Abstract

The process of devolution in Wales over the past decade has changed the role and profile of the Welsh language as it has been increasingly promoted and utilised in policy by the National Assembly for Wales (NAW). Little research has been carried out to explore the affect of these changes on specific spaces with specific area-based characteristics and issues. This paper explores the role of the Welsh language in policy at local level of governance within areas experiencing high levels of multiple-deprivation. Using data from an empirical research study the paper applies one stand of the Linguistic cosmopolitan theory to identify and explore the issues surrounding Welsh language policy promotion and utilisation within regeneration partnerships in Wales. The paper concludes that in connection with a lack of Linguistic cosmopolitanism in the case study areas, deprivation can be seen to have a direct impact on the way that partnerships promote the Welsh language. The paper states that the NAWs strategic promotion of the Welsh language, arguably, has a little impact in the context of deprivation. From this statement the paper seeks to draw out significant related issues to do with exclusion, the utilisation of culture and identity and their policy implications.

Introduction

Since 1999 the Welsh language has become a principal cross-cutting policy within communities, housing and regeneration under the National Assembly for Wales (NAW). The advent of devolution under New Labour has boosted the role, profile and legislative status of the Welsh language in Wales following on from a language policy trajectory by a minority of influential Welsh activists and political organisations prior to 1997. Academic studies on the Welsh language prior to 1997 look at the role of the language within nationalist movements and from a historical perspective (e.g. Balsom et al., 1984). A number of studies on the Welsh language after 1997 have tended to focus on the social significance of the spatially-based distribution patterns of Welsh speakers in Wales (Aitchison and Carter, 2004, Coupland et al., 2006). Few studies combine distinctive area-based characteristics and spatial-dispersal patterns of Welsh speakers with Welsh language policies and approaches since 1999.

In August 2008 Leighton Andrews, the Deputy Minister for Regeneration in Wales, announced a £1m package as part of the Assembly’s pledge to incorporate the Welsh language into regeneration. This stems from the Welsh Assembly Government’s (WAG) policy document ‘Iaith Pawb (Everyone’s Language): a national action plan for a bilingual Wales’ (2003). The Assembly’s pledge highlights the utilisation and cross-cutting approach to language promotion in Wales and the perceived connection between language-based identity and regeneration. Within this context the paper assesses the role of the Welsh language within the framework of tackling area-based deprivation through urban regeneration in Wales.

Using the regeneration programme Communities First (CF) in Wales as a case study this paper explores the use, status and perception of the Welsh language through in-depth interviews with local regeneration partnership representatives at community and ward level.

CF is a flagship programme under the NAW which has been running since 2001 to tackle deprivation in Wales’ poorest areas.  The programme has 150 partnerships and 10 themed areas running in the most deprived areas of Wales.1  CF partnerships are required under the Welsh Language Act of 1993 and according to the Welsh Assembly.  Government’s (WAG) Action Plan for a bilingual Wales (2003) to produce every printed document bilingually. CF guidelines (2007) produced by the WAG state that all notices, advertisements and other written documents should be printed bilingually. They also state that the ‘linguistic profile’ of an area should be taken into account when recruiting partnership employees in deciding whether or not to appoint a fluent Welsh speaker. It is the partnership co-ordinator’s responsibility to ensure linguistic equality.

‘In areas where the Welsh language is widely spoken in the local community, consideration should be given to whether the ability to speak, read, write and/or understand Welsh should be a requirement for the post’ (WAG 2007: 80)

In studying cross-cutting policy implementation CF partnerships bridge a strategic gap in governance between the NAW, LAs and local communities in a process referred to as ‘double devolution’ (Miliband, 2006; Adamson and Bromiley, 2008).

In 2009 there will be a budget-cut in the funding given to CF partnerships for bilingual publications;2 this is in direct contrast with the ethos of the £1 million for incorporating Welsh into regeneration and raises questions about the way that Welsh is being promoted and, linked to this, the motives behind its promotion in relation to regeneration and area-based deprivation. In light of this the Welsh Language Board’s approach to marketing the Welsh language refers to its value as a tool for economic and social advancement.

‘The perceived instrumental value of the Welsh language, providing access to socio-economic advantage, is claimed as a marketing tool by the WLB…’ (Chriost, 2007: 75)

It is possible to argue that the unofficial ‘switching’ of funding allocation from partnerships to a higher profile initiative highlights the importance of visible marketing techniques in promoting of the language. One possible motive for promoting the Welsh language linked to this issue is the symbolism of language-based identity rooted in nationalism. Within the context of regeneration the use of symbolism in policy shifts the debate on nationalism to one of area-based characteristics and disadvantages. While people in deprived areas use Welsh in their day to day life this research shows what appears to be a gap between the policy and the practice in language promotion and policy in deprived areas.

This paper begins with the development of a linguistic cosmopolitan theory drawing on findings from two pieces of research on bilingualism and minority languages in two different geographies. The theory is then laid out as a hypothesis through which to frame the subject of bilingualism in regeneration. Details from the findings of empirical research are given followed by an analysis which applies the findings to the hypothesis. The paper ends with concluding commentary on the possible meanings and implications of minority language policy in the context of regeneration and area-based deprivation.

Linguistic cosmopolitanism and Wales

A ‘new wave’ of academic research on minority languages in policy and governance took place during the 1990s following the creation of the European Charter for Regional or Minority Languages (1992). The Charter pledges commitment to the protection and support of minority languages in EU Member States (See Mouthaan 2007 for critique on the Charter).

Research following the increased profile and promotion of minority languages has tended to focus on the impact of national policies and trends on minority languages and their speakers. These have been conducted in areas where minority languages exist such as Wales (Cardinal and Denault 2007), French-speaking Quebec, the Basque Country and the Autonomous Community of Catalonia (Newman et al., 2008). Cardinal and Denault (2007) and Newman et al. (2008) look at the implications and impact of policy development and global trends on the role of minority languages in the countries and areas where they exist.

Cardinal and Denault (2007) conduct a comparative study of language policy and group politics in the context of neo-liberalism in French-speaking Quebec and Wales with the aim of uncovering the affects of global neo-liberal policies on minority language and group politics. Their research assesses whether or not neo-liberal language policies and changes can empower minorities. Newman, et al (2008) study the ‘language attitudes’ (2008) of ‘linguistically diverse’ young people in Catalonia following the instigation of official policies favouring Catalan. Both papers argue that there exists ‘a new cosmopolitanism’ (Cardinal and Denault 2007) or a ‘linguistic cosmopolitanism’ (Newman et al 2008) in which linguistic pluralism becomes a tool used both to the advantage of its speakers and as a reflection of a progressive and cohesive society. Cardinal and Denault (2007) argue that despite neo-liberal influences language plans in Canada and Wales show a promise of empowering minority groups. They conclude that the empowerment of minority groups through language policy depends on ‘…the capacity of groups to engage with existing opportunity structures and to use them in atransformative way‘ (2007: 455). This implies that organised collectives can potentially have influence over policy proceedings to their advantage.

The process of regeneration takes a holistic approach to tackling deprivation, involving a focus on the social, geographical and economic aspects of the problems. This paper acknowledges that structural factors play the key role in tackling area-based economic problems and that bilingualism does not necessarily have a large-scale impact on this aspect of deprivation. However, drawing on the research by Cardinal and Denault (2007) and Newman et al. (2008) there can be non-economic advantages to bilingualism, for example, in the case of minority groups influencing policy and political actors or linguistic pluralism as a refection of a progressive society, which could lead to indirect economic, social and geographical advantages for an area. Linguistic cosmopolitanism as a term to describe the potential area-based social, economic and geographical benefits of bilingual policy will form the hypothesis for this piece.

The following sections apply the theory of linguistic cosmopolitanism in the context of area-based approaches to tackling deprivation through partnerships in Wales. This aims both to develop the theory within a different economic, geographical and social context and to explore its use as a framework for expanding on the issue of bilingualism within CF.

The empirical research carried out in this study was completed in two stages using primary quantitative data to narrow the focus for a more in-depth qualitative approach. The research method used in the initial phase was an email survey sent out to individuals within CF partnerships. The results were then used to design a set of research questions within a semi-structured interview for use in the second part of the research.

In stage one of the empirical research two partnerships were chosen randomly from each of the 22 local authority (LA) areas in Wales (three in one LA area) to achieve an even spatial distribution of partnerships within the framework of the LAs in Wales. The six areas created in the Wales Spatial Plan (WSP) (2004) were used as a means of better understanding the area-based characteristics pinpointed by the Assembly from within which the 53 partnerships were chosen. The Welsh Index of Multiple Deprivation (WIMD) (2005) was also used as a way of deciphering the areas of high deprivation on which CF partnership distribution is based.3

From this selection process 53 CF partnership co-ordinators were contacted by email and asked for any information relating to projects involving bilingualism and specific work or documentation citing bilingualism as a tool within partnership working. Of the 53, 10 co-ordinations responded (19 per cent response rate). To compensate for the low response rate desk-based research was carried out using the Communities First Support Network (CFSN) to identify any bilingual projects taking place in the 53 partnerships; this produced no significant findings. The majority of respondents (eight out of 10) stated that they had no specific bilingual projects running in their areas:

‘Bilingualism is not a hot issue on these two estates, so we do not have any work to share with you’ (CF Coordinator)

The findings showed only one example of a significant move towards utilising bilingualism; as a way of encouraging youth participation. In this case there was already a strong Welsh-speaking majority in the area. The initial scoping stage highlighted a fundamental lack of positive responses from partnerships where Welsh was not already an issue. In light of the low response rate and the lack of positive response regarding the use of bilingualism in the selected partnerships, the area-based selection process for further research took a more strategic approach. Using the WSP (2004), information from the WLB and the 2001 National Census three LAs with a Welsh speaking majority (of 50 per cent or over) were selected. The LAs that fit this criterion were: Gwynedd (68 per cent are Welsh speakers), Anglesey (59 per cent) and Ceredigion (51 per cent). By narrowing the spatially-based parameters of the research in this way the project targets LAs in Wales where Welsh is more likely to be a significant issue.

All 15 CF partnership co-ordinators within the three LAs were contacted by phone4 and asked if they would be willing to participate in the project, of these six agreed to take part. Interviews were semi-structured and conducted over the telephone with five CF partnership co-ordinators and one development worker. The interview schedule was designed to move from questions on the bureaucratic and obligatory approaches to promoting bilingualism in partnership through to questions on the impact of bilingual policy within the NAW on the partnerships. The schedule was also aimed at gaining a clearer picture of the way that partnerships juggle the incorporation of the Welsh language policy into projects, either because of obligation or on their own initiative, with the interaction that they had with a mix of bilingual and non-bilingual community members.

The six areas had a similar population size of around 2,000 people and high derivational problems dating back to the closure of traditional industries, including coal and slate mines, during the 1980s and 90s. Of the areas that had available micro-level data relating to ‘knowledge of the Welsh language’ most had a higher number of people who could understand spoken Welsh than people who could read, write or speak Welsh. Unlike the other partnerships both the wards in Gwynedd self-reported a low number of Welsh speakers and were referred to in the interviews as ‘anglicised’.

Findings

Within the interview analysis there were a number of significant and recurring themes which can be summarised into five key points:

• The Welsh language is a low priority for partnerships in the context of tackling deprivation
• There is a lack of motivation to promote the language within partnerships beyond the bureaucratic (relating to the first theme)
• There are issues around ‘different types’ of Welsh, for example, that are taught formally within the context of a policy initiative as opposed to learnt socially as a mother tongue
• There is an ethos of ‘supporting’ rather than ‘promoting’ the language
• The future of Welsh as a minority language in the context of global and national political changes.

The findings have implications for bilingualism both in the context of area-based deprivation and in terms of the linguistic cosmopolitan hypothesis. One possible implication is that the NAWs promotion of bilingualism has not materialised in the context of tackling deprivation and regeneration as it has done in other policy areas, such as education. When considering the holistic nature of regeneration this could have further area-based implications for the existence of a linguistic cosmopolitanism. Despite being conducted within areas with a Welsh-speaking majority, the research findings do not identify any instance of Welsh being used as a tool to further collective or group agendas, neither is linguistic pluralism seen as a reflection of a progression or cohesion within communities. This is unsurprising considering that all speak English as a common language. Bilingualism as a tool, and viewed in this way as a problematic side issue for partnerships working to tackle deprivation, is therefore unlikely to have as significant a political influence as the linguistic cosmopolitan theory implies. Inside and outside of the context of regeneration partnerships the findings point to the possibility that bilingualism is, at best, a secondary policy issue in these areas of high deprivation.

The Welsh language is seen as a low priority for partnerships in the context of tackling deprivation which is understandable given the high levels of deprivation in all areas. In the context of CF partnership remits, a number of respondents stated that the Welsh language is a secondary obligation at best. Promoting language is seen as relatively unimportant for partnerships in the context of tackling deprivation and focusing on themes for improving and regenerating communities. Welsh is cited as ‘taking a backseat’ to more important issues such as reducing crime rates and unemployment. In the two most anglicised areas both co-ordinators refer to a lack of interest in promoting the language due to a lack of demand for this from local people. The reasons given for this included the arrival of incomers to the community with ‘different values’ who did not see the Welsh language as important; this was especially prevalent in two of the areas that have high numbers of tourists:

‘I can’t see [the Welsh language] taking any higher profile than it does now, people moving into an area bring different values and to them the language doesn’t seem important’ (CF Partnership Co-ordinator)

This quote touches on the possibility of Welsh as an impediment to cohesion and the implications of language-based identity separate from government policy. Another reason given for not prioritising promotion of the language is the fact that local people speak Welsh with each other but see CF partnerships as separate from everyday social life involving Welsh:

‘Communities First is seen as an English thing … and the partnerships themselves don’t push the Welsh language issue’ (Partnership Co-ordinator)

This points to people being in control of when and with whom they speak Welsh and seeing the governance of language as at best unimportant and at worst threatening. In the context of area-based deprivation, and separate from the partnership’s work, one co-ordinator also points out the potential difficulty of pro-actively learning a second language for people living in poor conditions.

‘You know, when you have to make ends meet and you don’t have as much time, you’re just surviving and the bills need paying, you don’t really see learning Welsh as an important thing. Middle-class people can make more of a choice and have the resources to make the effort’ (CF Partnership Co-ordinator)

Because of the issues outlined above there is a lack of motivation to promote the language within partnerships beyond the bureaucratic. The co-ordinator quoted above stated that ‘a strong culture would sustain itself’ and that ‘pushing’ the language onto people would ultimately lead to its decline. In contrast to issues such as high crime rates, high unemployment and drug problems the profile of the Welsh language does not directly affect the quality of peoples’ lives and therefore does not take as high a priority in this context.

Issues around ‘different types of Welsh’ emerge and the divide between socially-used Welsh and Welsh that is taught and promoted because of bureaucratic requirements is made by more than one respondent; in some cases this divide equated to the difference between spoken and written Welsh and first-language and second-language speakers respectively. The phrase ‘language on the street’ or similar is used more than once in referring to the difference between socialising in Welsh on a day to day basis and being able to speak, read or write Welsh formally; one respondent referred to ‘social Welsh’ being in decline. Another co-ordinator stated that if Welsh is not used socially then its evolutionary progression will halt and the language will become tokenistic.

‘When the kids from our town speak to the kids in [the neighbouring town with a higher percentage of Welsh speakers] they sound like country bumpkins … if children do not speak Welsh at home or with friends then their knowledge of the idioms and evolutionary changes in the language is low’ (CF Co-ordinator)

In all partnerships there were two or more primary schools teaching Welsh as a first-language but only one High School teaching Welsh as a first-language. Respondents stated that in most cases parents chose the English first-language high schools for relatively unrelated reasons including distance, location and the quality of schooling. Significantly, another reason for parents choosing English first-language high schools included actively not wanting their children to learn Welsh for fear it might confuse them. Contrary to this one respondent notes that Welsh is playing more of a role in partnership working than two years ago:

‘There are now Welsh courses run here and the residents are attending Welsh courses also parents are showing an interest because they want to understand their child’s homework’ (CF Development Worker)

There is also some anecdotal evidence of a divide between the politicisation of Welsh and the elderly Welsh speakers. In one case an elderly first-language Welsh speaker said that he ‘doesn’t want anything to do with those Plaid Cymru snobs working from their planning offices in Cardiff’. This implies a gulf between the promotion of the Welsh language by government and the indigenous Welsh speakers who don’t feel a part of that promotion; a tentative example of possession of language promotion meaning possession of language for political purposes. There is also a class issue hinted at which could be an important avenue for further research in the context of bilingualism and deprivation. The idea that ownership of language by individuals or governments plays a part in compounding class divides and creating a ‘Welsh speaking elite’ could be drawn out of these findings, however further research would be needed in this specific area to make any solid claims or statements.

Most respondents state that their partnership aimed to ‘support’ rather than ‘promote’ the Welsh language. By complying with CF Guidelines (2007) on bilingualism and giving people the option to interact and conduct business in Welsh, the majority of respondents do not see the need to actively promote the language and encourage people to speak. In one case the respondent feels that actively promoting or ‘pushing’ a language would ultimately discourage people of all ages from learning.

Another respondent feels that there were a number of Welsh speakers living in the area that were not becoming involved with partnership working, termed ‘hard to reach’ residents. The respondent thinks that this needed to be addressed by the partnership as a problem. This again suggests another avenue for further research around the issue of inclusion and exclusion within multi-level governance.

In terms of putting Welsh in a minority language discussion on a global scale, one respondent was aware of global factors effecting minority culture and languages such as ‘the commercial mechanisms of Americanisation’. In addition respondents felt that the Welsh language was becoming increasingly tokenistic.

‘In 10 – 20 years time the Welsh language will be worn as a badge to enable the people who can speak it to get better jobs, but it will not be used in day to day life’ (CF Partnership Co-ordinator)

Analysis

Linguistic cosmopolitanism, as defined earlier through drawing on the work of Cardinal and Denault (2007) and Newman et al (2008), involves the utilisation and promotion of minority languages both for the purpose of pursuing policy aims and in reflecting a more harmonious and cohesive society. What is highlighted in the findings of this research and in the context of regeneration and deprived areas is that language is not a reflection of a cohesive society, rather it is seen more as an obligation that is difficult to fulfil due to time and resource restrictions within partnerships working in deprived areas. It raises important questions around whether area-based disadvantage has any bearing on the emergence and existence of linguistic cosmopolitanism. While the findings discussed here hint at some possible conclusions, it is not possible to provide any definitive answers to these questions based on this small scale study.  However, in terms of applying the theory of linguistic cosmopolitanism it is possible to make some headway.

Based on the findings in this study that the role of the Welsh language in partnership working to tackle deprivation is seen as unimportant and mainly a bureaucratic obligation carried out through supporting rather than promoting, then this highlights some weakness in the linguistic cosmopolitanism theory. Perceptions of the Welsh from a top-down and bottom-up perspective create a juxtaposition that is difficult to resolve. Cardinal and Denault (2006) pointed to minority language as a route through which influential minority groups could negotiate political agendas, however, from a partnership perspective it is a peripheral issue. This reaffirms the divide between Welsh spoken by people day to day and Welsh as tool to pursue policy objectives. Linguistic cosmopolitanism is not materialising in the deprived areas studied for reasons such as lack of resources and time to learn and lack of motivation in light of bigger problems. Linguistic cosmopolitanism failing to flourish in deprived areas because they are deprived points to the issue of the possibility of compounded area-based inequality. The incorporation of bilingualism into area-based measures to tackle deprivation is extremely problematic for those working to tackle deprivation. These complex issues relate to two problematic areas; whether or not the NAWs promotion of the Welsh language and its ineffectiveness in certain areas is exclusionary and whether or not it is feasible to politically engineer a identity without being exclusionary.

Welsh does not appear to have a significant role in day to day partnership activity and its inclusion is more of a ‘tick-box exercise’. While there is a realisation among partnership co-ordinators of the importance of having some Welsh language skills and of the vocational value in speaking Welsh in a country where it is pro-actively promoted by government, this is not notably pursued or indeed always possible to pursue. The advantages of speaking Welsh are engineered and produced by the NAW and are therefore inherently a product of Welsh promotion as a symbolic political issue. What real advantages can these have for people living in deprived areas? From the top-down perspective Welsh is a product rather than a ‘real’ encapsulation of area-based identity. From the bottom-up it is a complex and organic entity which comprises different meanings to different people. The assimilation of these two perspectives is one which can only be achieved through a combination of cultural and policy processes.

The emerging issue of ‘formal Welsh’ or ‘taught Welsh’ versus the social use of Welsh shows a consensus within the research that Welsh as a social and day to day vernacular is in decline and that the promotion by the NAW is limited to bureaucratic actions, for example the fact that a number of people fill forms out in Welsh but do not speak it. The use of the words ‘taught’ and ‘learnt’ by co-ordinators to describe different ways of gaining knowledge of the Welsh language bring to mind a distinction between imposed and voluntary. This raises questions about the way that the NAW is promoting the Welsh language and the motives behind a bureaucratically-centred and rhetorical approach. A bureaucratic approach to language promotion does not necessarily directly increase the aspiration among people living in Wales to speak the language. In light of the fragmented and diverse distribution of Welsh speakers in Wales it could be argued that the NAWs bureaucratic promotion is not addressing this diversity in full.

Welsh was not mentioned in the context of capacity building or in playing an active part in project work. This could suggest a flaw in the promotion of the language by the NAW because it is certainly an indirect advantage for people living in Wales to have some knowledge of Welsh for better employment opportunities. Therefore in terms of long-term life improvement, having some knowledge of the Welsh language could potentially improve the vocational chances of the people living in deprived areas.  In this case using Welsh in capacity building could be a positive move. Another area of separation between local governance and individuals in this context was highlighted in the feeling that people who speak Welsh in the community do not do so in the context of partnership involvement because of the perceived view of CF as ‘an English thing’.

In terms of looking at language as a commodity in linguistic cosmopolitanism it is possible to argue that the NAW’s promotion of the Welsh language exemplifies utilisation of the language as a commodity. This is also evident in WAG policy documents outlining the value and potential gains in giving Welsh equal status to English in order to make it a source of pride for people living in Wales. However, the outcomes from the research show that ‘on the ground’ and in the context of community regeneration, there is little motive or action for partnerships to get involved in promoting the Welsh language. If linguistic cosmopolitanism, as it is developed in the studies by Cardinal and Denault (2007) and Newman et al. (2008), means minority language as an social, economic and political advantage to the people who speak it in terms of reflecting social cohesion and gaining political influence, then little of this is being played out at a local level of governance in the case studies explored here. The theory could therefore be criticised for failing to account for a partnership perspective and for failing to acknowledge the complexity of distinctive area-based characteristics and different groups of people.

Issues around which spatial or geographical level the Welsh language policy is impacting on and subsequently whether there is an added disadvantage here for people living in deprived geographical areas are highlighted as important for further consideration within this paper. In addition, the context of Welsh language policy promotion shows significant difficulty in aligning a cross-disciplinary policy.

In terms of tackling deprivation through regeneration, while CF partnerships are successful in their support of Welsh language speakers through bureaucratic processes that respond to demand, there is little scope for partnerships to go further. Within the development of the theory of ‘new cosmopolitanism’ and the role of minority language within this Cardinal and Denault (2007) state that:

‘…neo-liberalism has provoked an important paradigm shift in the area of language policy making and planning. It has led to a new attitude towards languages informed by a new cosmopolitanism which places linguistic pluralism at its centre’ (2007: 16)

This study suggests that this is not the case in area-based regeneration partnerships tackling deprivation. The reasons for this identified in the paper can be explained, among other factors, through a structural problem with the way that the Welsh language is being promoted by the WAG. Linked issues include the juxtaposition of the research by Cardinal and Denault (2007) using Wales as a case study in developing the theory of linguistic cosmopolitanism and the lack of evidence to suggest it exists in reality as a concept in the context of deprived areas. Further research would be required to delve any deeper into the issue of whether or not deprivation has an influence over the existence of linguistic cosmopolitanism and the potential class issues attached to this, however it is fair to say that partnerships working to tackle deprivation at present do not find it easy to pro-actively promote the Welsh language.

Conclusions

The theory of linguistic cosmopolitanism cites language as a commodity which, in some cases, is promoted by governing structures and policies. Based on the perceptions of CF partnership co-ordinators within this study there is an understandable lack of priority given to promoting the Welsh language in the context of tackling deprivation. There are two main overarching issues of significance that stem from the discoveries made in this paper. Firstly the issue of deprived and non-deprived areas and the way partnerships within those areas promote the Welsh language. By limiting the research to areas in need of regeneration to tackle deprivation this research cannot make the claim that non-deprived areas have more success in promoting Welsh or what this could mean for area-based inequalities, however the findings from this study clearly show that deprivation has a direct affect on partnerships’ and actors ability to promote Welsh. Secondly, and linked with the first issue is that of modes of minority language promotion. The bureaucratic approach to promoting the Welsh language is not perceived to be successful or effective in the context of area-based regeneration partnerships. Based on this the NAW could, arguably, have more success in reaching ‘hard to reach’ Welsh speakers and influencing non-Welsh speakers if it branched out its currently limited marketisation approach. This conclusion is rooted in the complex contextual variations of meaning and values placed on the Welsh language by different people and the symbolism of the issue as a policy domain.

Through the application of the concept of linguistic cosmopolitanism to the empirical study of the role of minority language in local regeneration, this paper argues that while bilingualism is potentially an advantage for individuals and groups in a country / region where the language is being promoted by government, there is a lack of resources and ability to do this through partnership working in deprived areas. The nature of NAW promotion of the Welsh language as a symbolic form of nationalism does not permeate to the partnerships in deprived areas and can therefore be seen as divisive and exclusionary. Linguistic cosmopolitanism raises issues of elitism and favouring of particular groups in both a positive and negative light, in the case of partnerships within the findings of this study we can see ‘the other side’ of the minority language coin. That is the side that is, arguably, not benefiting from the symbolic policy promotion.

The issue prompts three related questions.  First, how can the NAW promote Welsh in a different way that would impact on partnerships and individuals in deprived areas?. A second more fundamental question is whether it is possible to politically engineer the revival of a language? Finally, there is the question of who benefits from successful language promotion. In answer to these questions this paper concludes that if the NAW wishes to promote the Welsh language for nationalistic, symbolic and cultural reasons and ‘as a source of pride for us all’ then it must take a more meaningful approach that goes further than the current ‘face value’ bureaucratic one in order to reach deprived areas. While engineering the benefits of speaking Welsh within employment in Wales the NAW must also be wary of the potential for exclusion and division through policy among the changing demographic of Welsh speakers in Wales.