8.1 Introduction

The title of this chapter should really be ‘Englishes in Wales’, in reference to the regional diversity of English dialects of Wales, resulting from different histories of anglicisation and varying degrees of influence from the Welsh language and English English (EngE) dialects (e.g. Parry Reference Parry1999; Thomas Reference Thomas and Trudgill1984, Reference Thomas and Burchfield1994; Penhallurick Reference Penhallurick1991, Reference Penhallurick1994; Awbery Reference Awbery and Tristram1997; Paulasto Reference Paulasto, Penttilä and Paulasto2009, Reference Paulasto, Durham and Morris2016; Jones Reference Jones2016a, Reference Jones2018). This diversity will constitute a running theme for this chapter. Another such theme is dialect change. By now it is clear that many of the traditional dialect features described in such formative works as the Survey of English Dialects (SED; Orton et al. 1962–Reference Orton, Barry, Halliday, Tilling and Wakelin1971) and, in Wales, the Survey of Anglo-Welsh Dialects (SAWD; Parry Reference Parry1977, Reference Parry1979, Reference Parry1999; Penhallurick Reference Penhallurick1991) have become either extinct or at least severely endangered. Other features, however, continue to be heard in the local vernaculars. The present chapter will therefore comment on the developments emerging in dialect data from SAWD onwards to the present day. A recent book-length description of Welsh English (WelE) (Paulasto, Penhallurick and Jones Reference Paulasto, Penhallurick and Jones2020) looks into the dialectal diversity and change in rather more detail than is possible in a single chapter. Hence the aim here is to obtain a bird’s eye view of the regional dialects and accents of English.

The groundwork for regional variation was laid during the centuries when the English language gradually made its way into Wales. The anglicisation of Wales is described in detail in several sources (e.g. Williams Reference Williams and Coupland1990; Thomas Reference Thomas and Burchfield1994; Paulasto Reference Paulasto2006; Morris Reference Morris2013), hence the present chapter will only draft the broad outlines of the process. Areas which anglicised early on, before or during the Anglo-Norman period, include the Border country, the southeastern coast, the Gower peninsula, and South Pembrokeshire. They differ with respect to EngE dialect influence, as the border and adjoining coastal areas absorbed dialect input from the West Midlands, while the southern pockets, surrounded by the sea on the one hand and Welsh speakers on the other, drew settlers from south-west England. They are nevertheless similar in displaying regionally specific, historical dialect features that largely derive from EngE rather than Welsh language contact. These are also the parts of Wales that first drew the interest of traditional dialectologists in the nineteenth century (Awbery Reference Awbery and Tristram1997; Penhallurick Reference Penhallurick1994).

The anglicisation of these areas was followed by a lengthy period of societal diglossia and regional demarcation of English and Welsh, with most of Wales remaining monoglot Welsh and bilingualism being restricted to towns and the linguistic boundaries. The diglossia was established quite firmly by the dual effect of the Laws in Wales Acts of 1536 and 1543, stripping the Welsh language and its speakers of linguistic rights in public offices, and the Protestant Reformation and subsequent translation of the Bible into Welsh in 1588, affording the language a refuge in religious functions in addition to domestic ones (Williams Reference Williams and Coupland1990:21). Hence the next major wave of anglicisation only began with the industrial period from the late eighteenth century onwards. Industrialisation concerned particularly the newly discovered south-east Welsh coalfield, which up until then had been predominantly Welsh-speaking. The region first became increasingly bilingual, with the workforce flooding in from England as well as rural Wales, until the area ultimately anglicised during the late nineteenth and early twentieth centuries (e.g. Williams Reference Williams and Coupland1990:31–6). Due to the speed and informality of the language shift, the dialect which emerged initially displayed a high degree of Welsh contact input at all levels of language (Ellis Reference Ellis1882; George Reference George1990; Paulasto Reference Paulasto, Penttilä and Paulasto2009), but the subsequent, primarily monoglot English-speaking, generations have moulded the southeast Valleys English into a variety that is quite distinctive: it retains Welsh input mainly in its phonology, some lexis and restricted features of morphosyntax, shares features of EngE dialects spanning across the border, and involves regional dialect features not found elsewhere in Wales. I am here including Cardiff English (CarE) in the broad southeastern dialect area; despite differences, there are also similarities.

The popularity of the English language was furthermore boosted by the 1847 Report of the Commissioners of Inquiry, which judged the socio-economic problems of the Welsh to be the result of speaking Welsh rather than English (Roberts Reference Roberts and Schwarz1996), and the subsequent Education Act of 1870, which made primary education free and compulsory though the medium of English alone. These developments had an impact on all of Wales. As indicated early on by Ellis (Reference Ellis1882) and followed up by Paulasto (Reference Paulasto, Penttilä and Paulasto2009), the anglicisation of the rural Welsh-speaking north and southwest is to a great extent – but not solely – the result of the educational system. Additional factors are English increasingly becoming a community language and the many twentieth-century developments, such as mass media, population mobility, and counter-urbanisation. Each of these factors has had varying degrees of impact on how dialects of English have developed and evolved in the rural southwest and north of Wales.

The datasets used for the present chapter include my sociolinguistic interview corpora from Llandybie in Southwest Wales (LC, recorded 1995–2000), North Wales (NWC, 2000), and Tonypandy (TC, 2012), and, to a lesser extent, transcribed sections of the Survey of Anglo-Welsh Dialects interviews (SAWD, 1974–1981) and Rob Penhallurick’s so-called Urban SAWD interviews from four Welsh cities, compiled in 1986. References will also be made to Ceri George’s interviews from the Rhondda (1979; see George Reference George1990). Significant recent data arise from the BBC Voices in Wales survey (BBC Voices 2004–2005) and the Millennium Memory Bank project (MMB, 1998–1999; see British Library Sounds 2009). For further information on the corpora and surveys, see Paulasto (Reference Paulasto2006:140–56) and Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:36–42, 45–7).

8.2 Phonology

The phonology chapter of the book Welsh English by Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:45–69) was authored by Rob Penhallurick. He compares the SAWD data (Parry Reference Parry1999), which most broad descriptions of WelE have thus far been based on (e.g. Penhallurick Reference Penhallurick and Britain2007, Reference Penhallurick, Kortmann and Upton2008), against the more recent data provided by the MMB and BBC Voices surveys and other recent research, producing an updated overview of the present-day phonology of the variety. Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020), as well as the present chapter, avail of the linguistic commentaries and phonological descriptions compiled by the Voices of the UK team, Robinson, Gilbert and Herring (Reference Robinson, Gilbert and Herring2010). This section summarises the main characteristics of current WelE and draws together observations on phonetic or regional variation. The regions are based on Penhallurick’s observations on the survey data, and hence, for example, South Wales here includes both the southeast and the southwest, unless stated otherwise.

The general phoneme inventory is described in Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:47) as follows:

• Short vowels: /ɪ ɛ a ʌ̈ ɒ ʊ/

• Long vowels: /iː eː ɛː oe: aː ɔː oː uː/

• Diphthongs: /ɪu ai au ɔi uə iə/

• Unstressed vowels: /i ɪ ə/

• Consonants: /p b t d k g f v θ ð ɬ s z ʃ ʒ x h tʃ dʒ m n ŋ l w j r r̥/

In the short vowels in kit, dress and trap, there is not a great deal of variation. lot/cloth tends to be pronounced as [ɒ], but words spelled with <a>, such as quarry or wash may adopt [a > æ]. Strut tends to be centralised as [ʌ̈], apart from in the northeast, where it emerges as a Cheshire-influenced [ʊ]. For overlapping in the trap, strut and start phonemes, see Hejná (Reference Hejná2018) and Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:50–1, 55–6). One and similar words spelled with <o> group with lot/cloth rather than strut, which may result from Welsh-influenced spelling pronunciation or EngE dialect contact, depending on the region in question. Foot is typically realised as [ʊ].

Regional variation emerges in long vowels and diphthongs, as described in Table 8.1. Excluded from the table for displaying very little variation are fleece [iː] and choice [ɔi]. Square [ɛː > ɛə > eə] as well as the set of triphthongs that typically contain a semi-vowel glide, cure [ɪuwə ~ uə > ɔː], power [auwə ~ auə > aə], and fire [aijə > aɪə > aə], seem to have no clear regionally based variation. Tuesday [iu ~ ɪʊ > juː], similarly, can vary between the diphthong and yod realisations, albeit the former is highly common (e.g. music [mɪʊzɪk], few [fɪʊ]).

A general characteristic of North Wales is pharyngealisation, which applies to most vowels bar the open ones. Changes from the SAWD data in Parry (Reference Parry1999) concern the fading out of r-colouring in nurse and start, for example, previously found to occur in long-standing English regions such as South Pembrokeshire, the Gower, or the Border country, although still attested on occasion (see ‘Rhoticity’ below). The characteristic monophthongs /eː/ and /oː/ seem to be increasingly replaced by diphthongs, especially in the south, and the above patterns (although based on limited data) indicate less of a connection between spelling and pronunciation than observed by Penhallurick (Reference Penhallurick1993; also Paulasto et al. Reference Paulasto, Penhallurick and Jones2020:54–5). The hallmark of Cardiff English is a long nasalised [æː] verging on [ɛ̃:] in the palm/start vowel in working-class speech (Coupland Reference Coupland1988:27; Collins and Mees Reference Collins, Mees and Coupland1990:95). There is evidence of the long [æː] in the Cardiff localities of MMB and Voices as well but not of the raised vowel sound.

The consonantal features of WelE have also experienced some changes between SAWD and the MMB and Voices surveys. The recent findings indicate the following distributions:

Many of the features which display Welsh language contact or convergence (see Morris Reference Morris2013), such as the realisation of dark and light /l/, geminate consonants, or devoicing of sibilants, are alive and well in the dialects. Some interesting developments in the consonant inventory include the emergence of TH-fronting in southeast Wales as well as increasing T-glottalisation, both characteristics of London English and increasingly used across southern England. Some of the Welsh-induced features are waning, however, such as rhoticity, and some traditional EngE features, such as initial voicing of fricatives /f, s/ in the Gower and Pembrokeshire, are no longer attested. Overall, however, the phonology of WelE is more resistant to change than the grammar. This also applies to the ‘lilting’ stress and intonation patterns, resulting in lengthened consonants, a potential for a ‘fuller quality’ (Walters Reference Walters2003:74) of unstressed vowels in final syllables (e.g. common [kɒmɒn], frustrated [fɹəstɹeːtɛd]), and a relatively high range of pitch movement and a delayed pitch rise (depending on the region; see Walters Reference Walters2001; Quaino Reference Quaino2011).

8.3 Morphosyntax

The Welsh accent – by which most people presumably mean the south(east) Welsh accent – is probably the most salient and easily recognisable characteristic of English in Wales, whether we are asking linguists or layperson observers. However, when it comes to morphosyntax, there is an interesting discrepancy between lay views and academic ones. The popular enregisterment of WelE dialect features can be witnessed in online lists and stories on WelE (e.g. Rhys Reference Rhys2014), and, more systematically, in Durham’s (Reference Durham2015) investigation of WelE dialect performance tweets on Twitter. Both approaches reveal that the majority of structural characteristics typically recognised as ‘Welsh’ are shared with EngE, although they do certainly appear in the dialect, especially in the southeast. The top three features of morphosyntax listed in Durham (Reference Durham2015) are right dislocation (by a wide margin), verbal ‑s and clause-final like, while features of a Welsh contact origin – exclamative there, invariant isn’t it, and focus fronting – gain just a few mentions. Recent lay perceptions are most likely influenced by popular TV programmes such as Gavin and Stacey or The Valleys, set in the southeast of Wales. Linguists describing WelE, on the other hand, are likely to draw their attention to characteristics that are original to the variety; what Thomas (Reference Thomas and Burchfield1994:134) calls ‘formative influences in syntax’. His list includes aspectual features (habitual periphrastic do and extended habitual progressive), focus fronting, exclamative there, embedded inversion, infinitive for participle, and extended possessive use of with. A further frequently occurring feature, invariant question tag isn’t it, is mentioned among ‘general vernacular features’ (Thomas Reference Thomas and Burchfield1994:134–41; the present terminology follows Paulasto et al. Reference Paulasto, Penhallurick and Jones2020, however). These features are also the subject of Penhallurick (Reference Penhallurick, Klemola, Kytö and Rissanen1996), Paulasto (Reference Paulasto2006, Reference Paulasto, Durham and Morris2016) and Roller (Reference Roller2016). Although some of them continue to be prominent in the predominantly English-speaking southeast and border areas, they are more likely to be found in the bilingual Welsh-speaking areas.

This chapter will examine the grammar of WelE from a geographic perspective, first identifying features of broad currency and then focusing on the two main dialect areas, the southeast and the rural north and west of Wales, and on the features that distinguish these areas at the level of dialect. The regions are obviously not clear-cut, but there are certainly differences as to the frequencies of individual features. The peripheries (i.e. the long-standing English pockets of the Gower and South Pembrokeshire as well as the Border country), are here left in a minor role. Historically, their regional dialects are largely EngE in origin, and of the traditional features of grammar, the vast majority are by now out of use (for example, archaic or switched pronominal forms, 3sg objective en, archaic verbal agreement forms, a-prefixing). Instead, these dialects have become amalgamated with the neighbouring Welsh and English ones, also drawing influences from other current sources of input, such as southern EngE.

8.4 Lexicon and Discourse Pragmatics

The geography of the WelE dialect lexicon is certainly quite different today from the traditional rural lexicon described in SAWD (Parry Reference Parry1999). It is a customary observation that Welsh-derived dialect lexis is scarce in present-day WelE (e.g. Thomas Reference Thomas and Trudgill1984:193; Penhallurick Reference Penhallurick1993:38–40), a matter which has not changed over the past few decades. There are nevertheless Welsh words which have taken on cultural significance, such as cwtch (or cwtsh) n. ‘hug, small storage space’, or v. ‘to hug, to squat down’, hiraeth ‘longing, homesickness’, and cawl ‘vegetable (and lamb) soup’, and which are thus known and used throughout the country. The Welsh terms for grandparents, nain and taid in the north and mam-gu and tad-cu in the south, continue to be popular as well, despite English-based competition from, for example, nan and grandad (Paulasto et al. Reference Paulasto, Penhallurick and Jones2020:151–3). The amount of Welsh language lexicon is considerably greater when words that display more limited use are taken into consideration.

The majority of dialect lexis in WelE today – as in SAWD, too – is of English language origin. Traditional EngE dialect words have entered Wales in geographic patterns and waves, which produce beautiful regional isoglosses in Parry’s descriptive maps (Parry Reference Parry1999:261–307). The vast variation in lexis, however, leads to much less systematicity in the regional dialects than in phonological or grammatical variation. Some items, such as poll ‘hornless cow’ and yorks ‘knee-straps for trousers’ have spread to nearly all parts of Wales bar the very west, while others have remained in use close to their county of origin (for example, words for newt: asker in Radnorshire and eft in Monmouthshire; Parry Reference Parry1999:294). These lexical isoglosses in the east and south of Wales are also indicative of the impact of informal and formal acquisition of English: to the west and north of them, the word in common use is often the StE one.

Recent lay observations on the matter produce current expressions such as lush ‘lovely’, tamping ‘furious’, tidy ‘nice, decent, excellent’, butt(y) ‘friend’ and mun, a term of address or pragmatic marker ‘man’ (Durham Reference Durham2015). Many of these are decidedly southeastern, as confirmed by the BBC Voices commentaries as well as recent surveys in Gwent (Jones Reference Jones2016b) and in Tonypandy in the Rhondda (see Paulasto et al. Reference Paulasto, Penhallurick and Jones2020:150–7, who discuss select items). The Voices Radio Wales survey, as well as the Voices Language Lab online survey, both extending across the UK and including a focus on regional lexis, are the most extensive sources of data on this subject of late (BBC Voices 2004–2014). Penhallurick (Reference Penhallurick, Upton and Davies2013) examines the compatibility of SAWD with Voices and their potential for carrying out diachronic analyses of dialect change. He concludes, however, that the problems are numerous, from different foci of lexis to different levels of systematicity in the data collection. A handful of the lexical items are the same (e.g. ill, living room, [outside] toilet), which offers some opportunities. Penhallurick (Reference Penhallurick, Upton and Davies2013:128–30) takes up one such item, left-handed, which produces phonetically and regionally similar terms cag[gy]-handed in SAWD and cack-handed in the Voices Language Lab. Welsh terms llawchwith and llaw bwt are used as well, the latter even by non-Welsh-speaking informants in Tonypandy. For more extensive overviews of WelE dialect lexicon, see Jones (Reference Jones2016a) and Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:133–57).

The discourse-pragmatic markers of WelE is an area which is in dire need of further research. A number of such markers have been identified in the dialect (see Williams Reference Williams and Tristram2003 and observations in Jones Reference Jones2018), but their functional and regional distributions are yet to be charted out:

• • now

  thematic marker: My brother now that died, Dai, when I started work, er, he was driving from [sic] my father, driving a horse then.

  (CGC, Williams Reference Williams and Tristram2003:209)

• • now then

  adverbial: so what do you do with yourself now then boy? (Tonypandy)

  discourse marker: Huw Edwards, now then, he’s a newsreader, okay, on BBC television. (NWC Pencaenewydd)

  pragmatic marker: we only speak English now then to someone who doesn’t understand Welsh

  (Llandybie)

• • like

  discourse marker: after the funeral he moved from here, and they emigrated then like (Tonypandy)

  quotative: where they’re more like, ‘yes’, like, ‘no’, and we just like roll all our words together

  (Voices Cardiff/Splott)

• • mind (you)

  yeah well that’s nice mind. Edinburgh is a lovely city that is

  (Tonypandy)

  there’s a lot of us, mind you a lot of old Criccieth people don’t agree

  (NWC Pencaenewydd/Criccieth)

• • see

  ◦ there’s Welsh schools opening see innit

  (Tonypandy)

Paulasto et al. (Reference Paulasto, Penhallurick and Jones2020:169–73) contains further details on the above kinds of usages.

8.5 Recent Research

Sociolinguistic research into Welsh English has a relatively long history, especially when it comes to dialect attitudes and perceptions (Durham and Morris Reference Durham, Morris, Durham and Morris2016; Paulasto et al. Reference Paulasto, Penhallurick and Jones2020:25–35). Recent years have seen a number of interesting contributions into this field, such as Mercedes Durham’s above-mentioned Twitter studies (Durham Reference Durham2015, Reference Durham, Durham and Morris2016). There is also Katja Roller’s (Reference Roller2016) in-depth analysis of the salience of WelE grammar features to Welsh and English informants and Benjamin Jones’s (Reference Jones2018) systematic investigation of the depiction of and attitudes towards WelE in English language fiction over the centuries. Both the present-day and historical sources reveal the salience of features such as focus fronting and invariant isn’t it in and outside Wales, although the history of WelE as a literary dialect really only begins in the late nineteenth or early twentieth century, in tandem with its rising national significance. Chris Montgomery (Reference Montgomery, Durham and Morris2016) uncovers folk linguistic perceptions of WelE dialect areas, including the broad distinctions of North and South Wales, followed by southeast Valleys and Cardiff.

Scholars have also turned their attention to dialect features and sociolinguistic variation. Paulasto (Reference Paulasto, Durham and Morris2016) looks at the levelling and maintenance of three Welsh-induced grammar features in different parts of Wales, while Podhovnik (Reference Podhovnik2008), Morris (Reference Morris2013) and Hejná (Reference Hejná2015) examine phonology, based on their own fieldwork. Besides linguistically detailed analyses, each of them provides sociolinguistic perspectives into their foci of interest. Podhovnik (Reference Podhovnik2008, Reference Podhovnik2010) studies the role of age, gender and educational background in the phonological system of Neath English, and Morris (Reference Morris2013), looking at both Welsh and English, finds variation based on locality, sex, home language, and/or style in his young North Welsh informants’ use of /r/ and /l/. Hejná’s (Reference Hejná2015) study includes the variables of sex and age in the use of pre-aspiration in Aberystwyth English. Research on variation and change in Cardiff English is also ongoing: Durham (Reference Durham2019) examines word order phenomena, while Rowan Campbell’s PhD study (Reference Campbell2021) charts recent sociolinguistic changes in the variety in broader terms. Of significant interest is also the growing immigrant population of Wales, likely to impact not just the linguistic ecology of Wales (Moraru Reference Moraru2016) but also the local ways of speaking English. English(es) in Wales will continue to evolve and require further scholarly attention in the years to come.