT), maintains an in-house translation and interpreting department with around 10 full-time and additional 10 part-time staff who are supported by external contractors; together they respond to around 50,000 annual face-to-face and telephone requests for interpreting in around 100 different languages. Other hospitals, and the city's emergency services, rely on a number of local contractors for interpreting and translation who often draw on the language skills of local residents. Manchester's General Practitioner (medical) surgeries register upwards of 15,000 interpreting requests annually. Languages with a high demand for interpreting services are generally those that are most widespread in the city, including Urdu, Panjabi, Arabic, Polish, Bengali, Persian, Kurdish, Cantonese, and Somali, as well as Romanian.
It is estimated that around 3,000-4,000 Manchester pupils take secondary school level exams in foreign languages every year, though
precise figures are difficult to come by since schools contract a number of different authorised providers of examinations. The majority of these qualifications are in European languages, mainly French, Spanish, and German; but hundreds of pupils each year complete exams in 'community' or 'heritage' languages (the label given to languages that are spoken by local immigrant communities, and more importantly, the way of designating language courses that are taken up largely by pupils with exposure to the language in the home context). These include, in Manchester, mainly Urdu, Arabic, Panjabi and Polish (cf. [Matras, Robertson 2015]). Upwards of fifty supplementary schools operate in the city, offering weekend classes in the city's principal community languages such as Arabic, Urdu, Bengali, Panjabi, and Polish, but catering also to some of the city's smaller or incipient communities such as Tamil, Turkish, Somali and Uyghur. Manchester's Haredi or so-called 'ultra' Orthodox Jewish community operates community schools that use Yiddish as a medium of instruction and Hebrew as the language of study of scripture. A good measure of community language vitality are the city's linguistic landscapes, which feature some 50 different languages on public signs of commercial outlets, cultural and religious institutions, and occasionally on public sector notices [Gaiser, Matras 2016b].
4. The MLM model
Multilingual Manchester1 was set up in the academic year 20092010 as a new model for high impact and participatory research. It was conceived of in pursuit of two main objectives: First, to secure a framework in order to make the archived outcomes of previous research on languages sustainable beyond the lifetime of the external funding that supported that research, by pooling together the work of various colleagues under an overarching umbrella that would be able to generate continuous resources for adequate infrastructural support. This was a reaction to the high volatility of IT support within the institution. It followed a period in the early 1990s during which considerable resources had been allocated by UK research councils to support the creation of digital resources in the humanities. But the absence of a local, institution-based framework to sustain that investment meant that there were no safeguards to protect the products of that research.
1 http://mlm.humanities.manchester.ac.uk.
Second, with student numbers in Language degrees beginning to decline and a trend gaining pace to shut down Language departments at various universities around the country, the future of Language studies seemed to depend increasingly on institutions' ability to offer innovative learning environments. The vision that guided MLM at that point in time was that Manchester could stand out among its immediate 'competitor' institutions by linking its teaching programme in languages to the city's linguistically diverse environment. This called for an approach that turned the city and its language communities, and the challenges that language diversity poses to public sector services as well to issues of community and civic identity, into an active, innovative and creative research setting. The pathway to achieve this goal necessarily involved establishing links with language communities and public services, in order to open up observation opportunities [Matras, Robertson 2017].
The concept failed to resonate at first with the management at the local level, where department and programme boundaries stood in the way of an integrated vision that would bring together a sociolinguistic agenda with the traditional focus of Modern Languages teaching. The idea of pooling together a technical and administrative infrastructure in order to sustain research outputs only appealed to a small circle of people, and was overshadowed by aspirations to create a university-wide archiving hub, plans that never materialised. The model did find support at the central level of university management. The university was at the time developing a new Social Responsibility agenda. This was a reaction to growing pressure to demonstrate a 'return' on government investment in research in the form of demonstrable impact on society, economy and policy, and growing competition among universities to recruit top students, and so to demonstrate a return on student tuition fees in the form of a unique 'student experience' as well as employability prospects. Two strands of the university's central administration supported the MLM idea: The first was a newly established fund for Social Responsibility in the Curriculum. It provided a modest award to support the creation for an online resource to archive original student research carried out as part of an undergraduate module on Societal Multilingualism, devoted to original observations carried out by students, under supervision, among the city's language communities. The second was the university's Business Engagement department. It was on the look out for new ventures to connect the university to a variety of local stakeholders, as part of the university's vision to revive its image as a 'civic university' that is tightly embedded into the
city's communal life and strategic development. With the help of Business Engagement, a 'branding strategy' was drafted for MLM. It identified potential partners for impact and the benefits that they would draw from the initiative, a communications strategy and measures of success, and the rationale for linking the model with the Manchester location.
The initial investment in an archive of student research developed into a sustained model of learning through research that led to the creation of an online archive of well over 100 research reports, constituting the largest online documentation of multilingual practices in any one city, and, with the cumulative number of contributors being around 500, the largest research resource authored exclusively by undergraduate students2. Themes covered in the reports include language practices in families and community institutions, language use in local businesses, language policy of public service providers, linguistic landscapes, and more. The archive attracted the attention of local stakeholders among public service providers and schools, who contacted MLM with requests for information. This led in a number of cases to research collaboration around questions of policy and service provision that were of direct interest to the external partners. In that way, it delivered on the agenda of the MLM brand, to be of direct benefit to external stakeholders in the local community. A key contribution that brings together external engagement with the ambition to strengthen the student experience aspect is the MLM student volunteer scheme. Launched in early 2013, it was the University of Manchester's first local (department-based) volunteering initiative, operating next to the centralised Manchester Leadership Programme, a credit-bearing programme that provided generic skills training combined with a work experience module.
The MLM volunteering scheme was not integrated into a credit-bearing academic programme, but provided students of all academic programmes with an opportunity to contribute, in a highly flexible and minimally scripted manner, to a variety of activities around the general theme of language diversity. Host institutions have included Manchester Royal Infirmary, where students accompany interpreters to record patient experience feedback from patients with other languages; Greater Manchester Police, where student volunteers carried out focus groups to collect feedback on letters to victims or crime and suggested ways of re-formulating them (for the benefit of all recipients, not specifically speakers
2 See http://mlm.humanities.manchester.ac.uk/reports.
of other languages); a group of speech and language therapists, whom students supported by creating a video outlining the benefits of family bilingualism; local schools, where students carried out events on multi-lingualism and language taster sessions; a series of voluntary sector organisations where students supported English conversation sessions for new arrivals, many of them refugees; public events at which they set up interactive exhibitions on language diversity; and more.
The scheme produced a range of deliverables: It gave participating students an immersion opportunity with the city's diverse population and the issues facing key service providers, and experience in working with people of a variety of backgrounds, including both 'clients' and professionals. It helped key institutions fill gaps in specific areas of service, and created in some cases lasting legacies in the form of new protocols (redrafted letters for the police force), new materials (toolkits for English conversation sessions, and interactive exhibition materials), and dissemination outputs (such as online videos for parents in bilingual families). It put engagement with the city's language diversity high on the agenda of the university's programme of outreach to the local community, and contributed to the university's reputation as a leading contributor to the public narrative on language and cultural diversity (see below). And it opened up observation settings and opportunities to carry out research on aspects such as public service engagement with language diversity, access to ESOL provisions, and narratives of language and identity. Building on the experience and success of the volunteer scheme, in 2017 a student placement scheme was set up, offering opportunities for students at different levels, including funded placements for several weeks for PhD students of all Humanities subjects. Host institutions for placements have included the city council, where students carried out focus groups and interviews and contributed to strategic reports, and community based supplementary schools, where students contributed to the delivery of logistics and marketing.
These activities opened up opportunities to collaborate with a series of public sector institutions: Open discussion forums, private conversations around arrangements for student placements, and student research helped identify research questions that were of interest to stakeholders. These became the focus of small teams of researchers who gained access to data in the form of statistical records, interviews and observations, leading to a series of co-produced research reports covering aspects such as use and accessibility of interpreting provisions in the health care service [Gaiser, Matras 2016a], organisational setup and practices
in community supplementary schools [Gaiser, Hughes 2015], learners' experience of English as additional language classes [Vasey et al. 2018], and more. The batch of these enquiry-into-practice reports further cemented the project's reputation among local stakeholders while demonstrating the relevance of questions of language diversity to the city's various institutions. Relationships were established which led to regular knowledge exchange activities, including providing training sessions of language diversity to members of the emergency services, training programmes for supplementary school teachers delivered by facilitators whom the project recruited, support for the drafting of various city council policy papers, activity days devoted to language diversity at local schools, and public events and interactive exhibitions. Many of the activities were accompanied by audio-visual documentation that was published online and on social media, creating a vibrant platform through which to forward inspiring imagery of active engagement with issues of language diversity across sectors3. Complementing the online documentation of activities is the project's work to develop new digital tools to capture language diversity: The LinguaSnapp mobile app and online database and mapping tool allow crowdsourcing of annotated images of multilingual signs. Launched in early 2016 at an event with the University's President and the City Council's Deputy Leader, the app has been employed to collect and archive data for research [Gaiser, Matras 2016b] and student projects, and versions of it have since been replicated in Jerusalem, Melbourne, St Petersburg and Hamburg. The Multilingual Manchester Data Tool, launched in late 2018, brings together a variety of statistical datasets on languages in the city, along with background information and a mapping tool, providing a prototype tool for a general coverage and data triangulation for multilingual cities. Since its release the model has attracted much attention and UNESCO has expressed interest in adopting it as a protocol for the documentation of languages in linguistically diverse cities.
5. Case study 1: Collecting data on language
As cities become more diverse and more complex, finding solutions for efficient delivery of services become more heavily reliant on close monitoring of data, and collaboration among networks of actors
3 See http://mlm.humanities.manchester.ac.uk/videos.
[Wong 2015; Amin, Thrift 2017]. Such notions of 'smart cities', as they are defined by researchers in geography [Albino et al. 2015; Mora et al. 2017], are usually applied exclusively to the physical environment, addressing issues of transport, environmental quality, and the accessibility of public services and amenities. Little attention so far has been devoted to the role that digital tools and networking can play in capturing cultural or language diversity. While language practice, as critical sociolinguists emphasise, is dynamic and not easily captured in discrete entities with strict boundaries, language can at the same time be a tangible indicator of an element of cultural practice: A request for an interpreter for a particular language, for example, represents an act that is intended to frame, through the use of a language label, the means by which a communicative interaction (usually in an institutional setting) can be facilitated. Often the language requested does not precisely capture the language of the actual interaction. For example, it has frequently been observed that clients ask for an interpreter in Urdu, but upon establishing that client and interpreter share knowledge of Punjabi, or another regional language such as Potwari, this becomes the language of interaction, while the request remains recorded in the log of interpreter jobs as Urdu. Likewise, a recorded request for Arabic does not capture the regional dialect in which the interaction actually took place, whether both parties spoke the same, a similar, or very different regional varieties, and whether differences may have impeded communication. Nonetheless, the record taken from a log of interpreter jobs at a particular institution such as a medical facility provides an indicator of the way in which actors (clients of the institutions on the one hand, and its agents on the other) attempted to frame a procedure to facilitate interaction in order to enable a transaction of some kind between agent and client.
Other kinds of records convey actors' various perceptions about their own or others' linguistic reality. The national UK Census of 2011 asked respondents to indicate their 'main language', allowing only a single choice. The attached guidance notes explained that 'main language' could be the language "you were brought up using, you feel most comfortable using, you use at home, or you use most often". Inevitably, this created ambiguity, with multilingual respondents tailoring their choice of just a single response to any of these different criteria. As a result the census data for 2011 may indicate for some respondents languages that they were brought up with, while for others they may indicate the language that they use more often. Consequently, data are not fully
comparable, nor are they necessarily complete in that many respondents will have under-reported languages that are part of their daily repertoire (even when disregarding languages that were acquired through formal instruction). Similarly, the annual School Census, which is carried out by individual local schools drawing a government template, reflects either parents' declarations about their children's 'first language' or else school staff's observations on pupils' 'first language'. This template, too, does not allow respondents to record more than one language, nor does it offer a consistent definition of 'first'. But in all these cases, records are provided of statements made by respondents, and these, in turn, document acts of framing interaction, repertoires or identities in a multilingual setting.
In response to the shortcomings of data collection tools on languages, MLM engaged in a discussion of instruments to compile data on languages, and developed a number of new tools. In Matras, Robertson, Jones [2016] we discussed problems with the School Census and introduced the School Language Survey. This was based on a method of direct face to face interviews with school age respondents, in which they were asked about the languages that they used, the persons they used them with, and language practices such as reading and writing, being read to, watching films, and attending weekend school or other extra-curricular structured language learning activity. We also piloted a rudimentary language proficiency test, design to obtain a quick and crude indication of respondents' level of fluency in each reported language. The test was based on the assumption that responses to short tasks would differ in their level of complexity and would therefore offer an indication of the level of fluency: Counting from 1-10 was expected to constitute a formulaic activity that would indicate superficial exposure to the language but not necessarily ability to communicate or even any degree of immersion in communicative routines. Naming body parts would indicate exposure to contexts where the language was being used for communicative purposes, but not necessarily an ability to hold conversations. Descriptions of daily routines would constitute the far end of the continuum and testify to at least some ability to communicate. Pupils were given a numerical score based on the spontaneity and apparent fluency with which they responded to the task. In this way the method allowed the team to collect observations on proficiency without the need to carry out audio recordings and without competence in the language. The results offered a clearer picture of pupils'
language repertoire, which enhanced the information that was available to schools from the official School Census. It also provided an indication of language vitality, by offering correlations between language proficiency and self-reported exposure to various language practices. In this way, the method went a step further not just in comparison to the School Census, but also in comparison with other research methods that relied on self-reported language preference and proficiency rather than on direct proficiency sampling, such as those described by Extra, Yagmur [2004, 2011].
The study of linguistic landscapes began to flourish in the late 2000s, as a way of investigating power relations among languages as well as the vitality of languages. Interest in correlating language, genre, message content and space gave rise to exploratory methods of mapping the location of languages on public signage [Bagna, Barni 2009, 2010]. In 2015, MLM designed and released a new digital tool to capture linguistic landscapes — LinguaSnapp4. The tool consists of three application components. The first is a mobile phone application that is downloadable for free and so accessible to all. It allows users to capture images of signs and upload them onto a database together with a set of analytical descriptors that can be entered using the application's pre-set menus. These descriptors capture information such as the number and names of languages and scripts, the position of the sign and the nature of the outlet on which it is found, and the hierarchical display of languages and accompanying multi-modality. Uploaded images with accompanying descriptors are stored in a database where they can be edited and released onto a map, which shows the image and its descriptors drawing on the GPS location sent by the mobile application. The online map is freely accessible and can be filtered based on any combination of descriptors. The release of LinguaSnapp introduced a new era into the documentation and archiving of linguistic landscape data, by allowing crowd sourcing and public sharing of a data corpus while also relying on key analytical descriptors [Gaiser, Matras 2016b]. Since its launch the application has served students writing essays about the city's linguistic landscapes, and schools engaging pupils with the use of online digital tools through the prism of community languages. With over 2,200 images it is currently the largest corpus of linguistic landscapes in an individual city that is publicly available. The application has
4 http://www.linguasnapp.manchester.ac.uk.
also been adopted and localised by several universities in other countries, the first such digital export of the University of Manchester's Humanities Faculty and a model for research, teaching and public engagement in multilingual cities.
Above I alluded to the ambiguity of the question 'What is your main language?' which first appeared in the UK Census in 2011. In Manchester, one person each put down 'Manx' and 'Cornish' as their 'main language', both languages that are considered extinct except in revivalist circles. By contrast, only very few people indicated minority or regional languages such as Romani or Yiddish, which are known to the MLM to be spoken by hundreds of people as the principal language of the home. Respondent feedback collected by the Office of National Statistics had confirmed user discomfort around the question, and as a result the Office for National Statistics (ONS) had classified the question as imposing a 'medium burden' on respondents. In its topic report on the language question from 2016, the ONS noted that "the relatively high demand for online help for this question indicated that some online respondents had difficulty interpreting the question", and also that "some non-UK born respondents were uncertain whether the question was asking about the language they first learnt or the language they most frequently spoke" (ONS Census Transformation Programme. The 2021 Census Assessment of initial user requirements on content for England and Wales, Language topic report, May 2016, p. 16). Following critique of the census question both in publications [Matras, Robertson 2015] and in media interviews, we set out to engage the ONS in a dialogue about possible amendments to the question, ahead of the final editing of the questionnaire for the upcoming UK Census 2021.
In correspondence and at a meeting with ONS officials in the spring and summer of 20185, we pointed out that other countries with an English-speaking majority have a more fine-tuned question on language in their census: New Zealand asks in which languages respondents can have a conversation about a lot of everyday things. Canada asks about proficiency in English or French, and then asks respondents to list any other language used in the home and any additional language used on a regular basis, and to state which language was learnt first.
5 For a documentation see http://mlm.humanities.manchester.ac.uk/census-2021-an-opportunity-to-acknowledge-multilingualism.
South Africa asks which two languages are spoken most often in the household. The US, Australia and Ireland ask respondents whether they speak a language other than English in the home. Some countries provide a list of pre-selected languages representing those that are most widely spoken, which saves many respondents the time and effort of writing down language names and facilitates the coding of data at the evaluation stage.
Since we appreciated that the census has limited space and resource capacity to include an entire of set of questions on the topic of language, we suggested that a rather simple amendment might add considerable value: Instead of the 2011 Census question pair
18. What is your main language?
19. (If the answer to 18 is not English) How well do you speak English?
we suggested that the 2021 Census might ask
18. Which languages do you use in the home?
19. If English is not your first language, how well do you speak English?
A possible alternative to 18 might be 'Which languages do you use regularly', but that might re-define the question from one about population composition to one about acquired skills (Canada and New Zealand ask this question).
At the time of writing, the 2021 Census is about to be discussed by the UK Parliament and unless amendments are adopted at this stage, it seems unlikely that the question will change. The response from the officials has been that they are reluctant to initiate a change for several reasons: First, the ONS follows a strict consultation process of its own in which it defines questions for consultations and puts them to a pre-selected groups of respondent organisations. Since the language question had not been put on that agenda, no responses were received on it; and since no 'user feedback' - as this particular process is referred to by ONS internally — was received on that question, ONS does not feel obliged to respond to it; the circularity in the argument is obvious. Next, officials said there was no space in the census questionnaire to accommodate amendments. They also suggested that processing multiple responses to the question would be costly. A further argument against a change was the need to ensure comparability of the data
across censuses. The officials went on to suggest that obtaining accurate data on language diversity was not a priority for the census's principal stakeholders, who are interested primarily in respondents' level of proficiency in English; question 18 on 'main language', they said, was merely a stepping-stone to question 19 on proficiency in English. Finally, changing the variables would also require time in order to test the new questions for quantity and the impact of data on users. The ONS's time plan foresees completion of a White Paper outlining the 2021 Census before the end of 2018, which would make it quite impossible to run such trials. Instead, there were two practical suggestions: The first was to improve the guidance notes for respondents in order to help remove the ambiguity of the question. The second, in the longer term, was to draw on other surveys in order to collate data on language diversity.
The first suggestion was taken up and our feedback is currently being considered for a reformulation of the guidance notes, though the problem of the absence of multiple entries remains. And so the best case scenario is one in which the ambiguity of the term 'main' might be reduced somewhat. At the time of writing, the outcome of the consultation process is yet to appear.
The latter suggestion — improving tools and collating data on language from different sources — is the objective of another MLM activity, coined the 'Multilingual Manchester Data Tool'6. The Data Tool, of which a pilot version was created online in 2018, brings together datasets on language for the Manchester area from a variety of sources: Census and School Census data, city library stock and loans and renewals by language, data on interpreter requests at hospitals and General Practitioner medical centres. These datasets have never before been collated in a single repository. The Data Tool offers a mapping element in which some of the datasets, those that cover multiple locations, can be plotted on a map, as well as a general repository where data tables can be viewed and downloaded. A set of static pages offers overviews by language including general background and selected statistics for the city, and by municipal districts (wards), featuring key data on language from the various sources including Census based self-reported level of proficiency of English by language. It is a way of raising general awareness of the city's language diversity and at the same
6 https://sites.manchester.ac.uk/mlm-datatool.
time a research tool and potentially an aid for understanding needs and planning provisions. Users can trace a language across areas and public services, discover the language profile of a selected area in the city, or assess the language needs in a particular service sector. The facility to triangulate datasets in each of these domains offers a more fine tuned path toward understanding the demographic distribution of speakers and an indication of language vitality, while combining the insights with the documentation of public signage offered by the LinguaSnapp tool enhances the picture to include the public presence of languages in the commercial and cultural sectors. The Data Tool also offers proof of concept: It demonstrates that cultural aspects of the city's population can find its way into the vision of smart cities, which so far has been devoted mainly to the physical environment (with a focus on such aspects as transport and physical accessibility of services). It also offers a comparison basis of different standards and practices of data compilation and formatting and in that way can help encourage service providers to share good practice and optimise approaches to data collection and data sharing.
In its current pilot format, some of the Data Tool's deficiencies in fact serve this very purpose, by providing an indication of just how difficult it is to obtain up to date datasets from key public services beyond issues of data protection. Not only are there gaps in annual records, for example, but in order to allow for a comparison, different datasets need to be adapted to a similar configuration around the three principal descriptor dimensions of time, place, and language. Of those, time is perhaps the easiest dimension to process, as data can usually be summarised by year, though here too there are mismatches, with some institutions gathering data by calendar year, and others by fiscal (budget) year. For place, data import protocols have had to convert outlet names such as schools or medical centres into postcodes (not least for data protection purposes) and then postcodes into municipal ward boundaries (which are potentially subject to change in between municipal elections). For languages, synonymous labels (such as Persian and Farsi), near-synonymous or related languages (such as Persian and Dari) and spelling variations (such as Bahdini and Badini) have to be taken into account along with the potential hierarchy of cover terms to the names of regional varieties (such as Kurdish, alongside Kurmanji, Bahdini and Sorani).
Image 1: Search result display from Multilingual Manchester Data Mapping Tool
6. Case study 2: A city language narrative
In the summer of 2013, we engaged a group of assistants among final year students of Linguistics to carry out a small survey of language provisions in Manchester. The outcome was a report entitled 'Multilingual Manchester: A Digest'7. The launch of the report in august 2013 at a public meeting with stakeholders from a variety of service providers was
7 http://mlm.humanities.manchester.ac.uk/wp-content/uploads/2015/12/ MLMDigest.pdf.
accompanied by a University press release8, which carried the headline "Manchester is Britain's City of Languages". The press statement reported that "the team based at the University's Multilingual Manchester project say there could be up to 200 languages spoken by long-term residents in the Greater Manchester area". On the same time, the Independent newspaper picked up the story, but with the headline "200 languages: Manchester revealed as most linguistically diverse city in Western Europe". Several other daily newspapers also carried the story. In March 2015, the UK's Association of Chief Police Officers launched the #WeStandTogether campaign to promote solidarity among groups of different faiths and cultures. The initiative was a response to the rise of far-right extremist attacks targeting in particular Muslim communities, in the aftermath of a wave of terrorist attacks by individual Muslim extremists across Europe. At the launch event, Manchester's Chief of Police, Sir Peter Fahy, referred to Manchester as a city of 200 languages. The statement was repeated many times on social media and became emblematic of the campaign's ethos and objectives to bring together people of various backgrounds and to celebrate diversity. In October 2015, MLM organised Levenshulme Language Day — a weekend event consisting of performances, stalls and family-friendly activities to celebrate languages. Local politicians were invited, and the local newspaper quoted a statement of support from the city council's Deputy Leader, who several months later, in February 2016, attended the launch event of LinguaSnapp at the University. In her speech at the event, she mentioned how languages can be a bridge to bring people together, and how they offer opportunities for the city's economy.9 Similar statements were made by other leaders of the city council at an event organised by MLM to mark UNESCO International Mother Language Day in February 2017, and were published on the MLM website.10
In May 2017, a suicide bomber attacked a music event at Manchester Arena, killing 22 people. The fact that the attacker, of Libyan background, was raised in Manchester, added to the shock of what was one of Britain's worst terrorist attacks. The city united in expressions of grief, which quickly embraced the motto of cross-community and inter-faith solidarity.
8 https://www.manchester.ac.uk/discover/news/manchester-is-britains-city-of-languages.
9 http://mlm.humanities.manchester.ac.uk/wp-content/uploads/2016/07/ Speech-23rd-Feb-2016.pdf.
10 http://mlm.humanities.manchester.ac.uk/living-in-a-city-of-languages.
The Guardian newspaper's report on the attack and the vigils that filled Manchester's streets cited in its second paragraph the fact that "200 languages are spoken in Manchester", linking, in its online edition, the University of Manchester press release from August 2013. The message was repeated many times on social media, as residents linked to the article and to the original University press release and expressed pride in the city's language diversity. In October 2017, MLM held its second Levenshulme Language Day. This time, many local politicians coordinated their visit to the event and broadcast it on social media, linking the message of language diversity with their commitment to support local community groups and community cohesion in general. In 2018, the city council's Libraries department adopted UNESCO International Mother Language Day as a regular annual event and invited MLM to participate in a planning group for activities across the city. In the spring of 2018, the city council initiated a crowd souring activity calling on residents to add their own lines, in their own languages, to a poem celebrating the city, called 'Made in Manchester', setting a target of 100 languages, while Manchester Museum launched its refurbishment programme under the motto 'hello future' in 50 languages and declared its intention to become a 'multilingual museum'.
turner Khan * mOM*
. ©umerknanMcr ^HHBv
"Over 200 languages are spoken in Gtr Manchester, reflecting the strength of our diversity' @gmpfahy #WeStandTogether
* .
■ M ■ fTtj [ m
........■
9:02 PM Way 2015
14 Retweels 11 Likes *.— ^ ^P£ & 9
Image 2: Twitter message citing former police Chief Constable Sir Peter Fahy at the launch of the #WeStandTogether campaign in 2015.
7. Conclusions
Above I discussed examples of the application of a new epistemol-ogy of studying multilingual communities, one in which researchers take an active part in shaping the public agenda of engagement with language diversity, and draw on that engagement as a setting for research. This approach creates new challenges. We might first ask: who are the beneficiaries of active engagement? The narrative that accompanies the research and that is communicated to the public, does more than just inform about enquiry, facts and discovery: It pitches language diversity as an advantage for the city and its residents, and awareness of language diversity and adequate responses to it as the responsibility of public agencies. The attempt to engage the public and public institutions with research findings, is thus an attempt to convince audiences of the benefit not just of the research itself but of a new assessment of the reality around them and of a vision for a society in which languages, and by proxy cultural diversity, are cherished.
At the wholesale level, the city and its residents are thus regarded as beneficiaries, as the public communications coming out of the project address issues such as improving accessibility of services (by assessing the efficiency of interpreting provisions, for instance), improving community relations (by celebrating diversity and community identities and building bridges among populations of different cultural and language backgrounds), and contributing to economic growth (by recognising languages as skills). Implicitly, the specific beneficiaries are those communities that are multilingual and seek the protection of public agencies for their efforts to maintain distinct identities including linguistic practices. Co-production ventures with public service providers in schools, the emergency services and the health care sector frame these service providers (front line practitioners) as explicit beneficiaries who stand to gain from the expertise and information that is gathered by the researchers. In the process, students are also beneficiaries by virtue of being the university's primary 'clients', through their direct involvement in project activities and the exposure opportunities that these activities open up to a range of civic processes, from the planning and delivery of public services to a culturally diverse population, and on to immersion with vulnerable or fringe groups such as refugees and minority community initiatives.
A core element of the new epistemology is thus the potential transformative effect that the research has on others, both within and outside the higher education sector. Viewed in its entirety, the model provides
a two-level script for the concept of the 'civic university': First, it creates an alternative narrative for the city itself. This is apparent in the manner in which the language diversity motto has been embraced in Manchester at a time when the search for a civic identity narrative has been propelled by a series of challenging events: The difficulties imposed by government austerity measures created a need to motivate community initiatives to take on new roles. The EU referendum campaign and the Brexit vote of June 2016 created tensions among communities. This was primarily a result of the Leave campaign's strong focus on curtailing immigration and embracing a national identity narrative that directly rejects globalisation and its implications for population diversity, often accompanied by explicit linguaphobic tones [Matras 2016]. The aftermath of the Arena Bombing, in this particular context, triggered a search for a binding communal narrative.
In this context, MLM's message is that language diversity should be embraced. Its outreach and engagement strategy is designed to alleviate the a priori abstractness of language by addressing practical issues around the role of language in improving service delivery, by highlighting language through performance and the arts as well as through technical visualisation, and by linking people's natural emotional attachment to languages, with a discussion of identity and community relations. The latter aspect in particular resonates initially with those who feel a need to compensate for a feeling of disempowerment, namely local immigrant and minority communities; but at a time of crisis it also resonates with mainstream organisations, finding its way to the top level of the city's leadership. Here, the iterative mode finds a clear expression in the fact that it is the research activity that first formulates principles. It then offers the city's leaders platforms at which to reiterate those principles. Finally, it documents that very reiteration and broadcasts it in order to amplify the message and reach new audiences. In this way, the stakeholders and declared target beneficiaries are in a certain sense also the conduits of the message from which they, and those whom they serve, stand to benefit.
In Matras, Robertson [2017] we addressed a series of challenges of the reciprocal model, some with particular reference to the competitive environment that rewards measurable research impact on non-academic audiences and competes to attract students by offering a uniquely distinctive 'experience'. These challenges included the intense process that is required to gain the trust of external stakeholders as genuine partners and not just as subjects of research and observation sites; the risk
of creating a dependency relationship once the research begins to deliver a regular service; the volatility of external partner organisations; the responsibilities associated with pastoral care of students around the volunteer scheme; and the dependency on short term resource investment from grants and university internal funds. To these we can now add three more areas of self-critical reflection that by necessity accompany the iterative engagement and outreach model:
First, there is the ethical dilemma of data management. Instruments such as the Data Tool and LinguaSnapp have the potential to expose language practices and the actors involved in them to scrutiny, and in extreme cases might even inadvertently help map them as targets for criticism or even attacks of various kinds (such as reinforcing the unfounded argument made during the EU referendum campaigns that there are "entire areas in our cities where nobody speaks English"). There are also the ethics of broadcasting a round number of languages, which in reality is difficult to support as a fully realistic representation of languages, which do not lend themselves to a simple, let alone permanent count as discrete and stable entities. The image of "200 languages" is necessarily a metaphor rather than a scientific finding. Language counting has been criticised as futile 'demolinguistics' in some of the recent literature on urban multilingualism [King 2015; Pennycook, Otsuji 2015]. But as explained above, the Data Tool shows the merits of capturing numbers as representations of singular events and acts of framing interactions. The metaphor of a large, round number of languages serves, in turn, a mobilisation objective, projecting a sense of multiplicity, which, in the socio-political context, challenges public narratives of cultural uniformity, fixed boundaries between communities and identities and national isolation.
Next, the model is pitched as having transformative potential for targeted beneficiaries who are public sector practitioners. In the impact-led research environment, there is an aspiration to measure such benefits by attesting how knowledge exchange has instigated a change in practice. But the measurability of such exchanges and their effect varies considerably. Schools, for instance, may regard interactive sessions on language diversity as an enrichment to curriculum activities and in this way be able to certify that new teaching content or delivery methods have been introduced. Providing expertise to public bodies such as the National Health Service, the Police, or the Office for National Statistics is, however, a very different process. Here, results may be acknowledged at a certain level of officials and may enhance their understanding
of problems but the pathways to change of practice are long and complex. In the NHS, we experienced interest in our findings from co-produced research on language provisions in access to primary care [Gaiser, Matras 2016a] but an inability to adopt structural changes at the level at which the research was commissioned, or to escalate the issues to the proper executive level. With Greater Manchester Police, very practical recommendations to change the style and content of letters sent to victims of crime proved difficult to implement due to a general overhaul of the force's approach to communications and the contracting of a new external provider for IT related matters, which includes the procedure for generating such letters. The reactions of ONS officials were described above.
Finally, as researchers seeking to influence public attitudes we are finding ourselves having to negotiate a fine line between the dissemination of research results and the propagation of possible solutions to problems, and the need to lobby, campaign and engage in advocacy in order to ensure that proposals are given due consideration and implemented. This means that the 'civic university' model requires researchers to have not just a methodology, but also a vision for society. Risk assessment and risk management strategies for this new role are still under development and will require considerable critical reflection in the future.
References
Albino et al. 2015 — V. Albino, U. Berardi, R. M. Dangelico. Smart cities: definitions, performance and initiatives // Journal of Urban Technology. 2015. Vol. 22. Issue 1. P. 3-21. Amin, Thrift 2017 — A. Amin, N. Thrift. Seeing like a city. Cambridge: Polity Press, 2017.
Arnaut et al. 2015 — K. Arnaut, J. Blommaert, B. Rampton, M. Spotti. Language and superdiversity. London: Routledge, 2015. Bagna, Barni 2009 — C. Bagna, M. Barni. A mapping technique and the linguistic landscape // E. Shohamy, D. Gorter (eds.). Linguistic Landscape: Expanding the Scenery. London: Routledge, 2009. P. 126-140. Bagna, Barni 2010 — C. Bagna, M. Barni. Linguistic Landscape and Language Vitality // E. Shohamy, E. Ben-Rafael, M. Barni (eds.). Linguistic Landscape in the City. Bristol: Multilingual Matters, 2010. P. 3-18. Barni, Extra 2008 — M. Barni, G. Extra (eds.). Mapping linguistic diversity
in multilingual contexts. Berlin: Mouton de Gruyter, 2008. Blackledge, Creese 2010 — A. Blackledge, A. Creese. Multilingualism: A critical perspective. London: Continuum, 2010. Blommaert 2013 — J. Blommaert. Ethnography, superdiversity and linguistic landscapes. Chronicles of complexity. Bristol: Multilingual Matters, 2013.
Blommaert et al. 2005 — J. Blommaert, J. Collins, S. Slembrouck. Polycen-tricity and interactional regimes in 'global neighborhoods' // Ethnography. 2005. Vol. 6. Issue 2. P. 205-235.
Blommaert, Rampton 2011 '— J. Blommaert, B. Rampton. Language and Su-perdiversity // Diversities. 2011. Vol. 13. Issue 2. P. 1-22.
Bonacina-Pugh 2012 — F. Bonacina-Pugh. Researching 'practiced language policies': insights from conversation analysis // Language Policy. 2012. No. 11. P. 213-234.
Cadier, Mar-Molinero 2012 — L. Cadier, C. Mar-Molinero. Language policies and linguistic super-diversity in contemporary urban societies: the case of the City of Southampton, UK // Current Issues in Language Planning, 2012. Vol. 13. Issue 3. P. 149-165.
Davies, Ziegler 2015 — W. V. Davies, E. Ziegler (eds.). Language Planning and Microlinguistics: from Policy to Interaction and Vice Versa. Houndmills, Basingstoke: Palgrave Macmillan, 2015.
Extra, Yagmur 2004 — G. Extra, K. Yagmur. Urban multilingualism in Europe. Clevedon: Multilingual Matters, 2004.
Extra, Yagmur 2011 — G. Extra, K. Yagmur. Urban multilingualism in Europe: Mapping linguistic diversity in multicultural cities // Journal of Pragmatics. 2011. Vol. 43. Issue 5. P. 1173-1184.
Gaiser, Hughes 2015 — L. Gaiser, P. Hughes. Language provisions in Manchester's supplementary schools. University of Manchester, 2015.
Gaiser, Matras 2016a — L. Gaiser, Y. Matras. Language provisions in access to hospital and primary care in Manchester. Report commissioned by the NHS Clinical Commissioning Groups, Manchester, 2016. http://mlm.hu-manities.manchester.ac.uk/wp-content/uploads/2016/09/Language-provi-sions-in-access-to-primary-and-hospital-care-Sept-2016.pdf.
Gaiser, Matras 2016b — L. Gaiser, Y. Matras. The Spatial Construction of Civic Identities: A Study of Manchester's Linguistic Landscapes, 2016. Accessed 17 July 2017. http://mlm.humanities.manchester.ac.uk/wp- con-tent/uploads/2016/12/ManchesterLinguisticLandscapes.pdf.
Garcia, Fishman 1997 — O. García, J. Fishman. The multilingual apple. Languages in New York City. Berlin: Mouton de Gruyter, 1997.
Heller 2011 — M. Heller. Paths to post-nationalism. A critical ethnography of language and identity. Oxford: Oxford University Press, 2011.
King 2015 — L. King. Multilingual cities and the future: Vitality or decline? // L. King, L. Carson (eds.). The multilingual city: Vitality, conflict and change. Clevedon: Multilingual Matters, 2015. P. 179-202.
Lamarre 2013 — P. Lamarre. Catching "Montreal on the Move" and Challenging the Discourse of Unilingualism in Quebec // Anthropologica. 2013. Vol. 55. Issue 1. P. 41-56.
Landry, Bourhis 1997 — R. Landry, R. Bourhis. Linguistic Landscape and Eth-nolinguistic Vitality. An Empirical study // Journal of Language and Social Psychology. 1997. Vol. 16. Issue 1. P. 23-49.
Li Wei 2018 — Li Wei. Translanguaging as a practical theory of languages // Applied Linguistics. 2018. Vol. 39. Issue 1. P. 9-30.
Liddicoat. Baldauf 2008 — A. J. Liddicoat, R. B. Baldauf. Language planning in local contexts: Agents, contexts and interactions // A. J. Liddicoat, R. B. Baldauf (eds.). Language Planning in Local Contexts. Clevedon: Multilingual Matters, 2008. P. 3-17.
Matras 2009 — Y. Matras. Language contact. Cambridge: Cambridge University Press, 2009.
Matras 2016 — Y. Matras. The case against linguaphobia. Manchester Policy Blogs, 2016. http://blog.policy.manchester.ac.uk/posts/2016/09/ the-case-against-linguaphobia/.
Matras 2017 — Y. Matras. Can global cities have a language policy? // Languages, Society & Policy. 2017. Vol. 1. Issue 1. Retrieved from http:// www.meits.org/languages-society-policy.
Matras, Robertson 2015 — Y. Matras, A. Robertson. Multilingualism in a post-industrial city: policy and practice in Manchester // Current Issues in Language Planning. 2015. Vol. 16. Issue 3. P. 296-314.
Matras, Robertson 2017 — Y. Matras, A. Robertson. Urban multilingualism and the civic university: A dynamic, non-linear model of participatory research // Social Inclusion. 2017. Vol. 5. Issue 4. P. 5-13.
Matras et al. 2016 — Y. Matras, A. Robertson, C. Jones. Using the school setting to map community languages: a pilot study in Manchester, England // International Journal of Multilingualism. 2016. Vol. 16. Issue 3. P. 353-366.
Matras et al. 2018 — Y. Matras, L. Gaiser, g. Reershemius. Multilingual repertoire management and illocutionary functions in Yiddish signage in Manchester // Journal of Pragmatics. 2018. No. 135. P. 53-70.
Mora et al. 2017—L. Mora, R. Bolici, M. Deakin. 'The first two decades of smart-cities research' // Journal of Urban Technology. 2017. Vol. 24. Issue 1. P. 3-27.
Peck, Ward 2002 — J. Peck, K. Ward (eds.). City of revolution. Restructuring Manchester. Manchester: Manchester University Press, 2002.
Pennycook, Otsuji 2015 — A. Pennycook, E. Otsuji. Metrolingualism: Language in the city. London: Routledge, 2015.
Stevenson 2017 — P. Stevenson. Language and migration in a multilingual metropolis: Berlin lives. Houndmills: Palgrave Macmillan, 2017.
Vasey et al. 2018 — H. Vasey, R. Tipton, H. Stanton, A. Robertson, E. Mosser, Y. Matras, K. Boddington. ESOl at Manchester Adult Education Service (MAES): Experiences of entry 3 and level 1 learners, 2018. http://mlm. humanities.manchester.ac.uk/wp-content/uploads/2018/10/ES0LatMan-chesterAdultEducationServicev6.3-1.pdf.
Williams 2003 — G. Williams. The enterprising city centre. Manchester's development challenge. London: Spon Press, 2003.
Wong 2015 — C. Wong. A framework for 'City Prosperity Index': linking indicators, analysis and policy // Habitat International. 2011. No. 45. P. 3-9.
Young et al. 2006 — C. Young, M. Diep, S. Drabble. Living with difference? The 'cosmopolitan city' and urban reimaging in Manchester, UK // Urban Studies. 2006. No. 43. P. 1687-1714.
