This thesis explores the link between domain specificity and the resulting attitudes towards Welsh and English among young speakers in Wales. The domain specificity refers to Welsh and English being compartmentalised for function (i.e. Welsh transmitted through the school domain). In order to investigate this link, this thesis investigates the attitudes of three separate cohorts of school pupils across Wales who receive daily contact with the Welsh language through the school environment. The education system plays a crucial role as a main agent of Welsh language reproduction, transmission and shift throughout Wales, particularly important in both the development of ‘new’ speakers in predominantly Anglophone areas (Williams, 1992) and in providing the language with a niche domain of transmission for ‘traditional’ speakers (May, 2000). Young people have been identified as the crucial demographic in language revitalisation (Fishman, 1991), thus the three studies composing this thesis consider speakers at post-16, mid-adolescence and transition stages. The growth in WME in anglicised areas of Wales has raised hopes for a revival (Williams, 2003:7; Coupland et al. 2005:2) although, as almost two thirds of primary pupils come from English-speaking homes, pupils’ Welsh language use beyond the classroom may be severely impaired (Edwards and Newcome, 2005:303). As education is pertinent to the revival and maintenance of the Welsh language, we must understand the attitudes which exist among speakers – particularly L2 learners (i.e. L2 learners/ ‘new speakers’) as they are a major focus of Government planning efforts to increase speaker numbers through WME. This is the focus of the first study of the thesis, exploring the extent to which compartmentalisation impacts speakers’ attitudes and behavioural output (i.e. Welsh and English viewed as functionally separate). From this, this thesis proposes the resulting diglossic relationship between Welsh and English among these speaker communities (i.e. Welsh transmitted through the school domain impacts use and application of Welsh in the informal communicative domains). While some negative attitudes exist among L2 learners relating to how Welsh is perceived in the school environment, negative attitudes have not been previously reported in L1 cohorts. However, the literature reports that both L1 and L2 cohorts demonstrate decreased use of Welsh in interpersonal social communication. There is mounting evidence which demonstrates that many pupils – both L2 and L1 fluent native speakers – seldom use Welsh beyond the classroom (Jenkins, 2001; Jones and Martin-Jones, 2004; Thomas and Roberts, 2011) with a marked decrease in Welsh use as children become older (Baker, 2003; Edwards and Newcombe, 2005; Hodges, 2009). As existing L1 reseach has measured attitudes overtly, it is possible existing research has not measured attitudes accurately, thus the discrepancy between seemingly ‘positive L1 attitudes’ yet tendency to use English over Welsh in informal contexts as reported in the literature. Thus the second study explores L1 covert attitudes towards Welsh and English. Crucially, if there is a perceived domain specificity between the languages, we would expect notable gender differences in speaker attitudes. Female speakers consistently hold more positive views towards the language of the formal domains (Eckert, 1989, 1990, 1998), whereas male speakers demonstrate less favourability towards overtly prestigious language varieties (Wang and Ladegaard, 2008). As such, investigating language attitudes and gender may provide crucial information regarding how Welsh and English are viewed, providing broader understanding as to why speakers may converge or diverge from certain language varieties depending on domain specificity. The three studies of this thesis differ in their approach and scope – in the research methodologies and the cohorts investigated. The principal research questions of the three studies are thus as follows: Q1: What attitudes exist towards the Welsh language among WM-educated L2 males in an Anglo-context? Q2: What attitudes exist towards Welsh and English among WM-educated L1 adolescents in an L1 Welsh community context? Q3: Which activities can promote Welsh transmission from the formalised school domain into the informal communicative domains? The first study explores the attitudes of Welsh-medium educated males towards Welsh. Methodologically, the study employs the qualitative focus group (FG) method in one Cardiff secondary school. Here, I introduce the concept of ‘inverted diglossia’, and analyse the results with reference to this theoretical approach. The results suggest that there is compartmentalisation of Welsh and English, with participants associating the former with formality and the latter with informality. These results highlight a diglossic relationship in this speaker community insofar as domain specificity reportedly impacts upon speaker attitudes and domain of application i.e. Welsh seen as the language of formality and is not transferrable into informal contexts. Moving forward from this, the second study explores covert attitudes towards Welsh and English among adolescents in an L1 Welsh community school. The methodology comprises a quantitative study as an adaptation of the matched-guise technique (MGT). The results indicate that English is rated more positively than Welsh, particularly evident among male participants as similarly reported in the speech community of the first study. The third study provides a combination of qualitative and quantitative research methods. Through this study, I provide an assessment of attitudes towards intervention approach (i.e. Forest School (FS) through semi-structured interviews with school staff and FGs with pupils). The third study provides design, implementation and measure of an intervention aimed at facilitating Welsh use among speakers in informal communicative contexts. Qualitatively, staff and students perceive the FS component positively as a means of transmitting Welsh into the informal domain. The pre- and post- intervention FG recordings provide the basis for the quantitative assessment. The results demonstrate increase in Welsh use in the post-intervention FG. It is argued that a third domain such as this provides an alternative (informal) domain of transmission for Welsh, thus circumnavigating the diglossic relationship which has developed through existing domain specificity. The results of these studies are discussed with reference to curriculum developments in Wales and minority language maintenance efforts in the international arena.