Migration through a language planning lens: a typology of Welsh speakers’ migration decisions

ABSTRACT

This paper seeks to explain the migration decisions of minority language speakers by investigating motivating factors. Viewed through a language planning lens, the study pushes the parameters of some of the discipline's more recent agency concepts within the context of migration. Semi-structured interviews were conducted with 60 Welsh speakers aged 18–40 who have stayed, left, or returned to the Welsh language ‘heartlands’ to create a typology representing the diversity in speakers’ priorities when deliberating migration decisions. The data shows that the Welsh language is a significant consideration in the migration decisions of some typology groups, however most groups prioritised other factors. It was found that, in some cases, employment was a means of sustaining speakers within the linguistic community or attracting them back, offering much-needed evidence to support key tenets of the Welsh Government's current language strategy. Furthermore, given the likelihood of minority language speakers’ decision to stay, leave or return to a language ‘stronghold’ to increase or limit opportunities to use the language, we argue that migration is an important context for probing the use of agency by minority language speakers at a micro level. Consequently, we contend that migration should receive greater attention from language planning scholars.

KEYWORDS:

• Language planning
• language policy
• migration
• agency
• minority language
• Welsh

Introduction

This paper investigates the migration decisions of minority language speakers through a language planning lens, offering a new perspective on an age-old process. Migration was recognised as a major trend for language planners several decades ago (Tucker, 1994) however the limitations of the language planning discipline to address such a phenomenon have also been reputed (Tollefson, 2002). More recent interest by language planning scholars in the concept of agency, and the belief that language planning takes place on a continuum of levels from the macro to the micro (Kaplan & Baldauf, 1997), has paved the way to consider the influence of migration processes on different types of language policy and planning, and vice versa (see Payne, 2017; Tunger et al., 2010; Wei et al., 2022). However, the demand from grassroots language planning practitioners for evidence of what specifically drives minority language speakers to migrate from areas where there is a higher density of minority language speakers, thus influencing the sustainability of these languages, remains an overlooked area of research (Mentrau Iaith Cymru, 2020; Powel, 2022). This paper responds to this demand from Wales by presenting a typology, based on the typology of ‘Stayers’ and ‘Leavers’ of rural island youth developed by Cooke and Petersen (2019), conveying the considerations that influence Welsh speakers’ decisions to stay, leave or return to the language's heartlands; whilst probing the parameters of some of the more recent agency concepts of language policy and planning within a new context. In doing so we contribute both practically and theoretically to the field.

Study background

The 2021 census reported 538,000 Welsh speakers aged 3 and over in Wales, a decrease in percentage (1.2%) and in absolute numbers (23,700) since the last census in 2011 (Welsh Government, 2022). This is the second consecutive census that has reported a decline in the number of Welsh speakers and the lowest percentage to be recorded in a census to date (Welsh Government, 2022). Although there was a slight increase in the percentage of speakers in the capital city and some post-industrial counties of the South, the percentage of speakers decreased in all other counties, including counties traditionally considered to be strongholds of the language such as Ceredigion, Carmarthenshire, Gwynedd and Anglesey (Welsh Government, 2022). Although the concept of a language ‘strongholds’ or ‘heartlands’ is disputed, with Aitchison and Carter stating in 2000 that ‘there is no longer a solid Bro Gymraeg¹ in the sense in which it was originally conceived’ (p. 135); these are the four counties in Wales with the highest percentages of Welsh speakers (Statscymru, n.d).

It is important to note that alternative data sources such as the Annual Population Survey (APS) contradict some of the census data. For example, although both sources include the same language question, the APS shows there to be many more Welsh speakers in 2021 (884,000) than the census (538,000) (Welsh Government, 2023). Inconsistency between the two sources is a longstanding issue, perhaps explained by differences in how language ability is reported, however this was the first time that the respective trajectories diverged (Jones, 2019; Welsh Government, 2023). As the discrepancy is not fully understood, the Welsh Government and the Office for National Statistics have committed to come together to examine some of the differences observed (Welsh Government, 2023). Meanwhile, the Welsh Government continues to use the census as its main authoritative source to monitor the number of speakers in Wales in pursuit of their target of one million Welsh speakers by 2050 (Welsh Government, 2023).

Welsh Government's language strategy, Cymraeg 2050, aims for a million Welsh speakers while promoting daily Welsh usage and fostering favourable conditions for language growth (Welsh Government, 2017). It stresses sustaining and growing communities with high Welsh speaker density, crucial for language vitality, yet facing mounting pressure. The percentage of small areas, known as LSOAs (Lower Layer Super Output Areas), with a population of 70% or more Welsh speakers has fallen from 3% in 2011 to 2% in 2021, and indeed, such areas are now only found in Gwynedd and Anglesey (Welsh Government, 2022). Moreover, Welsh speakers are increasingly unlikely to reside in Welsh-speaking communities in 2021 compared to 2011 (Welsh Government, 2022), indicating a potential threat to the sustainability of these areas.

Jones (2010) contends that any reduction in the number of Welsh-speakers is likely to be due to outward migration (other than deaths, and if the possibility of losing the ability to speak the language is discounted) and any reduction in percentages is likely to be due to immigration (as cited in Wynn, 2013). The Welsh language can therefore be supported by encouraging its speakers to stay or return to Welsh-speaking communities and the Welsh Government believes that supporting the socio-economic infrastructure of these communities is a means of doing so (Welsh Government, 2017, pp. 59–62). However, the possibility of losing the language in situ cannot be discounted; and whilst the high-profile slogan of Cymraeg 2050 may be to increase the number of speakers to 1 million, increasing language usage is also a key aim of the policy (Welsh Government, 2017). As such, migration is but one policy concern within a much broader policy agenda including, for example, language acquisition, education, the workplace, access to Welsh language services and community use of the language (Welsh Government, 2017). Nevertheless, migration remains an important area of policy and stubbornly persists as a challenge for policy makers with language sustainability goals.

Arfor, aligned with the Welsh Government's language strategy, aims to bolster Welsh-speaking communities through economic interventions, fostering daily use of the Welsh language (Arfor, 2022, p. 9). A key goal is to retain or attract young people and families to these communities by assisting in business development or securing suitable employment aligned with their aspirations (Arfor, 2022, p. 11).

The correlation between economic dynamics and linguistic sustainability has been highlighted by prominent scholars (Crystal, 2002; Grin, 1996; Kaplan & Baldauf, 1997; Williams, 2000). Aitchison and Carter (1987) specifically observed a trend of young people migrating from Welsh language strongholds in the west of Wales to pursue employment opportunities in Cardiff. This underscores the relevance of initiatives like the Arfor programme as a proactive response to academic findings, despite the passage of over three decades since Aitchison and Carter's research. However, there is a dearth of evidence to support the hypothesis that providing employment within the Welsh heartlands is an effective means of alleviating the emigration of Welsh speakers from these areas (Bonner et al., 2024). Furthermore, Lewis and McLeod (2021) caution that providing jobs within the heartlands of a lesser-used language does not benefit the language unconditionally; it could contribute towards the normalisation of the majority language if a workforce is employed that does not use the minority language. Consequently, the primary aim of this paper is to address the need to improve our understanding of the role of employment, alongside other motivating factors, in the decision-making process of Welsh speakers to migrate, advocating an evidence-based approach to policies designed to support Welsh language speakers to stay or return to Welsh speaking communities.

Research context

Although it is a commonly held belief that factors influencing the migration decisions of young people from rural areas of Western Europe are numerous and complex (Alexander, 2016; Ní Laoire, 2000), earlier studies have attempted to identify a dominant motive, such as economic or academic considerations (Ní Laoire, 2000). More recently, the literature has turned its attention to the role of attendant influences. For example, Bjarnason and Thorlindsson (2006) found that whilst economic considerations influenced young people's intention to leave, local identity and national pride were identified as factors that reduced these intentions. Similarly, whilst Alexander (2016) does not dismiss the importance of the ‘economic component of decision making’ they argue that social and cultural proximity are also important considerations for young people deciding where to live or study (p. 187). However, except for limited examples (see Stockdale, 2002), few studies distinguish between groups whose motivating factors vary. This study contends that appreciating that different groups prioritise contrasting considerations is necessary if public policy makers are to address the needs of specific groups in a targeted manner, thus leading to more effective results.

A rare example of a study that distinguishes between the priorities of different groups is a typology developed by Cooke and Petersen (2019) which investigates the dilemmas faced by rural island youth as they make decisions in relation to education, employment and living location. The types of decisions are allocated to groups within two main categories, namely the ‘Stayers’ and ‘Leavers.’ Consistent with the international literature (Alexander, 2016; Ní Laoire, 2000), the findings revealed that there were various and interrelated factors influencing individuals’ decisions including the education/employment opportunities available locally, lifestyle options, family and social ties, and characteristics of the individual, further demonstrating the complex nature of young people's decision-making processes.

Cooke and Petersen (2019) suggest that the challenges facing young people from rural areas will be similar, whether they hail from an island or the mainland, but argue that ‘this dilemma can be exacerbated for those living on islands because of the logistical and psychological complexity of the “stay or leave” decision’ (p. 101). When investigating the decisions of Welsh speakers in relation to migration, we posit that given the cultural and economic significance of the Welsh language, the language may be a potential complicator for its speakers too.

Academic studies investigating the Welsh language and migration are rare. One such study, conducted by Wynn (2013), examined the outward migration of young people from the Welsh heartlands. Motivated by research developed by Jones (2010) that found young people born outside Wales were four times more likely to leave the Welsh language heartlands than young people who were born there, Wynn's study focused on the hopes and aspirations of young people from families who had migrated to the Welsh speaking strongholds. Jones (2010, p. 123) suggests that people moving from Welsh language heartlands to England are likely to be doing so for work-related reasons; similarly, Van Langevelde (1993), from a Friesian perspective, argues that economic factors and employment shortages drive emigration trends in marginal areas. However, Wynn (2013) argues that whilst a deficit in economic considerations was an important factor in the young people's decisions, they did not ‘overcome’ cultural reasons. Alternative factors ‘related to the sense of belonging, levels of integration and social opportunities’ were found to play an important role in the young people's decisions, and thus economic factors were not necessarily a means of stemming the immigration of young people from Welsh-speaking communities (2013, p. 259).

The study discovered differing relationships with the Welsh language between young people from migrant families and those from families born in the area. Whilst the latter prioritised language preservation, the former viewed the language as integral to integration and belonging (Wynn, 2013). This suggests that lacking these elements could influence their decision to migrate. The difference between the two cohorts reinforces the need to give equal academic attention to the role of Welsh in the migration decisions of Welsh speakers whose families were born in the area in addition to those from migrant families. There is a gap in our understanding of what drives young Welsh-speaking adults to stay, leave or return to areas with a higher density of Welsh speakers, and the role that the minority language plays in their decision-making processes. Consequently, a secondary aim of this study is to further our understanding of the role of Welsh in the migration decisions of its speakers by adapting Cooke and Petersen’s typology for a context in which a minority language is spoken.

Theoretical foundation

Put simply, ‘language planning is an attempt by someone to modify the linguistic behaviour of some community for some reason’ (Kaplan & Baldauf, 1997, p. 3). Early literature explored language planning as an activity for governments at the national level in response to decolonisation processes (Kaplan & Baldauf, 1997; Ricento, 2005). In this respect, it was viewed as a deliberate act on a large scale, from the top down, motivated by the belief that linguistic problems could be solved by planning for the future (Rubin & Jernudd, 1971; Ricento, 2005; Zhao & Baldauf, 2012). References to the role of actors were very general (Kaplan & Baldauf, 1997; Zhao, 2011) and confined to governmental and bureaucratic domains (Cooper, 1989). But as the field of language planning evolved with the recognition that it could be an activity operating on a continuum of levels – from macro, meso, to micro (Kaplan & Baldauf, 1997) - the question of who are the actors and who has agency has received increasing attention from scholars (see Zhao, 2011; Gruffydd et al., 2023; Hodges & Prys, 2019; Finardi & Guimarães, 2021; Vennela & Kandharaja, 2021).

In a general sense, agency can be defined as the ‘intention or capability of an individual to act, initiate, self-regulate or make differences or changes to their situation’ (Liddicoat & Taylor-Leech, 2021, p. 1) and in this sense it can be argued that individuals implementing migration decisions use their agency. Within the context of language planning, Zhao and Baldauf allocate individual actors to four categories, namely (i) People with power, (ii) People with expertise, (iii) People with influence, and (iv) People with interest. The latter group is most relevant to this study as these are individuals who do not possess power or prestige like the first three groups, but rather, are ordinary people that can be defined as follows:

These people are ordinary citizens at grassroots level, who passively or unconsciously get involved in making a decision on language use for themselves, often accidentally, partially because of their occupations, sometimes ‘bumping into’ LPP. Their below-the-radar participation seems to be extremely intermittent and ad hoc, and is limited to rare occasions, but their individual attitudes toward language use taken as a collective can affect societal language behaviour in a significant way. (Zhao & Baldauf, 2008; as cited in Zhao & Baldauf, 2012, p. 910)

It could be inferred that anyone who speaks more than one language is a person with interest as their decision to speak the language, even if ‘passive’ or ‘unconscious,’ in conjunction with the decisions of other speakers, is likely to influence the community use of the languages in question. We suggest that Welsh speakers implementing migration decisions are people with interest as their decision to stay, leave or return, is likely to influence the opportunities available for them to speak Welsh personally and together with the decisions of others, has the potential to influence language use within the area sending or receiving migrants.

Liddicoat and Taylor-Leech (2021), however, interpret the concept of people with interest in a more limited manner:

Those who get involved in LPP decision-making at the grassroots level because of their interest in language issues. (p. 7)

which suggests that the decision made by speakers to use a language (or not) is directly related to issues concerning the language, rather than any influence on language use as a consequence of another decision, such as to stay or leave. This interpretation is reinforced by Zhao and Baldauf (2008, as cited in Zhao, 2011) and offers examples of people with interest as business owners deciding on what language to use when providing services (Kaplan et al., 1995; Marriott, 1991; as cited in Zhao, 2011). Referring back to Liddicoat and Taylor-Leech's interpretation, it could be argued that ‘get involved […] because of their interest in language issues’ implies active rather than passive involvement and in a manner that is conscious rather than unconscious of their actions. We suggest that this limits the original interpretation as it excludes individuals making decisions over language use for themselves in a ‘passive’ or ‘unconscious’ manner (Zhao & Baldauf, 2008; as cited in Zhao, 2011, p. 910).

Other studies using the concept of people with interest have confined studies of actors’ agency to educational contexts (see Fenton-Smith & Gurney, 2016; Finardi & Guimarães, 2021; Ning & Zheng, 2022; Vennela & Kandharaja, 2021). Thus, despite the breadth of the original definition, the way in which the concept of people with interest has been interpreted or applied within other academic contexts seems more confined.

The final aim of this paper will be to investigate to what extent it can be argued that all Welsh speakers, or any minority language speaker for that matter, who implement migration decisions, are people with interest – even if their decisions, although perhaps influencing language use, are not necessarily related to language. In doing so, we use a language planning lens to add to our understanding of individual agency within a different context – that of migration.

Method

Weber (1920/1947, p. 88) describes sociology as ‘a science which attempts the interpretive understanding of social action in order to arrive at a causal explanation of its course and effects.’ This study adopts this perspective to understand migration decisions through the eyes of participants, in an attempt to define and explain motivating factors.

Study sample

A purposive criterion approach (Bryman, 2012) was adopted for this study. Criteria were determined based on the Welsh Government and Arfor's aspiration to retain young Welsh-speaking adults within areas with a higher density of Welsh speakers. As such, 60 Welsh speakers, aged 18–40, who were raised in the four counties with the highest percentage of Welsh speakers, namely Ceredigion, Gwynedd, Carmarthenshire, and Anglesey (15 from each county) (Statscymru, n.d) were recruited to take part in the study. Participants were recruited with the assistance of organisations working with Welsh speakers, with further participants identified through snowball sampling (Bryman, 2012). Consent was obtained from all participants. The sample included a range of ages and included a relative balance of males (27) and females (33). Representation was sought from participants who had decided to stay (12), leave (25) or return (23) to the research area.

Research area

This study focuses on the Welsh-speaking heartlands, delimited to the four Welsh counties with the highest proportions of Welsh speakers, namely Ceredigion, Gwynedd, Carmarthenshire, and Anglesey. Following De Haas (2021, p. 31), who posits staying and leaving as ‘complementary manifestations of the same migratory agency,’ we are interested in the motivating factors leading to the decision not only to leave, but also to stay, and return as these outcomes are of equal significance when considering language maintenance within the heartlands.

‘Staying,’ ‘leaving’ and ‘returning’ are defined as staying within, leaving, or returning to the research area as defined above. For example, if someone originally came from Anglesey and chose to study in Ceredigion, the study considers them to have ‘stayed’ as they continue to live within the research area, even though they have moved within this area. Similarly, if someone hails from Anglesey, and returns after a period in England to Gwynedd, the study considers them to have ‘returned’ – despite the fact that they have not returned to the county of their upbringing. This definition is logical given the ambition of the Welsh Government (2017) and other bodies to retain or attract Welsh speakers to return to Welsh speaking areas. However, it must be acknowledged that this approach obscures some interesting circumstances relating to migration within local areas and at a community level, which are discussed in more detail in the results section below.

It is worth noting too that people who leave the research area are not necessarily crossing borders; despite leaving the language heartlands they may remain within Wales where opportunities to use the language may still avail. Others will have moved to England in-keeping with a long tradition of intermigration between the two countries (Giggs & Pattie, 1992) or elsewhere within the UK. As Castles and Miller explain, some migrants ‘are virtually indistinguishable from the receiving population’ or are ‘quickly assimilated’ (1993, pp. 13 and 14) which may argubly be the case in these instances. Others may have moved further afield of course, perhaps adding to a growing trend of transnational and multilingual families (Hua & Wei, 2016, p. 655).

As well as being the four counties with the highest concentration of Welsh speakers, another characteristic of the research area is that it is to a greater extent rural. As such, trends in rural outward migration and counter-urbanisation observed globally can be applied within this context (see Champion, 1999; Milbourne, 2007). Outward migration, including an exodus of young people for economic reasons, can be said to define the best part of the nineteenth and twentieth centuries in rural Wales (Milbourne, 2011). However a notable shift in migration patterns towards the end of the last century are in line with counter urbanisation trends, with an influx of English migrants retiring to rural Wales (Milbourne, 2011). Despite this, depopulation is still a defining feature of some rural communities in Wales, particularly in less accessible areas such as Ceredigion (Woods et al., 2021).

Research process

Semi-structured, interviews were conducted online and in-person, lasting between 30 and 105 min. The aim of the research was not only to identify the factors considered when reaching a decision to stay, leave or return but also to collect a wealth of data explaining the process of deliberation and compromise needed to reach a decision through the eyes of the participants – in line with the belief that the social world should be interpreted from the perspective of the people being studied (Bryman, 2012, p. 399). Participants were asked to trace their history since leaving school and detail what drove significant decisions in terms of their living location. Questions were also asked relating to their relationship with the Welsh language. The intention was to gather a range of rich and in-depth information rather than attempting to categorise the individuals in a mathematical manner or quantify the sample (Bryman, 2012, p. 405).

Analysing the data

The audio recordings were transcribed, the scripts inputted into Nvivo, and the software was used to assist with the analysis and interpretation of the data by encouraging an explicit and reflective process (Bryman, 2012). The case attributes were created, namely the ‘Stayers,’ ‘Leavers’ and ‘Returners’ categories, and the individual participants were allocated to these groups. As one of the fundamental aims of this study was to test the applicability of Cooke and Petersen’s (2019) typology for a context in which a minority language is spoken, that is, to assess what was similar and what was different, we adopted both an inductive and deductive approach (Bryman, 2012). In this regard, the ‘sets’ feature was employed to create typology groups, and ‘codes’ to create themes a priori – based on the findings of the original researchers, following a deductive approach. In parallel, new themes (codes) based on influential factors in the individuals’ migration decisions were recognised and typology groups (sets) were created in an inductive manner. As such, this typology builds on the earlier typology developed by Cooke and Petersen (2019) where the original sample included 30 young adults and 81 older adults. Cooke and Petersen (2019) conceptualised the first and third-person accounts of these 111 individuals to create the original typology. Where relevant to the first-hand experiences of the 60 individuals in this study, many of Cooke and Petersen's typology groups have been incorporated or adapted within our own typology. The new typology groups are therefore grounded in both published literature and empirical data.

Using the ‘memo’ feature summaries were written for each individual based on their experiences and placed within a specific typology group. In this way, a thematic analysis approach ensured that salient themes received attention and consideration but in a way that did not separate them from the narrative and the wider context of the individual's experience, a risk that can be associated with the use of software such as NVivo (Bryman, 2012). The next step was to compare and contrast the abstracts within each typology group, and then compare and contrast the typology groups. The aim here was to refine the descriptions and characteristics of the various groups to ensure clarity. A table was created to present the different typology groups, a description of each, and their most prominent influential factors (Table 1).

Table 1. Typology of minority language speakers’ migration decisions.

Category	Description	Dominant factor(s)/ issue(s)
The ‘Stayers’ among young adults who speak a minority language
i. The Rooted	Are rooted to the community by choice. Choosing current location due to their social networks, and community use of the minority language.	Social characteristics (positive social connections); Cultural characteristics (e.g. minority language).
ii. The Faithful	Rooted within the community out of duty accompanied by a sense of belonging and/or commitment to a way of life.	Social and/or cultural characteristics, but acting as a tether.
iii. The Disengaged	Unsure about what to do, and that means staying. Lacking drive and direction to acquire more skills and experience.	Individual characteristics (e.g. apathy, low self-esteem).
iv. The Strategic	The ones who want to acquire valued skills and see the value in doing so close to home, and who are willing to stay if good local programmes are available.	Community characteristics (attractiveness of local education/employment options); Individual characteristics (e.g. discipline and foresight).
v. The Stymied	The ones wanting to stay but perceive there to be no employment or suitable housing. Thus, are flight risks.	Positive social and cultural characteristics, but concern for employment and housing.
vi. The Uncertain	Excited by options available elsewhere (employment and/or lifestyle) but lack the confidence to leave, and therefore make the most of what is available locally.	Individual characteristics (e.g. lack of confidence, uncertainty). Recognise that moving involves pros and cons.
The ‘Leavers’ among young adults who speak a minority language
vii. The Cosmopolitan	Attracted by city life.	Individual characteristics (prefers or craves an urban lifestyle).
Viii. The Urbanites	Feel the community is too small or feel like they’re attracting unwanted attention.	Community characteristics (aspire to the anonymity of the city and/or a more inclusive community).
ix. The Reluctants	They want to stay locally, but suitable employment options are lacking locally for themselves and/or their partner.	Community characteristics (perceived unattractiveness &/or unsuitability of employment conditions trump positive social connections).
x. The Ramblers	Keen to leave for new adventures, and therefore leave for work and/or education. Comfortable with change.	Individual characteristics (i.e. inquisitive, comfortable with risk). Seeking new adventures in new places.
xi. The Academic Achievers	Select reputable education programmes or notable institutions for a subject.	Prioritises educational goals.
xii. The Careerists	Goes to the best job opportunity, regardless of the location. Driven by career goals.	Community characteristics (relative strength of local labour market. Willing to relocate to the one perceived to be the best, wherever that might be).
xiii. The Minority Language Cosmopolitans	Want to enjoy city life through a minority language e.g. Welsh	Community characteristics (craving or preferring a city lifestyle together with a perception that there are more attractive and/or suitable employment opportunities); Cultural characteristics (wanting to live life through a minority language).
xiv. The Lonely	Feels lonely without peers or a partner.	Social characteristics (too few peers with common interests or difficulty finding a partner).
xv. The Transplanted	Originally left for education and/or work reasons, but have now developed stronger social networks in their new location, whilst networks have weakened back in the places they were raised.	Social characteristics (strengthened away while weakened back home).
xvi. The Happy Migrants	Having moved to another country the perception of a higher standard of living keeps them from moving back.	Community characteristics (perception of a higher standard of living away).
xvii. The Influenced	Those who follow the crowd to university without really considering what other options may be available to them locally.	Individual characteristics (influence of peers and ‘the norm).
The ‘Returners’ among young adults who speak a minority language
xviii. The Vulnerable	Return to a safe place because they, or their circumstances, are vulnerable.	Social characteristics (family providing safety net).
xix. The Pragmatic	Didn't intend to return but unable to turn down a suitable and/or attractive job offer.	Community characteristics (competitive local labour market).
xx. The Family-orientated returnees	Have returned to raise a family through the minority language.	Cultural characteristics (use of the minority language within the community and provision of education through the minority language); Social characteristics (close to family to help with caring responsibilities); Community features (access to the outdoors).
xxi. The Opportunists	Have returned to take advantage of an opportunity or offer that allows them to create a lifestyle.	Environmental characteristics (inheriting property or taking advantage of an opportunity), Personal characteristics (enterprising).
xxii. The Nostalgic	Moving back in response to a sense of hiraeth^a (nostalgia) (associated with belonging to a particular culture, language, landscape and community).	Social and cultural characteristics (hiraeth, and a sense of belonging).
xxiii. The Connection-keepers	Those who use their childhood home as a base but leave to take advantage of education and/or work opportunities not available locally.	Social characteristics (strong social ties), Community characteristics (lack of job and education opportunities locally).

^a Hiraeth is a Welsh work which roughly translates as a longing for home. It has no direct translation in English but exists in other languages e.g. ‘heimat’ in German, and ‘saudade’ in Portuguese. (Harris, 2012, p. 1).

Limitations

One limitation of this paper is that due to space constraints, we were unable to provide a deeper analysis of demographic differences such as age, gender and geographic location. We share Cooke and Petersen's hope that future studies will focus on particular demographic groups (as this paper does with regard to minority language speakers). In a similar vein, it should be noted that we, like Cooke and Petersen (2019), view this typology as ‘a starting, not a finishing, point’ (2019, p. 118). It is our hope that it will be embraced by researchers and policy makers and utilised to meet their own needs.

Results

The typology presented below categorises young Welsh-speaking adults by the dominant factors or issues which drive their decisions in relation to migration. The typology groups are placed within the sections of ‘Stayers,’ ‘Leavers’ and ‘Returners’ to the study area. As already explained, this typology is based on recent work by Cooke and Petersen (2019). It is not possible to discuss all typology groups due to the confines of this article, and therefore attention is given to the typology groups most relevant to addressing the research aims of this study. Pseudonyms are used to protect the identity of contributors.

It became clear when analysing the data that some participants’ summaries corresponded with Cooke and Petersen’s (2019) typology groups and therefore the descriptions of these typology groups were adopted, virtually unchanged, within the typology of this study. Examples of these categories include ‘The Disengaged,’ ‘The Stymied’ and ‘The Ramblers.’ This shows that there is a similarity between factors driving the migration decisions of young people from rural islands and those from areas in which a minority language is spoken. This may be due to similar socio-economic characteristics of the two areas as well as common personal characteristics or trends associated with young adults.

In other cases, where the overall essence of the group's description rang true, but some characteristics varied, the typology group was adapted for the context of this study. ‘The Socially Contented’ within Cooke and Petersen's (2019) typology, for example, become ‘The Rooted’ within the typology of this study.

Cooke and Petersen (2019) describe ‘The Socially Contented’ as follows:

Are rooted to the community by choice. Choosing current location due to their social network, and willing to accept available training and work options. (Cooke & Petersen, 2019, p. 112)

As the name suggests, ‘The Rooted’ (like ‘The Socially Contented) are also embedded within the community due to social connections, however for this cohort, this feeling of ‘rootedness’ goes hand in hand with the aspiration to live within a Welsh-speaking community. Living in an area where the minority language is spoken at a community level is integral to this group's decision to ‘Stay.’ In Cadi's words:

Speaking Welsh was something really important to me. And one of the reasons I wanted to stay in Wales. Doing it in a social way is important to me, I don't like to speak English at all, it doesn't flow as easily as Welsh for me, I don't think. So the opportunities are important to me, just going to the pub and speaking Welsh, I like to be able to do that. (Cadi). [Author's translation.]

Furthermore, whilst the aspect of having to compromise on training and/or work opportunities applied to some individuals within this category, it did not define this group as in the case of ‘The Socially Contented.’ Indeed, some individuals within this typology group had studied Welsh at university and due to this skill secured employment within a language-related position, and considered themselves fortunate of this:

So yes, obviously, I think subconsciously I’ve considered what opportunities there are (locally) but also I’ve been terribly lucky that it hasn't affected you know, the fact that I want to live in Anglesey, that I want to work in Anglesey, have a job with a decent salary in Anglesey, and that didn't affect the kind of degree I wanted to do and the kinds of things I like. (Llinos). [Author's translation.]

Llinos has not been forced to compromise on her education or career aspirations in order to stay, and more than this, her Welsh language skills have been an asset to her in securing comparably good quality work in the area. This perhaps lends support to the importance of policies that offer sponsorship to students who study part of their course through the medium of Welsh and strategic investment in language-related sectors.

As already explained, the typology of Cooke and Petersen (2019) is divided into two categories, ‘Stayers’ and ‘Leavers.’ This study builds on the original typology by introducing a third category, entitled ‘Returners.’ Bearing in mind the Welsh Government's strategic aim to attract Welsh speakers back to the language heartlands, it is argued that understanding the factors that provide a catalyst for return is a vital and significant addition to our understanding of the field.

An important group within this category, from a minority language perspective, are the ‘Family-orientated returnees.’ The Welsh language, along with other considerations such as being close to family to give or receive support and to take advantage of outdoor activities, is a key factor in their decision-making. Liam explains:

… I was actively looking at the Estyn schools’ reports, I was actually looking at how many children came from homes where Welsh is spoken as a first language and the percentages […] I felt pretty strongly about moving to a more Welsh area for the schools – and that's what we did. (Liam). [Author's translation.]

Liam's decision to return to north-west Wales was due to having family in the area and that he had secured employment locally, but he was also keen to ensure that his children attended a school where Welsh was spoken not only in the classroom but in the playground too. Consequently, he researched the linguistic make-up of the area's primary schools by scrutinising Estyn reports (an education and training inspection body in Wales) and this was one of the main reasons he decided to move to a particular area of the county – so that his children could attend a school where he presumed Welsh would be spoken at break-times and not restricted to the classroom. This is a clear example of Liam using his agency to increase the opportunities for his children to use Welsh daily and an indication of the significance of Welsh medium immersion education and policies aimed at increasing the social use of Welsh.

We conclude that although there are similarities between the context of the rural islands and that of this study, the minority language (in this case Welsh) can be a key factor in the migration decisions of its speakers. In this respect, and in contrast to Cooke and Petersen's findings, cultural factors were found to be an important consideration within a migration context where a minority language is spoken.

Although the influence of the Welsh language on the decisions of the individuals represented by some of the typology groups was clear, the majority did not prioritise the Welsh language when reaching a migration decision. As per Cooke and Petersen’s (2019) study, the data showed that a combination of social (e.g. relating to proximity of family and friends), and environmental factors (e.g. relating to educational/employment opportunities or provision of activities/services), along with personal characteristics (e.g. individuals’ values, aspirations and qualities) also played a role. This reflects the complexity and multi-layered nature of the decision-making process, and the data shows that Welsh does not always take precedence in the decisions of its speakers.

Given the emphasis afforded by the Welsh Government and the Arfor programme to the role of employment in attracting young Welsh speakers to stay or return, special attention was given in this study to the influence of factors relating to employment upon the decisions of young Welsh-speaking adults in relation to migration. The decisions of ‘The Strategic,’ ‘The Stymied,’ ‘The Reluctant, ‘The Careerists,’ ‘The Pragmatic,’ ‘The Opportunists’ and ‘The Connection-keepers’ are mainly based on factors relating to employment or the economy. Of these 7 groups, The Careerists are the group least likely to be tempted to stay in the research area as they prioritise their career over all else, and often choose a career based on specialist interests, where there are not necessarily relevant opportunities available to them locally. As such, it is suggested that limited resources may be more effective if directed towards those who want to or can be persuaded to, stay or return.

Of the remaining groups, (‘The Strategic,’ ‘The Stymied, ‘The Reluctant,’ ‘The Pragmatic,’ ‘The Opportunists’ and ‘The Connection-keepers’), all but one group wish or prefer to stay or return – providing there are suitable employment opportunities available to them locally.

A feature of ‘The Strategic,’ for example, is that they use their foresight to consider what skills are needed to develop a successful career locally. It is therefore interesting to note that the participants within this typology group are apprenticeship holders and see this as an effective means of realising this goal. Aled shares his experience:

I’ve started a degree apprenticeship with [company name]. I signed a contract for 5 years and I need to go to Uni in [the name of an English university]. The University side is completely up to me, but they pay my fees and pay me to go there. In the end, because they’ve paid for everything, I need to.. I’ve signed to do 2 years of work for them. Yeah, in terms of them paying my university fees. So I won't be in debt when I finish. (Aled). [Author's translation.]

While Aled admits this to be a significant commitment, he also considers his situation to be ‘7 years of guaranteed work and income.’

This indicates that access to an apprenticeship programme can attract young Welsh-speaking adults to stay, thus strengthening the case for sustained investment in the sector.

However, whilst there were opportunities for study participants to use Welsh with co-workers within the workplace, they did not necessarily follow the taught element of their apprenticeship programme through the medium of Welsh. This demonstrates the importance of ensuring opportunities and encouragement for young people to continue to develop bilingual skills after leaving statutory education – a view supported by the Welsh Government who see it as an important step towards increasing the number of Welsh speakers (2017, pp. 32, 40).

The ‘Stymied’ and ‘The Reluctant’ represent two sides of the same coin. Both groups were keen to stay, however, the first group was concerned with having to leave due to a perceived lack of job opportunities and the latter had already left due to the inability to find suitable employment. Interestingly, for some participants within ‘The Reluctant’ group, it was not the participants themselves but their partners who were unable to find suitable employment – leading to the family unit moving away from the research area. In one case, the participant believed that her partner's inability to speak Welsh contributed to his negative experiences when looking for work. In Ceri's words:

He couldn't get a job. He was an attorney. So, he had like, he was very well qualified, but he, he applied for a lot of jobs at the university and in town. And I think he was either like over-qualified … The fact that he didn't speak Welsh was a hindrance. And the fact that he wasn't from Wales or from [town name] ‘cause [town name] can be quite close-minded at times. And I don't think that helped but yes, he couldn't get a job the whole like three years that we were there. And so then that ultimately led to the decision to move back to [location abroad] so that he could work […]. (Ceri)

This finding is supported by participants from other typology groups who cited their partners’ deficiency in the minority language as a factor relating to their inability to find suitable employment. This supports the finding that Welsh speakers are less likely to be unemployed, particularly in the Westerly counties, perhaps due to legislation aimed at strengthening the Welsh language (Drinkwater & O’Leary, 1997). It therefore seems possible that there are unintended consequences to legislation aimed at empowering the Welsh language where young families who have the potential of passing on the language to their children are lost, perhaps offering an example of unplanned language planning (Egginton, 2002). Conversely, if these families decide to stay but their partners and children do not acquire Welsh, then the language would be adversely influenced in this situation, as Austin and Sallabank (2011) and Ceallaigh (2020) warn. This demonstrates the importance of schemes such as Welsh for Adults. It is interesting to note that the families involved in this study migrated not only from the research area but from Wales and indeed from Europe, settling on entirely new continents. Their use of Welsh is now limited to occasional conversations with family over the internet and therefore shows the consequences of migration decisions on the opportunities to speak Welsh, whilst also demonstrating the role technology can play in processes of language maintenance within families (Åsa Palviainen, 2020, p. 104).

Unlike the groups already discussed ‘The Pragmatic’ do not necessarily prefer or desire to live locally, but rather feel unable to turn down attractive employment within the research area. In contrast to what they envisioned, an employment opportunity has arisen locally, and they have moved back to take advantage of it. Returning to the research area has increased the opportunities for them to use the language and in more than one case, where Welsh was not the language spoken at home, it is now the language of the office and thus the opportunities for these individuals to use the language have increased significantly. This supports the findings of previous studies demonstrating the importance of the workplace as a domain to use Welsh as well as stimulating muda (plural mudes), or ‘specific biographical junctures’ (Pujolar & Puigdevall, 2015) where new speakers change their linguistic habits (Hodges, 2024; Morgan, 2000; Tilley, 2020). Some of the job opportunities described by the participants within this category were related to macro linguistic policies that have led to the funding of Welsh medium jobs within the education sector. This suggests that such language planning initiatives are effective in terms of maintaining Welsh speakers within Welsh-speaking communities. However, an important consideration from this group's perspective is the need to establish career paths, or there is a risk of losing the speakers once again. As Sera explains:

Right now I’m happy. There are other colleagues who have been in the job for 10–12 years now and I don't see myself staying that long. I want new things, a more sort of managerial role – stepping up a little bit. (Sera). [Author's translation.]

While there is evidence that employment can be a means of attracting Welsh speakers back to the minority language heartlands, there are deal-breakers such as a lack of career paths that can lead to outward migration in the future. Some researchers posit limited opportunities to progress as a hallmark of a ‘bad job’ (Kalleberg et al., 2000; McGovern et al., 2004) and possibly a trend more prevalent within smaller companies (Olsen et al., 2010), such as those found in marginal areas where minority languages are often spoken.

Returning to a point made in the methodology section, namely that the way ‘stay,’ ‘leave’ and ‘return’ are defined within this study masks some interesting trends in relation to mobility within smaller geographic areas. For example, there were participants from agricultural families keen to stay or return to the very farm or ‘square mile’ where they grew up and were frustrated at the housing options available to them. In contrast, there were several participants who wanted to return to the region, but not to the specific village of their childhood home. To offer an example, whilst Lowri would like to have moved back to her hometown, she perceived it to be ‘too rural’ for her partner who grew up in an English city and didn't speak Welsh, and that the larger nearby town was a fair compromise. Indeed, Caernarfon and Aberystwyth seemed to be attractive locations for Welsh speakers as they offer a relatively good variety and standard of facilities as well as an opportunity to speak Welsh within the community. It is therefore suggested, in passing, that ‘inter-rural’ migration is an area of research that requires more attention, as suggested by Stockdale (2002), as the trend has the potential to significantly influence the use of minority language within specific communities.

Discussion

The practical contribution of this study is to define and interpret motivating factors to explain Welsh speakers’ migration decisions in order to help policy makers support minority language speakers to stay or return to areas where the minority language is spoken as a community language. Whilst there is evidence that young Welsh speakers leave the language's heartlands for economic reasons (Aitchison & Carter, 1987), others assert that there is a dearth of evidence to support the perception that interventions aimed at creating employment within these areas are an effective means of retaining or attracting back Welsh speakers (Bonner et al., 2024), especially given that other factors are also involved in individuals’ decision-making processes (Wynn, 2013). However, the empirical findings of this study show that matters pertaining to employment are a priority for some, and in this respect are a means of maintaining Welsh speakers within areas where the language is spoken as a community language (e.g. The Strategic), or motivated to return to these areas (e.g. The Pragmatic). Indeed, examples of language and economic policies achieving this feat are cited by participants of this study. A key policy priority should therefore be to invest in the economy of areas with a higher density of minority language speakers to create attractive employment and thus sustain speakers within these areas. However, simply creating employment of any nature is unlikely to address the specific needs of the various typology groups. We therefore advocate a suite of initiatives targeting the needs of the distinct typology categories.

For example, in line with other studies (e.g. Magnussen, 1995), the importance of language-related industries is highlighted, specifically in relation to creating competitively paid jobs within a field relevant to individuals’ expertise. The sector enjoys a good reputation amongst young adults for interesting opportunities and progression routes, within a Welsh-speaking environment. We suggest that other sectors, for example, green energy or social enterprise, should aim to develop a similar reputation. However, when it comes to employment, what emerges from the data is that there are potential deal-breakers along the way, such as a lack of career development pathways. This situation is recognised by Woods et al. who explain that in-house training, up-skilling and succession planning are often restricted due to the comparatively small scale of rural businesses (2021, p. 75). They advocate a number of strategic interventions including positioning self-employment as an aspirational career path, and improving opportunities for sole traders, microenterprises and small organisations to access apprenticeship schemes (2021, pp. 75 and 76). It is not a question therefore of creating jobs of any kind within a minority language stronghold, there must be upskilling opportunities and progression routes, as well as opportunities to use the minority language within the workplace.

There are also examples of where a perceived or actual shortage of suitable employment opportunities has led to outward migration (‘The Stymied’ and ‘The Reluctants’), a finding consistent with the international literature (Cooke & Petersen, 2019; Stockdale, 2002; Thissen et al., 2010). It is suggested that ‘suitable’ employment refers to opportunities that are aligned with the educational qualifications of young adults. Whilst it may be possible to diversify the rural economy to a certain extent, it is unlikely to be able to compete with the breadth of opportunities available in more urban environments. It is important therefore to encourage young people to gain qualifications and experience in areas where they are more likely to find good quality employment in the heartlands. In this respect, we concur with Woods and Utz (2022, p. 61) that more should be done to bring together employers and education providers to highlight pathways to the employment opportunities available locally.

Overall, the evidence accessed as part of this study validates the rationale supporting the Welsh Government's language policy and schemes such as Arfor, that is, that employment has a role to play in maintaining Welsh speakers within Welsh-speaking communities, or in attracting them back to these areas. However, as we have argued previously, simply creating jobs of any kind is unlikely to be successful in encouraging young Welsh-speakers to stay or return to the heartlands. The typology presented here advocates a more nuanced and targeted response which is not restricted to language planning but includes socio-economic policy implications targeting the diverse needs of the individual groups. The typology demonstrates empirically that language vitality is deeply intertwined with socio-economic factors and as such policy makers must adopt a holistic approach to include a wide range of targeted linguistic, economic and social policies similar to those outlined above and aimed at ensuring that Welsh-speakers prosper in the heartlands.

A secondary aim of this research was to adapt Cooke and Petersen’s (2019) typology for a context in which a minority language is spoken with the aim of furthering our understanding of the role of that language in the decisions of its speakers. It was clear that Welsh was a very important influence on the migration decisions of some typology groups, namely ‘The Rooted’, ‘The Cosmopolitan minority language speakers’ and ‘The Family-orientated returnees.’ However, whilst Wynn (2013) acknowledges that language should not be dismissed as a factor in young people's aspirations, they found that it ‘did not specifically influence young people's choices and that language was not of significant importance to them’ (p. 249) [Author's translation]. In contrast to Wynn's conclusion, the data collected as part of this study shows the Welsh language to be a significant influence on the migration decisions of at least some of its speakers. Furthermore, even where factors relating to the Welsh language were not a priority in the individual's migration decisions, there were instances where factors relating to the relationship between economy and language were significant to the success of the individual or the viability of the linguistic community. For example, some found the ability to speak Welsh to be an advantage when looking for work, and it was found that the workplace provided an important space to use the language. As such, considerations relating to the Welsh language apply to all typology groups, whether they prioritise factors relating to the language or not.

The final objective of this work was to consider to what extent Welsh speakers can be regarded as people with an interest in order to add to our understanding of the individual agency in relation to migration through a language planning lens. It can be argued that the individuals represented by ‘The Rooted,’ ‘The Cosmopolitan minority language speakers’ and ‘The Family-orientated returnees’ are actors who use their agency in favour of the Welsh language. The first two groups use their agency to live in areas where they can use Welsh daily within the community, and the latter group choose locations that are likely to facilitate the transmission of Welsh to their children. They purposefully and consciously make use of their agency in relation to the use of Welsh at a micro level.

In line with the recommendation made by Hodges and Prys (2019), it is suggested that policymakers should pay attention to the use of such agencies. The impact of the outward migration of Welsh speakers from the rural West to the capital city on the linguistic regions involved in well known (Aitchison & Carter, 1987), however, the findings of this study highlight implications at a more local level. There are examples of parents targeting specific areas where there is a perception that Welsh is the language of the schoolyard (rather than being confined to the classroom) and of individuals targeting specific areas because it is their perception that Welsh is a community language within these areas. However, could further empowering the Welsh language within areas where it is perceived to be strongest forsake the language in other areas? It is suggested that migration patterns need to be investigated further at a more local level, giving particular attention to the use of individual agencies at a micro level.

Zhao and Baldauf's definition of people with interest includes those who make decisions about their personal use of language in a passive or unconscious manner, and that these decisions, in conjunction with the decision of other speakers, are likely to influence the community use of the target language (Zhao & Baldauf, 2008; as cited in Zhao & Baldauf, 2012, p. 910). This position is supported by the data, for example, in how participants increased their use of the language after returning to take up a job where Welsh was used in the workplace, or in the way that participants who migrated away lament the limited opportunities for them to speak the language in their daily lives since leaving the Welsh-speaking heartlands. Although these people do not make migration decisions based on the minority language, it can be argued that the impact of their decisions can be found on the language. It is therefore proposed that, in addition to being people with interest, minority language speakers should be seen as people of interest – to linguistic planners, at least. Policy makers concerned with minority languages must respond to the needs and aspirations of minority language speakers as their migration decisions influence not only the opportunities individuals have to use the minority language but also the potential they have to contribute towards the viability of the language within the communities in which they live. This paper provides empirically grounded examples of how the typology can be used to identify distinct policy interventions aimed at addressing the needs of individual typology groups. It is not likely possible, or indeed desirable, to maintain or attract back all Welsh speakers to the research area, and so it is suggested that limited resources may be more effectively targeted at those who want to, or can more easily be persuaded to, stay or return. The strength of this typology lies in its ability to identify which groups to target with this goal so that policy makers may tailor a broader range of more nuanced interventions accordingly.

Acknowledgements

The authors would like to thank the research participants for their participation. Diolch i chi gyd. Ethical approval was granted by Bangor University's ethics committee (HLSS202122S041).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This paper is based on a PhD project supported by the Martin Rhisiart scholarship awarded by the Coleg Cymraeg Cenedlaethol.

Notes on contributors

Elen Bonner

Elen Bonner is a PhD student at the School of History, Law and Social Sciences, Bangor University. Her research explores aspects of the relationship between the economy, migration, and the Welsh language.

Cynog Prys

Cynog Prys is a Senior Sociology Lecturer at the School of History, Law, and Social Sciences at Bangor University. Dr Prys’ current research interests include the sociology of language and bilingualism, specialising in the use of Welsh in civil society and the third sector in Wales.

Rhian Hodges

Rhian Hodges is a Sociology and Social Policy Senior Lecturer at the School of History, Law, and Social Sciences at Bangor University. She is currently the Director of the M.A. Language Policy and Planning and has published widely within the fields of Welsh medium education, new Welsh speakers and language planning in Wales.

Siwan Mitchelmore

Siwan Mitchelmore is a Senior Lecturer at Bangor Business School, her research interests include Small and Medium Enterprise business development, ambidextrous leadership, female entrepreneurship, and innovation.

Notes

1 The ‘Bro Gymraeg’ is a term used to describe areas where the Welsh language is at its strongest (Wynn, 2013, p. 46).