2.

WELSH ENGLISH: PAST AND PRESENT

2.1. What is Welsh English?

The Welsh dialects and accents of English are relatively unfamiliar to people outside
Britain, which is why they require a short introduction. The purpose of this subsection
is to sketch the outlines of Welsh English: how it is defined by linguists – the present
writer included – and what its primary characteristics are. ‘Linguists’ is the operative
word in the first instance, as the research interviews have shown that the speakers of
WE themselves rarely agree on the term or indeed are even aware that Welsh varieties
of English might be structurally distinctive in some ways. Let us, however, concentrate
on the scholarly opinions for the moment.
The description of the characteristics of WE will focus on syntax, the main
object of study in this book. The distinguishing features in phonology and lexicon will
receive brief treatments. Section 2.1.3 contains a few words on reports of earlier WE,
whether in literature, letters or other non-linguistic accounts. Quite little is known
about the early forms of WE, but what the sources reveal points towards consistency
in two significant respects: the Welsh language has affected spoken English in Wales
for many centuries, and Welsh-influenced English has never been esteemed particularly
highly (beyond quaint or comical), whether in Wales or outside.

2.1.1. Definitions

WE is characterised somewhat differently by different scholars. Visser (1955: 276)

defines the “Anglo-Welsh dialect” as “English spoken by people who think in Welsh”.
In Awbery’s (1997: 88) opinion it is the English of the traditionally Welsh-speaking
areas, heavily influenced by Welsh speech patterns. To Penhallurick, a knowledge of
Welsh is not an essential factor, although the form taken by WE in each area is
connected to the position of Welsh. He states (1993: 33) that “in broad regional terms,
the pattern of Welsh English is fairly closely aligned with the ‘three-Wales model’” of
Balsom (1985) and Osmond (1988), who divide Wales into the political areas of “Y
Fro Gymraeg”: the Welsh-speaking heartland; “Welsh Wales”: the southern industrial
valleys; and “British Wales”: the border and coastal areas (Osmond 1988: 129, cited in
Penhallurick 1993: 30). This division implies that the influence of Welsh is at its most
noticeable in the English spoken in the heartland, less so in the predominantly English
industrial valleys, and negligible in the border and coastal parts of Wales. All three
varieties can nevertheless be filed under the term Welsh English.
Awbery (1997: 86-88), on the other hand, divides English spoken in Wales into
three distinct varieties according to a somewhat different principle. “Welsh Wales”
refers here to the bulk of Wales: the former consistently Welsh areas, where English
has been introduced in the 19th century and where it “tends to be stereotyped as
8

Welsh English”. In addition, there are the long-standing English areas with a solid
English-speaking population since the early Middle Ages, i.e. Gower, southern
Pembrokeshire and the eastern border area; and the conurbations, most notably
Cardiff and environs, which have developed English dialects of their own. Awbery
does not use the term WE when discussing the latter varieties.
The simplest solution is to define WE as English spoken in Wales; after all, Irish
or Hiberno-English generally refers to English spoken in Ireland and Indian English to
English spoken in India. The above historical and linguistic distinctions and the
heterogeneous nature of present-day varieties of WE nevertheless appear to resist such
a generalisation. It is therefore worth pointing out that in spite of the variation, at least
some of the features characteristic of the Welsh-speaking regions can surface in the
highly anglicised regions, too, be they at the levels of intonation, phonology or syntax.
Coupland (1988: 36-37), for example, observes that the invariant tag isn’t it and fronted
sentence structures are used in Cardiff in much the same way as in WE on the whole.
A special case are the longstanding English pockets of the Gower peninsula and
southern Pembrokeshire: these dialects are the offspring of southwest English dialects,
preserving pointed similarities with their parent dialects which set them apart from the
neighbouring varieties of WE (see, e.g. §3.2 below). These dialects are distictive in their
history, as their origins are external to the country. They are, nevertheless, also varieties
of Welsh English, as their relative seclusion from their mother country has moulded
them in their own directions. Over the past century their connection with the
neighbouring dialects in Wales has become stronger and they have lost much of their
distinctiveness (e.g. Penhallurick 1994: 208-217). In the end, the only viable definition
of WE seems to be the one based on geography: any local dialect or accent of English
which contributes to the dialect map of Wales should be conceived as a variety of WE.
In spite of our best efforts to pin it down, WE remains far from a clear-cut
concept. Thomas (1984: 178-179) observes that it is difficult to determine how many
of the varieties of English spoken in Wales are structurally different enough to be
considered subvarieties or separate dialects, or how many are considered different
dialects by the speakers themselves. He distinguishes two main varieties: northern and
southern. The northern variety represents the transitional stages of language shift with
its stronger dependence on the Welsh sentence structure, whereas in South Wales the
language shift process has ended: the large majority of the people are monolingual
English speakers. Consequently, the Welsh influence gradually fades out towards the
industrial and urban areas. The two varieties are further affected by the neighbouring
English dialects, Liverpool in the north and the Midland and southwest English
dialects in the south. Thomas points out that the division is rough; e.g. the
southwestern dialects, spoken by bilinguals, are more akin to the northern variety in
their propensity to language transfer. The dialect areas have no clearly definable edges:
with the current mobility of the population the dialects are more and more scalar
throughout Wales. Because of the distribution of the population, however, the
southern variety of WE is the most commonly used one.
Garrett, Coupland and Williams (1999: 324) differentiate six to eight dialect
areas of WE “with very different profiles along several dimensions: perceived linguistic
features, affective qualities, prestige, urban/rural character, and ... perceived
Welshness”. In their view, the Welsh heartland, although generally associated with the
 9

prevalence of Welsh, can also be detected in Welsh English. On the English dialect
map it covers not only the traditional north and southwestern parts of Wales, but also
the southeastern valleys, whose English receives high scores for Welshness (op. cit.:
342-344). The schoolteachers taking part in the study evaluated the rural southern
varieties most positively in affective terms, while the urban dialects spoken around
Cardiff and Liverpool, and the Valley dialects in particular, received more negative
assessments. Teenaged informants agreed that the heartland varieties sounded most
Welsh, and they were evaluated fairly positively. However, they found the northeastern
and Cardiff varieties the most socially attractive (1999: 330), which is indicative of the
status of these regions as present-day centres of linguistic innovation (see §3.2).
According to Williams et al. (1996: 180), the localities most salient as geolinguistic
centres of dialect regions are Cardiff, Merthyr Tydfil and Swansea, followed by
Carmarthen, Cardigan, Caernarfon and Bangor. Coupland et al. (1994) have also been
able to find a contender for the position of ‘standard Welsh English’: the accent of the
rural southwest. Unlike any other Welsh accent, it was evaluated positively on all three
scales proposed by Zahn and Hopper (1985): prestige, pleasantness and dynamism.
Furthermore, Coupland et al.’s informants considered the variety highly Welsh-
sounding and it was most frequently named as the Welsh accent that “has very wide
social acceptability over others for life in general in Wales” (1994: 484). The writers
point out (op. cit.: 485), however, that the Welsh dialect map is diverse enough that
caution must be exercised when using labels such as Welsh English or southern Welsh
English. Although WE is employed in the present study as a cover term for the Welsh
varieties of English, it is precisely the diversity which the investigations focus on.

2.1.2. Syntactic characteristics

It is not feasible to describe the varieties of Welsh English conclusively here. This
subsection therefore presents a summary of syntactic features which are common in
spoken WE and/or frequently reported in various descriptions of the dialect. These
are, in other words, some of the features which differentiate WE from StE and other
rural British dialects. Most of them are found in southwestern WE, the variety which
is used also in the Llandybie region, central to the present study. The examples below
are from the Survey of Anglo-Welsh Dialects (i.e. SAWD), either Parry (1999) or the
present corpus (§5.3.3), as well as from other present corpora (§5.3).
As the focus of this study is on the syntax of WE, it will suffice to mention that
also the phonology of WE is distinctive in a number of ways. The orthography of
Welsh is fairly closely aligned with its pronunciation, which is probably the reason why
WE follows the English orthography in some respects. Other noticeable differences
from RP are the use of the monophthongs /e+ / and /o+ / in words such as great and
late, or coal and pole, the voiced, alveolar, rolled /r/, and the consistent use of a similar
/l/ regardless of its position in the word: a clear [l] in South Wales and a dark [{] in the
North, modelled on the respective accents of Welsh (see Parry 1999 for details). The
intonation is also striking, particularly in the Valley dialects and in southwest Wales.
The pitch range tends to be wider than in spoken StE, with a higher degree of
movement. The Welsh stress also functions differently from English, resulting in a
10

distinctive rhythm which is carried over into WE. Middle consonants, for example,
become lengthened after a stressed vowel (e.g. cooking [k k +iõ], ladder [lad +c], busy [b Iz +I],
any [en+I], etc.).
As for the lexicon, there are very few Welsh words – proper names excepted –
which have become integrated into the monolingual English-speakers’ vocabulary.
Eisteddfod, the name of the Welsh cultural festival, is an example of a loanword for
which there is no equivalent in English, and which is therefore commonly used. The
bilinguals may occasionally drop in a Welsh word such as bach, a common term of
endearment meaning ‘dear’, ‘little one’ or ty bach, ‘little house’, a euphemism for
‘outhouse, toilet’. I have also heard South Welsh bilinguals end their English sentences
with the Welsh tag ydy fe instead of the English ‘is it’ (see below; George 1990: 243).
Translation loans are another language contact element, particularly in older speakers’
English: using the verb stop instead of wait (cf. Welsh sefyll ‘to stand, to wait, to stop’),
or turn the tea instead of stir the tea (cf. Welsh troi ‘to turn, to stir, to revolve’; Morris
Jones 1990: 202) may be influenced by the respective Welsh words containing both
meanings.

(2.1) ...she used to go, you see, “you stop indoors now” she said, “I’ll go to see if the
bus is comin’” at half past four in the m ornin’. [...] An’ in the evening comin’
home then, she used to stop up for me, yes. (LC: BE)1
(2.2) draw a photograph ‘to take a photograph’ (cf. Welsh tunny llun ‘to pull a picture’)
(Parry 1999: 149)

The grammar of spoken Welsh English is for the most part very similar to any
Mainstream Modern dialect: it has many of the vernacular elements which are
commonly found in other parts of the English-speaking world as well (cf. Trudgill
1999: 1-6), and in most respects, it fits effortlessly into the spoken British English
continuum. This is perhaps the main reason why its speakers often do not distinguish
WE as a dialect of its own. WE does, however, possess certain structures which are
characteristic of Wales, and many of these structures have been forged by the language
contact situation. In the case of bilingual speakers, these structures are sometimes the
result of direct language transfer from Welsh to English, or they can be substratum
features, relics of the language shift process, which remain a part of the speech of
monolingual English speakers as well as the bilinguals (see chapter 3). Sometimes the
structure of Welsh merely reinforces the use of an already existing standard or
vernacular English structure, in which case the dialectal character of these features is a
matter of frequency and/or function. The effects of language-historical processes and
of other dialects must not be overlooked, either. Southern WE, for example, has also
received influences from the dialects of Southwest England.
All the features below can be found in the Welsh-speaking areas of Wales,
although some of them are perhaps more localised than others.

1
The SED corpus (§5.3.6) reveals that stop is occasionally used with the meaning
stay in the traditional English dialects, too, which makes it less distinctive of WE. In the above
case, however, the connotation of stop is wait, which is unusual in EngE.
 11

The definite article can be used in WE somewhat differently from StE. It can
precede the names of common ailments and languages, for example, and generally be
used in contexts deviating from StE, although its use is not nearly as widespread as in
Hiberno-English. The use of the indefinite article, absent in the Welsh language, can
also be irregular in the English of the elderly bilingual speakers.

(2.3) She could come back to the heat again (SAWD: Gn 9: 2)

(2.4) ...put them peas in for the cows and for the cattle... Very few put the beans for the
animals. (beans mentioned for the first time) (SAWD: Gn 7: 3)
(2.5) Yeah some say that it’s go- the Welsh is going to die, to be wiped, you know- be-
the language would be gone, you know,
[*>Mm.]
like<* the- in Scotland now it’s the Gaelic, you know,
[Yeah.]
well they don’t speak a lot of Gaelic now in Scotland you see. (LC: AM)
(2.6) I got sister, she’s having a birthday that day. (SAWD: Gn 8: 1)

In spite of the Welsh parallels, article use in WE does not always distinguish it from
other spoken varieties of English: expressions such as (got) the toothache and (got) the
headache are also widespread in the SED, and the dropping of the indefinite article is
likewise recorded in the traditional dialects all over England (Parry 1999: 107-8).
However, certain usages, such as definite articles with the names of languages, are
specifically characteristic of WE and likely to arise from Welsh (see, e.g. Thorne 1993:
98), which strengthens the case for substratum influence in the irregularities found in
article use more generally.
The use of periphrastic verb structures in nonstandard contexts is common in
WE, although the forms they take are not as elaborate as, e.g. in HE. What mainly
separates WE verb structures from mainstream English is their less constrained
aspectual use. The expanded ‘be + present participle’ structure, usually progressive in
meaning, can be used in habitual or stative contexts atypical of StE, or it can be
attached to other verb structures to convey habituality. The English progressive form
resembles the Welsh periphrastic structure consisting of bod ‘be’, imperfective marker
yn, and a verbal noun, which corresponds in meaning both to the inflected and the
progressive forms in English. Periphrastic do, found in some southern varieties of WE,
is also habitual in meaning and probably of southwest English origin (§7.2). The
progressive form and its nonstandard uses are central to the present study and will be
discussed further in §4.3 and in chapter 7 below.
Although the above Welsh structure is close to the English progressive form,
the verbal noun does not match the English present participle. The infinitive and the
uninflected verbal noun appear interchangeable, which is reflected in some WE
structures. Parry (1979: 155) notes that in certain areas of Wales stop is used instead of
prevent, and where English usually has the present participle, WE has the infinitive
form.

(2.7) We were stopping them to do that. (SAWD: D 5)

(2.8) To stop the wood to wear out. (SAWD: D 1)
12

A similar structure is found in Welsh, with the particle yn followed by an uninflected

verb. This is the equivalent of the English infinitive (Thomas 1985: 217). The reverse
situation can be explained on the same grounds.

(2.9) you’ve got to put this sharp side... to cut the mouth... to make it bleeding (referring
to the process of breaking a horse by using a special bit) (SAWD: Cl 7)

The structure for to + infinitive is used in the same way as in Hiberno-English,

corresponding to the StE in order to. It is mainly found in the anglicised areas of Wales,
which suggests that it is of English dialect origin, as this structure is also widely
recorded in SED.
The word order of the English indirect question is sometimes inverted in WE,
i.e. the word order of the direct question is retained (e.g Thomas 1985: 217). Indirect
WH-questions have a simple word order inversion, and in the case of indirect yes/no-
questions the connecting if/whether is left out.

(2.10) an’ I am not sure now what would that be exactly... (SAWD: Cl 2: 1)
(2.11) And he asked him would he be interested in coming up as a family. (LC: MT)

Thomas concludes that this feature stems from the corresponding Welsh form,
where the order of the verb and whatever follows right after it is identical in direct and
indirect questions. Omitting the conjunction if/whether is also of Welsh origin: the
corresponding Welsh conjunctions os/a are generally elided (Thorne 1993: 535).
The use of the grammatical focusing devices, fronting and clefting, differ
somewhat from mainstream English usage. Focus fronting, or predicate fronting, is a
common device of speech found all over Wales. The same structure is also used in
informal StE, but it is believed to be more frequent and varied in use in WE (e.g.
Williams 2000: 212). According to Pitkänen (1998), fronting is the more frequently
used of the two focusing devices. What points to Welsh influence most strongly with
respect to clefting is the occasional existential cleft sentence, as in example (2.14).

(2.12) [Would there be a- a local blacksmith as well?]

Yes.
[Where would he be in this area?]
Just up the road he would be. (SAWD: Gn9: 4)

(2.13) [What do you say you do when you strip the feathers off a dead chicken?]
Feather them.
[Yes, would you ever say you pluck them?]
Pluck them, yes, pluck them we do, yes. (SAWD: Cl 1: 2)

(2.14) At one time, it was only these four houses that was here. (LC: LZ)

Focus fronting (FF in brief) is another central construction in the present study, and
further investigations on its use in English, Welsh and Welsh English follow in §4.2
and chapter 6.
 13

Another distinctive feature in WE is its prepositional usage. The Welsh language

lacks a verb equivalent of the English verb have, and possession is indicated by an
existential form of the verb bod ‘be’ followed by the object, gyda ‘with’ and the
possessor, resulting in constructions such as mae car ’da John, ‘John has a car’. A
construction similar to that of Welsh can also be found in WE, receiving the form
‘there’s a car with us’ (Parry 1999: 117-118). In addition to have, the structure ‘with +
pronoun’ sometimes replaces the possessive pronoun in WE, or indicates a more
general ‘possessive’ relationship. All the instances of this structure that are mentioned
by Parry are from Dyfed, southwest Wales. Number (2.15) is an example of the first
type, whereas (2.16) represents a variation of the theme.

(2.15) There’s no horns with the sheep around this way. (SAWD: D 3)
(2.16) Well the boys are grown up with her now, you see? (LC: LZ) ‘her boys have grown
up’
(2.17) [It was when she got older, she had problems with her]
*>Yes, yes, yes.
[hip as well I think but it<* was operated on.]
Oh but it was painful with her. (LC: MP)

Example (2.17) is probably a modified translation of the Welsh phrase oedd poen ’da hi,
meaning ‘she had pain’ (lit. ‘was pain with her’) (cf. King, op. cit.: 236; Gwen Awbery
p.c., Nov. 2005). In most cases, however, it appears that the nonstandard use of with
has expanded in WE beyond its original use on Welsh. Other prepositions used
differently from StE include on, for, of and out of.

(2.18) There’s no Welsh name on these. (Gn 9: 1) cf. Welsh yr enw ar ‘the name on’
(2.19) They do, on times. (Dy 8); cf. Welsh ar brydiau
(2.20) to go for bed (Dy 14); We know for places that haven’t got any (Dy 17). (Parry 1999:
118)
(2.21) Well, the- the queen, th- hey gotta make a ring for her, well every queen has been
ma- a ring has been made out of Wales. (Cl 2: 1) cf. Welsh o Gymru, ‘from, out of
Wales’

In addition, certain phrases in WE originate from Welsh. These include the

distinctive use of there is/are in expletives instead of the standard how, as in ‘There’s twp
you are!’ (LC: BE). This correlates with the Welsh structure dyna twp wyt ti, lit. ‘there
silly are you’. Another, very common substratum feature is the use of the nonspecific
tag is it or isn’t it (innit in colloquial usage) instead of practically every other tag
question.

(2.22) Well he must have passed his BA to be a pharmacist, you know, isn’t it? (LC: ED)
(2.23) Is it ragwort, is it?
[Ragw... *>Er...]
The yellow<*, yellow...
[Yes, yes, could be.]
Bit of a nuisance, that, is it, you know. (SAWD: Gn8: 2)
(2.24) Why are you stayin’ so far away? Travellin’...
[Because I, er, spend my days at the university mostly.]
14

Down in Swansea?
[Mm.]
Is it?
[Yeah, yeah.] (LC: BP)

Williams (2003) explores the development of these tags in WE from their original,
focalising function towards increasing invariability (see also §[Link]).
Further details about the above and other phonological and syntactic features of
WE can be found in several sources. General descriptions are given, e.g. by Parry
(1999), Penhallurick (1993) and Thomas (1984, 1985, 1994). Descriptions which focus
on certain regional varieties are, e.g. Parry (1977, 1979), and Penhallurick (1991, 1994).
Coupland (1990) contains papers on various southern accents and dialects of WE.

2.1.3. Early Welsh English

There is some evidence that a Welsh variety of English was already recognised both in
and outside Wales in the late Middle Ages (see also §2.2 below). Some of the earliest
representations of WE can be found in the work of sixteenth and seventeenth century
playwrights. Although Thomas (1994: 107) observes that they often used dialectal or
colloquial language for the purposes of characterisation or sheer comedy, without
being too specific of the region, some of these dialect features can safely be identified
as ‘Welsh’. Shakespeare, for example, must have had a fair idea of how the Welsh
spoke English, as he manages to plant some quite typical features of WE into the lines
of Sir Hugh Evans in The Merry Wives of Windsor and into those of Fluellen in Henry V
– one of the most famous Welshmen in the history of literature. Fluellen (i.e.
Llewellyn), of course, is an English-speaking Welshman, which was not very common
in Shakespeare’s day and less common still in the days of Henry V, two hundred years
previously. However, Fluellen comes from Monmouthshire in southeast Wales, where
English was already spoken in the fourteenth century. He is also a nobleman, but not
of the most sophisticated sort, and Shakespeare appears to use his Welshness as comic
relief.

FLUELLEN:
Your grandfather of famous memory, an’t please your
majesty, and your great-uncle Edward the Plack
Prince of Wales, as I have read in the chronicles,
fought a most prave pattle here in France.
KING HENRY V:
They did, Fluellen.
FLUELLEN:
Your majesty says very true: if your majesties is
remembered of it, the Welshmen did good service in a
garden where leeks did grow, wearing leeks in their
Monmouth caps; which, your majesty know, to this
hour is an honourable badge of the service; and I do
believe your majesty takes no scorn to wear the leek
 15

upon Saint Tavy’s day.

-----
GOWER:
I think Alexander the Great was born in Macedon; his
father was called Philip of Macedon, as I take it.
FLUELLEN:
I think it is in Macedon where Alexander is porn. I
tell you, captain, if you look in the maps of the
’orld, I warrant you sall find, in the comparisons
between Macedon and Monmouth, that the situations,
look you, is both alike. There is a river in
Macedon; and there is also moreover a river at
Monmouth: it is called Wye at Monmouth; but it is
out of my prains what is the name of the other river;
King Henry V, Act IV, Scene VII

The phonological characteristic of using p for b and t for d is a reflection of the

strongly aspirated voiced consonants in certain varieties of WE, which the
unaccustomed hearer can mistake for voiceless ones. In Thomas’s (1994: 107-108)
opinion, the omission of the initial w in ‘world’ may also be caused by language
transfer, as Welsh has “no sequences of semi-vowel followed by a homorganic or near-
homorganic vowel”. The present WE corpora contain an instance which supports the
hypothesis:

(2.25) it was an wooden block

[I see.]
yes, we were doin’ it ourselves, er… (SAWD: Dy 4: 1)

The informant’s use of the article an is perfectly justified, considering that he

pronounces the following word as [d cn] rather than [w d cn].
Fluellen’s grammar also stands out from the other characters’ sophisticated
language. The lack of concord in cases such as ‘the situations is both alike’ is a
common nonstandard feature of English, but it also corresponds to the Welsh
structure, where the verb expresses no singular/plural contrast when the subject is a
lexical noun. A pronoun as the subject, on the other hand, receives a verb in
concordance with it. This is also the principle of the Northern Subject Rule, widely
used in Northern English and North Midlands dialects, and presumed to be of Celtic
origin (see, e.g. Klemola 2000). Another nonstandard verb structure found in
southeastern WE and typical of the southwest English dialects is the use of the
auxiliary do in affirmative sentences, a feature also known as periphrastic do (‘a garden
where leeks did grow’). Although this is not presumed to be directly a Welsh-
influenced feature, Celtic contact is nevertheless one of the suggested theories for its
development (see §7.2). A third feature often heard in Wales is the use of inverted
word order in indirect questions, such as in ‘it is out of my prains what is the name of
the other river’. It is not uncommon in general vernacular English either, but in the
speech of bilinguals its use is reinforced by the influence of Welsh, as stated above.
Finally, the use of cleft sentences (‘it is in Macedon where Alexander is porn’),
16

although completely grammatical in StE, can also be reinforced by the Welsh language,
in which clefting and fronting function as the primary focusing devices (see §4.2.3).
Thomas (1994: 109) points out the unfortunate fact that there is little first-hand
evidence of WE before the twentieth century. Such evidence could only have survived
in writing, but as the degree of literacy in English among the native Welsh population
was quite low, the educated bilinguals were often used as scribes in both official and
personal matters. The model for written English was that of StE, which is why the
letters, wills and other legal documents preserved from the seventeenth century
onwards reveal little of the English spoken in Wales at the time. Thomas (op. cit.: 110-
111) quotes a letter, written by W. Davies in 1826, which among other colloquialisms
or mistakes contains some structural features affected by Welsh:

(2.26) ‘I am against you to come’

cf. ‘wyf yn erbyn ichwi ddod’ (lit. am(-I) against for-you come)
(2.27) ‘I think to be at Newtown… ’
cf. ‘wyf yn meddwl bod yn… ’ (lit. am(-I) ‘in’ think/ expect/ intend be in… )

Structures which clearly originate from Welsh are however scarce in the written
documents. Thomas (op. cit.: 107-112) contains a brief survey of the available material.
Having also sat in the document room of the National Library of Wales in
Aberystwyth for several days, I can only verify his conclusions: most of the letters are
official in nature, which necessarily affects the style used, but even the personal letters
from the 18th and 19th centuries tend to be highly formal. The Voelas estate
correspondence from Anglesey, also mentioned by Thomas, is a more likely source of
nonstandard language, as these are mainly short notes, e.g. between farmhands and
estate owners, and thus either not as polished as regular letters or written by fairly
uneducated people. The correspondence contains some nonstandard forms, such as
concordance errors:

(2.28) C 180 ‘...my later neighbours there knows this...’

‘...I suppose the landlord and tenant together reaps the benefit...’

The following structure, on the other hand, possibly results from the lack of an
existential there in Welsh, the initial verb mae ‘is’ being existential in itself. The omission
of the indefinite article may be of similar language contact origin.

(2.29) C 782-3 ‘All is finished inside but painting . roof is on the [Link]’
cf. ‘mae to ar y tí’ (lit. is roof on the house)

The bilingual writer may switch between Welsh and English, as in some of the Charles
and Davies Letters from Carmarthenshire (A 107-111): in a letter to his sister Eliza,
from the year 1825, David Charles the younger uses Welsh when quoting his
acquaintances and even mixes Welsh words and phrases into the English narrative.
However, structural transfer does not take place.
A few isolated comments on early WE can be found elsewhere. For example, it
becomes clear that the commissioners who conducted the fieldwork for the Inquiry into
 17

the State of Education in Wales (1847) (see §2.2.3) did not appreciate the Welsh accent.
Roberts (1996: 187) cites their description of the speech of schoolchildren, who “recite
material in English learned by heart, ‘in a Welsh screech which seems expressly devised
to annihilate all chance of expression or modulation of tone’” (op. cit.: 25). Rev.
Richard Warner, on the other hand, has not a bad word to say about the English of the
bilingual children he encountered on this travels (1798/1800: 262-3, cited in Ellis 1882:
189-190): “all the children of Flintshire speak English very well, and were it not for a
little curl, or elevation of the voice, at the conclusion of the sentence (which has a
pleasing effect), one should perceive no difference in this respect between the North
Wallians and the natives of England”.
Welsh English, as a dialect of English with a Welsh substratum, arose most
strongly in the industrial valleys of Glamorgan in the late 19th and early 20th centuries.
This development will be followed more closely in the sections below.

2.2. The advance of English into Wales and the history of Welsh English

This section deals with the anglicisation of Wales, from a thoroughly Welsh to a
predominantly – and increasingly – English-speaking land. Language contact situations
and their outcomes have their own, unique characteristics. As the history of Welsh and
English in Wales plays a very important role both in the linguistic features examined
here and in the sociolinguistic positions of the two languages today, it is essential to
examine the past circumstances which have led to the present ones.

2.2.1. The Britons and the Anglo-Saxons

The Celts first arrived at the British Isles, inhabited by the Picts at the time, in the
middle of the first millennium BC. They spoke two kinds of Celtic languages: Goidelic
or Q-Celtic in Ireland and Scotland (north of the Firths of Forth and Clyde; Jackson
1953/1994: 7), and Brythonic or P-Celtic in the rest of Britain, i.e. in modern England
and Wales. Most of the continental Gauls were also speakers of a P-Celtic language
(op. cit.: 6), but there are varying opinions regarding the linguistic origins of the
Hispano-Celtic people. The Celtic languages constitute a branch of the Indo-European
language family along with, e.g. Germanic, Italic, or Balto-Slavic languages, but they are
typologically quite distinct in many ways. The features most characteristic of Insular
Celtic languages are mutations of word-initial consonants, inflected prepositions, VSO
word order and the use of particles to mark subordinate clauses or illocutionary force
(Fife 1993a: 8-24; see §4.1 below).
During the first centuries AD the culture and society of the Britons became
affected by those of the Roman invaders, who established their position as the
aristocracy of the country. The collapse of the Empire and the retreat of the Romans
left a void in the society, which the Germanic tribes came to fill in the mid-fifth
century. Groups of Angles, Saxons and Jutes had been arriving to the isles in earlier
times as well, but it was now that they seized the opportunity to take control over the
southeast of the island.
18

Venerable Bede dates the arrival of the Anglo-Saxons to AD 449, and makes
reference to the invincibility of the new invaders as opposed to the cowardly and
treacherous Celts. Tristram (1999: 5, 8-9), however, questions the objectivity of “early
medieval Christian moralists” such as Bede (AD 731) who – like many of their
contemporaries – deeply resented the Britons for their obstinate political and religious
independence. She points out that the 19th century ‘Anglo-Saxonist’ approach relied
on these biassed historians, creating a notion of Anglo-Saxon purity, which required
the interpretation that the Britons were simply slain or expelled from England.2 The
Anglo-Saxons did conduct raids across the island and they expanded their territories
through battles and colonisations over the following centuries, forcing the Brythonic
kingdoms to retreat to the west and the north (Jackson 1953/1994: 198-219). In
Higham’s view (1992: 224), the Britons who became the Anglo-Saxons’ primary targets
mainly belonged to the Romanised aristocracy, estranged too far from the masses to
rise to defend themselves or their subjects, and so the majority of the population was
left to “make what terms they could” with the invaders. A part of the population from
the southwest of the country fled across the sea to Brittany, where their descendants
have carried on speaking the Breton language to this day. In spite of this, the Brythonic
peasant population remained widespread in England.
Modern historians and archeologists increasingly recognise that to a great extent,
the Britons merged with the Anglo-Saxons and gradually adopted their culture. This is
evidenced by archeological findings such as culture-specific burial rites and styles of
manufacture, and paleobotanical findings, exhibiting “many signs of continuity in land
units, in enclosed landscapes, and therefore within the rural population and the
geography of exploitation” (Higham 1992: 229; see also Härke 2003). Cultural
exchange between the Anglo-Saxon minority and the Brythonic majority was probably
mutual: German (2000: 368-9) suggests that while the fusion of the Britons and Anglo-
Saxons occurred rather rapidly, over two or three centuries, there was widespread bi-
or multilingualism among both groups during this period. Large numbers of Brythonic
place names remain in England (see the gazetteer in Coates & Breeze 2000), which is
an indication of a relatively successful coexistence between the Britons and the
invaders, although a much greater number of place names was replaced by Anglo-
Saxon ones in the sixth and seventh centuries (Higham 1992: 229-230).
It has been acknowledged that Brythonic was influenced by the language
contact, but the convergence of Old English towards Brythonic is a matter which has
attained growing interest in the late twentieth century. According to the Anglo-
Saxonist tradition, again, lexical or structural transfer from Brythonic to Old English
was impossible, the Britons being considered a subordinate people. However,
historians and archeologists having thrown new light on the subject, a whole body of
research currently focuses on the early contact between the two language groups and
its possible outcomes in the English language through substratum influence (see, e.g.
Tristram 1997, 2000, 2003 and Filppula et al. 2002).

2
For a discussion on the rise and fall of the Anglo-Saxonist tradition, see German
2000.
 19

The Anglo-Saxons called the Britons wealas ‘foreigners’. This is etymologically a

Germanic term applied to the Celtic tribes in Europe (Friedman 2001), but after the
Romanisation of the Celts, Germanic speakers extended the term also to other
Romanised peoples along the borders of the Empire: the Walloons of Belgium, the
Welsch of the Italian Tyrol and the Vlachs of Romania (John Davies 1993: 71). The
Welsh, on the other hand, referred to themselves as Y Cymry, ‘the compatriots’, and to
their language as Cymraeg. The English are still called Saesneg in Welsh, ‘Saxons’.
According to Davies (1993: 68), Brythonic survived as a spoken language in
many parts of eastern England until the eighth century. By this time the division
between west and southwest Brythonic - Welsh and Cornish - was already in existence:
Jackson (1953/1994: 25) observes that the two varieties already differed in the first
century AD, but that the differentiation accelerated in the fifth century. The campaign
of Dyrham in 577 and the subsequent Saxon occupation put a stop on further
interaction between the speakers of the two varieties. Similarly, the northwestern
branch of Brythonic was cut off from the southern ones, developing into Cumbric,
which ceased to be spoken early on – probably in the eleventh century, when the
kingdom of Strathclyde was merged in the kingdom of Scotland (op. cit.: 9). Cornish
survived as a community language until the eighteenth century, but not until the
nineteenth (Stalmaszczyk 1999). In Wales, Welsh remained the sole language of the
population until the arrival of the Normans in the eleventh century. The language of
the new invaders, French, contributed to the lexicon of Welsh as well as to that of
English. This is also the period which witnessed the beginnings of WE in Wales.

2.2.2. From the early Middle Ages to the 18th century

English has spread into Wales so gradually that it makes Welsh English both a
historical and a fairly fresh language variety. As mentioned in §2.1.1, Awbery (1997)
divides the English language as spoken in Wales into three varieties, which in some
ways depict the stages in which English was introduced to Wales. The varieties are the
English spoken 1) in the longstanding English areas, 2) in Welsh Wales and 3) in the
conurbations. In the first regions, English has been the majority language for centuries.
Williams (1935: 242) suggests that English may already have been the spoken language
of the eastern Radnorshire plain in Mid-Wales as early as the eighth century AD.
Southern Pembroke and the Gower peninsula, along with the eastern edges of the
border area also belong to this category. They were located within the territories of the
Anglo-Norman Marcher Lords, but furthermore, in the early twelfth century, Henry I
permitted English and Flemish colonists to settle in South Pembroke and the Gower,
because “a dense settlement of peasants is always a more effective way of consolidating
conquest than a thin layer of gentry”, as John Davies (1993: 114) points out. Welsh was
consequently uprooted in the Gower and southern Pembroke, and being little pockets
of English surrounded by Welsh-speaking villages, they developed distinctive dialects
of their own (see Penhallurick 1994 for a description of Gower English and Parry
1990b for a summary on South Pembrokeshire English). They do not seem to have
expanded from their original size over the centuries, and neither have they experienced
significant re-cymricisation. South Pembrokeshire is still commonly known by its
20

medieval nickname ‘Little England beyond Wales’. Ellis (1882: 176) points out that
Flemish bore such a strong resemblance to West Saxon at the time that it has probably
left no mark at all in the present dialect.
The anglicisation of Awbery’s ‘Welsh Wales’ is a two-staged process which
continues to this day. It began from the English border, the coastlines and the towns,
but it was accelerated significantly by the industrial revolution. The first phase was
characterised by a clear-cut separation of Welsh and English, peasantry and gentry,
with a marked geographic distribution and an extremely slow infiltration of English
among the peasants from the east. The boundaries that were established by the early
12th century between Pura Wallia, the Welsh kingdoms in the north and west, and
Marchia Wallie, the territories of the Anglo-Norman lords in the east and south of
Wales, are already indicative of the modern language divide. Still, in many parts of the
Marcher lands the bulk of the population remained consistently Welsh in speech and
culture until the industrial age (Williams 1985: 65).
After two centuries of warring and partial colonisation since the coming of the
Normans, the English finally gained political power in Wales through the subjugation
and death of Llywelyn ap Gruffudd (Llywelyn II), the last indigenous prince of Wales,
in 1282 (John Davies 1993: 158-61).
The division between the gentry and the peasants in Wales became further
heightened by their different languages during the Middle Ages. Janet Davies (1993:
19-22) states that by the fifteenth century English had replaced French and Latin as the
language of law and administration, whereas the common people remained monoglot
Welsh. After the defeat of Llywelyn ap Gruffudd, the English strengthened their
position by establishing walled and fortified towns along the English border and on the
northern coast. Legal transactions were increasingly conducted in English, and
although Welsh had been the language of both Church and State in Wales, the Welsh
gentry began to feel that a fluent command of English was vital for them. Individual
bilinguality was nevertheless rare and existed mainly in the Marcher lands and in towns.
Williams (1990: 19) observes that “on a world historical scale, Wales might be
considered as the first colony of an expanding British state”. Wales was annexed to
England first through bringing the local Welsh nobles under the authority of the
English King, and formally with the Acts of Union in 1536 and 1543. The Laws in
Wales Act determined that English was to be the only official language in Wales, and
no Welsh-speaker was allowed an official position. The following extract from the first
Act is very straightforward and extremely discouraging for Welsh (cited in Aitchison
and Carter 2000a: 27):

And also by the cause that people of the same dominion have and do daily use a speech
nothing like or consonant to the natural mother tongue used within this realm some rude
and ignorant people have made distinction and diversity between the King’s subjects of
this Realm and his subjects of the said dominion and Principality of W ales. Whereby
great discord variance debate division murmur and sedition have grown between his said
subjects ... his highness therefore minding and intending to reduce them to perfect order
notice and knowledge of the laws ... and utterly to extirpate all and singular the sinister
usages and customs differing ... his said country or dominion of Wales shall stand and
continue from henceforth incorporated united and annexed to and with his realm of
England.
 21

The Acts of Union alienated the monoglot Welsh population from their own
courts of law and administrative activities, thus forcing their language to retreat into
the kitchen and the church. Linguistic difference fed class difference, and vice versa,
until the Welsh gentry completely abandoned the Welsh language in the fear of having
their “ability to speak polished English ... marred by even a nodding acquaintance of
Welsh” (Janet Davies 1993: 23). The Welsh gentry were by and large anglicised by the
eighteenth century.
The Welsh language received an institutional refuge from external oppression
with the translation of the Bible in 1588. Several scholars, including Williams (1990:
21), assert that this was an absolutely crucial event for the future of the language. While
Welsh was a highly sophisticated language with a strong literary tradition beginning in
the early 7th century, English was in the process of taking its place as the language of
all high domains. Religion, however, became a domain which could not be conquered,
and as Janet Davies (1993: 26) concludes, it helped the Welsh build themselves a “lofty
image” of their language. She also observes that Welsh was the only one of the non-
state languages in Europe to have a published translation of the Bible less than a
hundred years after the Protestant Reformation, which partly explains the difference
between the subsequent history of Welsh and that of other similar languages, such as
Irish and Scots Gaelic. By the mid-nineteenth century the Anglican Church was
practically replaced by Nonconformism, which was to become one of the greatest
advocates for Welsh.
Granting the Welsh language a religious role was not, however, merely an act of
good will from the part of the English government. As Thomas (1994: 97) points out,
it was also a tool for a better control of Wales in isolating the monolingual, socially
disadvantaged peasantry from the political aims of the English-speaking gentry. Those
aims were, of course, to create a strong, unified kingdom, in which the Welsh speakers
were to have no place. Permitting the use of Welsh for religious functions but
forbidding it from the secular created “an effective language buffer” between the
native Welsh peasantry and the anglicised gentry, a buffer which neither party was keen
to break down. There was only a narrow belt where both languages were spoken
surrounding the Welsh heartland, which by the eighteenth century had contracted only
slightly (see map 2).
As the two languages had distinctly different roles in the society, there was
diglossia in Wales, but there was little individual bilingualism. The two language
communities had little interaction with each other. Fishman (1972: 143) refers to
Friedrich (1966), when stating that

this pattern is characteristic of polities that are largely economically underdeveloped and
unmobilised, combining groups that are locked into opposite extrem es of the social
spectrum, and, therefore, conducive to groups that operate within extremely restricted
and discontinuous linguistic repertoires.
22

Map 2. Language zones in the mid-eighteenth century (Pryce 1978: 242;

© W.T.R. Pryce 1999. Reproduced with the permission of the author).
 23

The language situation of Wales at this time was probably similar in many ways to
Friedrich’s Tsarist Russia, where the aristocracy spoke French. Interpreters were
needed in various contexts: Thomas (1994: 96) mentions that in the fifteenth century
there were small numbers of educated gentry, who functioned as representatives for
the uneducated and monolingual peasantry in official matters. However, bilingualism
among the gentry was more or less a transitional phase towards English
monolingualism, because as Jones (1993: 541) points out, the Welsh language already
had a low status at this time, and the aristocracy aspired after the air of prosperity and
sophistication represented by English.
Before the beginning of the industrial era there was relatively little migration
within Wales and thus the diffusion of English into Wales remained slow. There was
no pressing need for the people of the rural heartlands to acquire English skills, but in
the eastern borderland villages, where the inhabitants interacted with English speakers,
the numbers of bilingual individuals were on the increase. English became not solely
the language of the upper classes, but it pervaded the entire population to an extent.
Pryce (1990: 52, 54) states that the River Usk formed a boundary between the Welsh
and English language communities in Gwent in 1771. His conclusions are based on the
Church in Wales Records which indicate which language was in use in the ancient
parish churches at that time. In the churches along the transitional zone, where
services were held in both languages, the parishioners were not likely to be bilingual.
Rather, there were two coexisting language communities in these villages. The bilingual
individuals probably acted as mediators between the language groups, because as
Thomas (1994: 99) observes, a lack of interpreters would have resulted in the creation
of a pidgin, “for which there is no evidence in Wales” (see §3.1).
The English spoken by the rural population of the border areas was likely to be
influenced by the Welsh language. MacCann and Connolly (1933: 56, cited in Pryce
1990: 50) tell of a seventeenth century gentleman in Abergavenny who sent his son to
London “so that he might acquire a sound knowledge of English ‘without any
corruption from his mother tongue which doth commonly infect men of our
countree’”. In addition to Welsh, the dialects of the West Midlands and the southwest
of England left their marks on spoken English in Gwent, and have since continued to
influence the southeastern dialects. This is evidenced by certain geographic patterns in
twentieth century and present-day WE (cf. §3.2).

2.2.3. The industrial age

Pryce (1978: 229-230) draws a sharp distinction between the pre-industrial ‘old’ Wales
– rural, Welsh-speaking, chancery towns being the most important regional centres –
and the industrial ‘new’ Wales – urban, increasingly English-speaking, population and
administration centralising to the south and northeast. Indeed, the development which
began in the 1770s changed the cultural and linguistic situation of Wales completely
over a century and a half. No major population movements within Wales had taken
place earlier, which is why the anglicisation of Wales had proceeded slowly, through
two restricted channels: the upper classes and the English border. During the industrial
era, however, people in Wales and in the rest of Europe left their home regions in
24

their thousands in search for work and new opportunities. The Welsh and English
languages collided and fought for dominance, particularly in the southeastern
coalmining valleys.
What set the migrations in motion was the massive population growth. In 1770
there were some 500,000 people in Wales (John Davies 1993: 320), but in 1851 the
population figure was 1,188,914, which by 1911 had ballooned to 2,442,041 (Jenkins
1998b: 1, with statistics derived from Jones 1998). Davies (op. cit.: 320-323) discusses
the reasons for the mushroom growth between 1770 and 1851, and concludes that
although Wales received more immigrants than any other country in Europe during the
nineteenth century, this was probably not the primary catalyst, as the numbers
emigrating to England and America during that time were large enough to compensate
for the immigrants. Neither was the growth restricted to the industrial areas, but it also
concerned rural Wales, as well as countries such as Ireland and Finland, which did not
experience an industrial revolution before the mid-nineteenth century. What we are left
with are the effects of the decreased death rate and increased birth rate: people lived
longer, and being confident that the economic prospects were good they also married
younger and had large families. The consequent surplus of labour triggered the
migrations, which in the Welsh setting were directed to England and America as well
as to the domestic coalfields.
Compared to Ireland, however, a significantly larger percentage of the people
were able to find work in their home land. Emigration was one of the key elements
which eroded the Gaelic language, but in Wales the direction of language shift was
momentarily changed as the monoglot Welsh speakers flooded the southeastern
valleys. Jenkins (1998b: 2) relates that by 1911 nearly two thirds of the Welsh lived in
the southeastern counties of Glamorgan and Monmouth. As the people gravitated to
the south, they became increasingly urbanised: two features which still characterise the
distribution of population in Wales. Williams (1985: 178) points out that Wales was
truly exceptional among the nations of Britain with its slow rate of emigration in the
late nineteenth century. In the decades before the First World War Wales became the
only country outside the USA to register net immigration.
The thriving coal industry of the late nineteenth century has been regarded both
as the curse and the blessing of the Welsh language: a curse because of the swift
advance of English during that time, but also a blessing for the reason that without the
employment it offered the fate of Welsh could have turned out worse still, as the
example of Ireland demonstrates. For the English language industrialism was, in any
case, a most effective vehicle indeed. Jenkins (1998b: 3) gives the figures for Welsh
speakers, from which we can deduce that the percentage of monoglot English speakers
was perhaps five per cent in the late eighteenth century, but rose to 55.4 per cent by
1911. Numerically the increase was enormous: from less than 30,000 to c. 1,350,000.
Only 8.7 per cent of the population remained monoglot Welsh speakers in 1911. Thus
the number of people with a knowledge of English had actually risen seventy-fold, and
bilingualism among the Welsh speakers had become widespread.
During the first half of the nineteenth century English-speaking immigrants to
the coalfield were far outnumbered by those originating from Wales, which
strengthened the position of Welsh in the southern towns and valley communities.
Welsh periodicals became numerous, Welsh cultural activities were blooming, and
 25

English-speaking immigrant miners often had to start learning Welsh to be able to

communicate with their co-workers. With the proliferation of the population, the
number of Welsh speakers kept growing until it peaked in 1911 at 977,366. Both
Williams (1990: 31-35) and Jenkins (1998b: 5-12) refer to Thomas (e.g. 1959 and 1987),
who argued forcefully for the significance of industrialism as a factor which created the
“cauldron of rebirth” in southeast Wales and therefore helped maintain the Welsh
language. Williams and Jenkins criticise Thomas’s conclusions on the grounds that this
state was not to last: by 1881 the percentage of immigrants from England, Scotland
and Ireland to Glamorgan and the coalfield had grown to 56.9 per cent, expanding by
1911 to 68.6 per cent (Jones 1998: 155). A development such as this could not help but
affect the linguistic situation. Furthermore, Welsh was not considered a prestigious
language, so there was no particular reason for its speakers to want to hold on to it.
The social position of Welsh in relation to English was fairly similar to what it had
been during the previous centuries; only now the English-speaking gentry had been
replaced by English industrial moguls. The English language offered a better access to
social advancement and many Welsh speakers seized the opportunity. In the decades
around the turn of the 20th century, the densely populated South Wales coalfield
remained linguistically diverse and fluctuating, or, in Jenkins’ (1998b: 11) words, “a
huge, complex, amorphous, even chaotic, sprawl of intensely divergent linguistic
communities ... [where] virtually no community was sheltered from prevailing English
influences”.
However, with the main weight of the population in relatively small areas, the
statistics distort the picture somewhat. Geographically the heavy anglicisation only
concerned the industrial counties of the southeast and northeast. The rural counties of
Ceredigion, Merionethshire, Caernarfonshire and Anglesey remained approximately 90
per cent Welsh-speaking in 1911. Although bilingualism became more common all
over Wales in the nineteenth century, the heartlands were able to resist monoglot
English invasion because the migrations were directed to specific areas. On the other
hand, these areas suffered a fate similar to other rural parts of Europe: “the pull of
urban and industrial employment weakened the demographic base of rural Wales
which was so vital to the reproduction of its economic and social order” (Williams
1990: 32). English was brought into the rural counties by the returning migratory
labour, increased tourism, and the newly-built roads and railways. Williams (1985: 145)
states that the working people of the rural uplands in south-west Wales obtained a
large part of their income through seasonal migration to the southern and eastern
valleys as well as to England. Odlin (1997) concludes that seasonal migrations from the
Gaeltacht were instrumental in introducing English into the monolingual Irish areas,
and it is likely that they had a similar effect in the heartlands of Wales.
Roberts (1996: 171-172) writes that

the mid-nineteenth century ‘culture of progress’ regarded the deliberate maintenance of

the d istinctive cultural practises of W elsh-speaking W ales as a threat, by its very
existence within a supposedly unified Britain, which must be removed in the interests of
both communities.
26

Whereas superficial cultural differences (represented e.g. by the Scottish tartans, kilts
and bagpipes) were tolerated, “to find fundamental cultural differences where only
superficial ones had been expected was, however, felt to be profoundly disturbing and
potentially dangerous; something clearly had to be done.” The position of English as
‘the natural Mother Tongue’ of the kingdom was therefore actively promoted through
the reformation of the educational system. The state of education in Wales was
investigated extensively by the Commissioners of Inquiry, and the results were
published in 1847 in a report subsequently dubbed as the Treachery of the Blue Books.
The commissioners found the Welsh language to be the stumbling block of the nation,
as it isolated the Welsh from ‘polite society’ and the benefits of progress; ‘progress’
meaning, in the language of industrialism, material wealth (Aitchison and Carter 2000a:
34). In the words of one of the commissioners, “the Welsh language is a vast drawback
to Wales, and a manifold barrier to the moral progress and commercial prosperity of
the people. It is not easy to over-estimate its evil effects” (cited in Roberts 1996: 185).
As it was unthinkable at the time that the position of Welsh should be
improved, it was instead deemed the reason for the vast class differences. Jenkins
(1998b: 12) states that the commissioners also expressed doubts concerning the
legitimacy of Welsh and the morals of the Nonconformist chapelgoers, which caused
outrage but also severely deteriorated the nation’s confidence in their language. By the
time that a new system of elementary education was being designed for the children of
Wales, both the English promoters and the Welsh themselves were ready to reject
education through the medium of Welsh as ineffective and useless (Jones 1993: 548).
Instead, a general introduction of English was considered the solution. Janet
Davies (1993: 48-49) relates that from the mid-nineteenth century onwards the
number of elementary schools began rising significantly, with annual tests for the
pupils to show that they were making progress in arithmetic and in the reading and
writing of English. A school could lose a considerable part of its state grant if the
pupils failed, which is why teachers were eager to favour English at the expense of
Welsh. Subsequently, the use of Welsh was often entirely forbidden in schools. The
most infamous means of punishing a child for speaking the wrong language was the
‘Welsh Not’ (also known as the ‘Welsh Knot’ or ‘Welsh Note’), a wooden plaque hung
around the neck of the poor offender. The only way that the pupil could rid himself of
the sign was by catching somebody else speaking Welsh. The child wearing the sign at
the end of the day was severely punished. Every Welsh-speaker since has heard of this
long-gone practise, as it has become an integral part of national memory. However,
Davies (ibid.) concludes that the ‘Welsh Not’ was probably not as widespread as
people today tend to assume.
Awbery (1997: 87-88) states that in previous times schoolchildren had often
learned very little English, since their parents could only afford to give them a minimal
education, but as elementary education became compulsory, free, and thoroughly
English with the Education Act of 1870, even the rural language communities gradually
became bilingual: nearly everyone could speak English as well as Welsh. Jones (1993:
549) demonstrates that the most rapid period of decline for the Welsh language - and
the most rapid advance of English - took place between 1871 and 1921, when the
double effects of immigration and the Education Act conjoined. When the negative
 27

attitude towards Welsh and the general unwillingness to support it are added, we can
only conclude that Welsh was given an extremely rough deal. How did it survive at all?
Throughout this period the number of Welsh speakers actually grew, and a
counter-movement for the Welsh language and education was begun at the end of the
century. In Jones’s opinion (1980: 58), one of the major forces which helped Welsh
survive through its ordeals was its resilient connection with nonconformist religion: “it
is arguable that so powerful was the alliance of language with religion that it was able
to contain the challenge of English culture up to and even beyond the first world war”.
Jones (op. cit.: 67-68) points out that for the common Welsh-speaking people their
language remained the medium of all social and public activities. It was a defence
against the anglicising forces of industrialism, which threatened to uproot the people
from their cultural heritage. On the general level, however, the two languages operated
through distinctly different functions, Welsh possessing only “restricted value” (op.
cit.: 61). Thus, the view that a knowledge of English was socially and economically
essential was stamped on the minds of the Welsh.
Pryce (1978) and Jones (1976; cited in Aitchison and Carter 2000a: 40) conclude
that at this time in southeast Wales, bilingualism was just a transitional stage in the
language shift process, a bilingual generation being followed by a monoglot English
one. Jones (1976: 347) views the anglicisation of the coalfield area through the shifting
language of the Baptist chapels. In his opinion the shift took place over the century in
three stages: from “uniformly Welsh” through “linguistic heterogeneity” to the
dominance of English.

2.2.4. Twentieth century developments

By the census of 1911 Welsh was no longer the language of the majority. The mid-war
depression, in turn, collapsed the Welsh economy and caused a flood of emigration.
Thomas (1987: 437, cited in Aitchison and Carter 2000a: 38) argues that this was the
greatest blow of all to the Welsh language:

Because of the dazzling heights reached just before World War I, the subsequent fall was
all the more disastrous. The class war in the coalfields intensified, and the clarion call
was Marxist not Methodist. What the potato famine did to the Irish economy, the great
d e p re ssio n d id to th e W e lsh e c o n o m y . In th e tw entieth century, econom ic and
dem ographic contraction, the decline of nonconform ity, severe unem ploym ent and
emigration ... have been a curse to the language.

Unemployment became so severe at the time that leaving Wales seemed the only
option. Janet Davies (1993: 59) reports that around 390,000 people moved out of
Wales between the wars, mainly to the Midlands and to the southeast of England.
Welsh began to seem like a language not worth passing to the children. As Thomas
indicates, it had also become scorned for political reasons in South Wales. It
represented the Nonconformist and Liberalist traditions of Wales, which the Socialist
ideology sought to replace in the early twentieth century. The industrial valleys in
particular turned into a bastion of Marxism, which religious life, and therefore the
28

Welsh language, had little part in. English was clearly better suited for uniting the
workers of the world (see also Jones 1980).
English began spreading into Wales all the more effectively with the
development of the media in the 1920s and 30s. Aitchison and Carter (2000a: 40) point
out that regardless of the previous development in the position of English throughout
the land, it was only with the coming of the mass media that Welsh speakers all over
Wales obtained direct contact with English and the geographic language frontier began
to melt away. English daily newspapers replaced weekly or monthly Welsh ones,
cinemas with their English and American talkies became enormously popular, and the
BBC began broadcasting from Cardiff and Swansea with overwhelmingly English
programming. With the proliferation of bus routes, cars and motorcycles, the
population became increasingly mobile and English tourists began to find even the
remotest Welsh villages (Janet Davies 1993: 60). English, the former language of the
upper classes, had become available to everyone, and its good command was not only
socially desirable, but an essential means of economic progress.
The state of Welsh was allowed to become somewhat critical before
administrative measures were taken. Its position was improved gradually, e.g. through
the Welsh Courts Act of 1942, the Welsh Language Act of 1967 and the Welsh
Language Act of 1993 (Aitchison & Carter 2000a: 46, 136), which finally gave Welsh
equal rights with English, such as can be guaranteed through legislation. Welsh-
medium education, at least on the primary level, became common around the turn of
the 20th century, and Welsh radio and TV channels were later founded to compete
with the flood of English which poured into Welsh homes.
Whereas the decline of Welsh was at its strongest in the industrial regions during
the economic recession of the 1930s and WWII, several of the rural counties
experienced severe decline in the percentage of Welsh speakers in the period of 1951
to 1971. Pryce (1978: 232) states that the twentieth century educational policies (prior
to the 1980s) were designed bearing in mind the need to establish bilingual schools in
the anglicised areas, yet under the impression that the Welsh-speaking pupils’ home
background was sufficient to guarantee the maintenance of Welsh in the heartlands.
This attitude may be one of the reasons why, in spite of the increased institutional
support, the percentage of Welsh speakers continued to go down. Baker (1985: 181)
was still quite pessimistic about the future of Welsh, when viewing the language
situation in the mid-eighties. At the time the census evidence did indeed point towards
inevitable decline. In the 1991 census, however, there was a visible change in the trend,
showing that the decline in the percentage of Welsh speakers had slowed down to near
standstill: from 18.9 per cent in 1981 to 18.6 per cent in 1991 (Aitchison & Carter
2000a: 51, 89).
The latest, 2001 figure is more promising still: 20.5 per cent or 575,640 people
in Wales aged 3 or over speak Welsh (Aitchison & Carter 2004: 49). This historical
census is the first ever to show an increase in both the percentage and the numbers of
Welsh speakers. The increase from the 1991 population figure of 508,100 is quite
 29

clear3 ; it is encouraging for the language that it has been able to sustain the growth in
spite of the general increase in population. When everyone in Wales with any Welsh
skills is included, whether in reading, writing, speaking or understanding the language,
the percentage rises as high as 28.4 (op. cit.: 39). This information cannot be elicited
from the previous census data.
Yet, there are also drawbacks: the percentage of those being able to speak, read
and write Welsh is only 16.3, and more significantly still, the increase in Welsh speakers
continues to concentrate in the cities and the heavily anglicised areas. The counties
showing the largest decreases in Welsh speakers are Carmarthenshire, Ceredigion,
Gwynedd and Ynys Môn4 , the traditional bastions of the Welsh language (Aitchison
and Carter 2004: 50). The language situation seems to be balancing: the Welsh
communities in the west and north of the country continue to anglicise, while the
numbers of Welsh speakers in the previously heavily anglicised regions are growing.
The most influential factor in the former development is migration. Out-
migration from the economically declining rural areas has continued throughout the
twentieth century and taxed the Welsh-speaking population, whereas the de-
industrialisation of the last decades has had English newcomers moving into Wales in
their thousands, attracted by the beautiful countryside and relatively cheap house
prices. The newcomers are often English-speaking families in search of a safer
environment to bring up their children, or senior citizens wishing to retire in the
countryside. Both in and out-migration continue to affect the current language
situation, and it is only this last wave of English speakers which has finally turned the
tables against Welsh in many areas.
Allan and Mooney (1998) have studied the language of counterurbanisation in
Scotland, stating that the in-migration has brought along certain terminology, which is
now an integral part of everyday language in the rural communities. Talking about
incomers, immigrants or settlers versus locals is also common in Wales, where the
clash can be further heightened by the language difference. Along with the mental
division of the community into two ethnic groups, the locals are increasingly
concerned that their indigenous language and culture will be taken over by the
dominant English culture, and consider the newcomers a threat to the ‘harmonious
rural life’ of the villages. Cloke, Goodwin and Milbourne (1995) find similar
experiences in the Welsh rural setting.

3
For comparison, the 2001 census also reveals that the number of all Gaelic
speakers in Scotland is only 58,652, i.e. less than the increase in Wales over the last decade
(see The Registrar General's 2001 Census Report to the Scottish Parliament:
[Link] [20.2.2003]).

4
Note that as a result of the 1996 reformation of the Welsh unitary authorities, the
counties of 1991 do not correspond to those of 2001, and figures from 1991 cannot be
compared directly with the fresh ones. Modern Gwynedd, for example, excludes Anglesey
(Ynys Môn), and there is a new county of Aberconwy and Colwyn between the former
Gwynedd and Clwyd. Ceredigion and Carmarthenshire, on the other hand, are parts of the
former Dyfed. The above observation, however, takes this change into account.
30

Aitchison and Carter (2000a: 111-123) divide Wales into five zones according to
the migration patterns evidenced in the 1991 census: the Welsh core, the borderland,
the transitional zone, the old industrial areas and the urban agglomerations. The
districts belonging to each zone are characterised by similar proportions of Welsh
speakers and similar numbers of in and out-migrants, changing the consistency of the
population in different ways.
The Welsh core receives a relatively high number of in-migrants from outside
Wales, which tests the endurance of these Welsh-speaking communities. The
borderlands on the other hand already share a stronger cultural and linguistic bond
with England than with the Welsh-speaking heartland. This is evidenced e.g. by the
relatively low percentage of Welsh-born population: up to 94 per cent of the in-
migrants in the border districts come from outside Wales (Aitchison & Carter 2000a:
119). The transitional zone is “somewhat ragged and ill-defined”; it “constitutes the
transition between the heartland and the border country” (op. cit.: 120). The
percentage of migrants from outside Wales is lower than in the two previous zones,
but there are fewer Welsh speakers among them than in the area itself, which suggests
that this language frontier is constantly threatened by an English invasion. The
industrial areas deviate significantly from any of the others: the percentage of Welsh-
born population is the highest of all, and the numbers of immigrants are the lowest.
Migration therefore does not appear to play a major role in language change. However,
there is little Welsh being spoken in these southeastern Valleys, and so, they might be
considered the core of the English-speaking Wales. The cities obviously attract the
greatest numbers of migrants, but Cardiff differs from Swansea and Newport in having
a higher percentage of in-migrants from Wales. There is actually a subtle increase in the
percentage of Welsh speakers in Cardiff because of the migrants, which is not
happening anywhere else in the country. Aitchison and Carter (op. cit.: 122-123)
conclude that the impact of migration “manifests itself in a pushing back of the
frontier between ‘Welsh Wales’ and Anglicised Wales on the one hand, and in serious
disruptions to the integrity of the heartland on the other”.

2.3. Wales as a bilingual nation: Welsh English in the society

Wales is both bilingual and bi-cultural. John Davies (1993: 109) refers to the
observation that the centuries long history of a dual perception of Welshness, of ‘inner’
and ‘outer’ Wales (or Cymru Gymraeg ‘Welsh Wales’ and Cymru-ddi-Gymraeg ‘Wales
without Welsh’, as e.g. in Jones 1980: 49), may have more significance in the minds of
the Welsh than the traditional division into north and south. The dual notion can be
said to originate from the early medieval separation of Pura Wallia and Marchia Wallie,
and it has persisted until present times, gathering considerable momentum with the
industrialization of the southeast. The existence of two different culture areas in Wales,
Inner Wales and Outer Wales, is first postulated by Bowen (1959 and 1964; reference in
Pryce 1978: 237-238), who bases the concept on the linguistic, cultural and geographic
patterns in Welsh history. Pryce summarises that the Inner Wales of the north and west
“has always drawn on its own resources, fostering and upholding Welsh culture traits”,
being able to maintain its identity until the arrival of English-medium mass
 31

communication. Outer Wales, on the other hand, has been easily penetrated by external
influences, making it culturally and linguistically more heterogenous. In his study, Pryce
follows the scholarly tradition of regarding “the incidence of spoken Welsh as a
surrogate measure of the distinctive culture that is Wales” (op. cit.: 236). The cultural
identity of Outer Wales is more difficult to define, which is perhaps why Pryce does not
tackle the issue.
The significance of the Welsh language for the Welsh culture is easy to
comprehend. Its struggle to regain the past prestige may even have increased its
cultural value. Its social value, however, has suffered blows from which it may not
recover equally well. Prior to the industrial era, the position of Welsh was very good on
the low level, as an everyday spoken language. Ever since the Acts of Union, however,
it had been poor on the high, institutional level. Although the English Law determined
that it had no official position in the society, in the eyes of the Welsh speakers their
language had a high prestige. In addition to being the language of worship, it had a
strong literary tradition reaching far back in history. Even as English gained popularity,
the indigenous Welsh culture was firmly rooted in the land. It is from these roots that
the current resurgence has also sprung.
The situation today is quite different: the support for the Welsh language is
strong on the high level, and it is often regarded as ‘more beautiful’ and ‘more lyrical’
than English by Welsh and English speakers alike, which is a sign of high prestige, as
pointed out by Lindsay (1993: 3). However, Welsh continues to lose ground on the low
level. Its use as an everyday community language between friends and family members
appears to be gradually declining, although not in every village; the speed and direction
of the process depend on the area (see Aitchison & Carter 2000a: 59-67).
There is no longer diglossia between the two languages, as Welsh has the
support of the authorities and most Welsh speakers are actively making use of their
rights. Traditionally Welsh was the socially inferior language, and generations of Welsh
speakers learned to feel ashamed of their mother tongue in front of the English
speakers, regardless of its glorious history. The enterprising English incomers had a
strong representation in the professional and manager-level jobs throughout the
country. Today, however, there is a great need for bilingual staff in administration,
media, education and other public services, and the division of labour has begun to
favour Welsh speakers.
Aitchison and Carter (2000a: 123-127) compare the three main ethnolinguistic
communities – bilingual Welsh, Anglo-Welsh and incomers born outside of Wales –
on the basis of the 1991 census figures. They observe that on the whole, bilinguals
today hold a greater percentage of higher class positions than monoglot English
speakers, who, correspondingly, are more likely to obtain a skilled manual, part-skilled
or unskilled job. However, it is still the non-Welsh incomers who take the majority of
the highest positions. The Anglo-Welsh, on the other hand, are at a disadvantage on
both accounts: they cannot compete with the bilinguals for jobs where Welsh skills are
required, but they also face strong competition from the incomers. The authors point
out that the situation varies according to the region: job opportunities are best for the
bilinguals in the anglicised southeast, around Cardiff area, whereas in the rural counties
most of the higher class positions go to the English-speaking incomers. The Anglo-
Welsh, again, draw the short straw both in the cities and in the countryside. Blackaby
32

& Drinkwater (1996: 168) have come to similar conclusions, stating that “for whatever
reason, Welsh-speakers appear to do better in the Welsh labour market than their non-
Welsh-speaking counterparts”. They also find that “the occupational advantage of
Welsh-speakers was most noticeable in areas where only a small minority of the
population were able to speak Welsh”, such as in the southeast of Wales.
Although there may no longer exist a division of functions between English and
Welsh, there are dialects of Welsh and dialects of English, all of which do not receive
the same level of regard. The main problem that the Welsh speakers face in this respect
concerns the differences between spoken and literary Welsh and the northern and
southern dialects. North Wales Welsh is typically considered more ‘deep’ and ‘pure’,
whereas South Wales Welsh has the tendency of borrowing English words more freely.
The use of ‘proper’ and ‘slangy’ Welsh is a regular topic of discussion for the Welsh
purists and both the North and the South Welsh confess to sometimes having
difficulty understanding the other dialect. The differences between the varieties have
been documented, e.g. by Awbery (1984) and Thomas (1992).
The problems of the Anglo-Welsh are of a more sensitive nature. How Welsh
are they exactly, not being able to speak ‘their own language’, Welsh? Particularly since
the founding of Plaid Cymru, the Party of Wales, in 1925, Welsh speakers have been
anxious to defend Welsh against the invasion of English. Smith (1984: 104) quotes the
words in which Saunders Lewis, the writer and inter-war President of Plaid Cymru,
reacted to the census figures of 1931.

Even in the Nationalist Party itself I fear that there are some who do not yet realise that
a ‘bilingual W ales’ is som ething to be feared and avoided, that the decrease in the
number of monoglot Welsh speakers is a tragedy, and that a Welsh-only Wales is alone
consistent with the aims and the philosophy of Welsh nationalism....We cannot therefore
aim at anything less than to annihilate English in Wales.... It is bad, and wholly bad, that
English is spoken in W ales. It must be deleted from the land called Wales: delenda est
Carthago.

Attitudes have softened, but the fact that English is the mother tongue of over
eighty per cent of the population has had little effect in making it a ‘language of Wales’.
Linguists find themselves tiptoeing around the topic of Welsh English. The cautious
attitude is reflected, e.g. in the title of the book English in Wales by Coupland and
Thomas (1990). As the authors are referring to the language of the born-and-bred
Welsh population, whether bi- or monolingual, the title might as well be ‘Welsh
English’. Penhallurick (1993: 31), however, points out that the choice is deliberate and
quotes the authors (Coupland & Thomas 1990: 2):

...the language question in Wales in sufficiently highly charged that some might infer that
even to pay analytic attention to English in W ales, or ‘W elsh English’...represents an
ideological position, perhaps even a form of capitulation, or collusion with forces
threatening the Welsh language.

Penhallurick moves on to argue that “English in Wales belongs to the Welsh

and is therefore Welsh English: nationalism cannot succeed, and perhaps the Welsh
language cannot prosper, without embracing the fact of Welsh English” (1993: 32).
 33

Welsh English is in need of national acknowledgement, but whether it will obtain it is

still an open question.
In spite of its low profile, WE appears to retain a measure of social and national
significance to its speakers. Numerous studies have been conducted on how the Welsh
varieties of English are perceived by the Welsh themselves. A large number of these
are reviewed in Giles (1990), and the general findings indicate some straightforward
patterns. In matched-guise experiments dealing with dialect convergence and
divergence, Anglo-Welsh listeners evaluated WE speakers diverging from the RP
interviewer as less intelligent but more kindhearted and trustworthy than speakers who
converged (e.g. Bourhis, Giles & Lambert 1975 and Bourhis 1977, summary in Giles,
op. cit.: 260-262). Therefore maintaining or emphasising one’s Welshness in intergroup
situations is in Giles’s (1990: 261) words “a valued tactic asserting cultural identity”.
The solidarity factor is significant for the maintenance of nonstandard speech forms,
as pointed out, e.g. by Coupland (1988: 93-94; cf. Labov’s ‘covert’ prestige 1972: 249-
250; §3.3). Integrative learners of Welsh also esteemed the Welsh accent as a symbol of
national identity, until such as time at least that it might be replaced by competence in
the Welsh language (Bourhis, Giles and Tajfel 1973; Giles et al. 1987, summary in Giles
1990: 263), which supports the notion that WE, as well as Welsh, possesses social and
cultural significance. A later study by Coupland et al. (1994), focusing on the dialect
evaluations of Welsh school teachers, showed that although southeast English accents
(e.g. Cambridge) were deemed highly prestigious, they received low evaluations on the
scales of pleasantness and dynamism. The WE varieties of Carmarthen and Merthyr
Tydfil, on the other hand, fared well in these respects, and Carmarthen and Newtown
received mid-level scores for prestige, as well. The research group consisted of both
bilingual Welsh speakers and non-Welsh speakers, and the Welsh speakers were found
to be less sensitive to the stigma attached to the heavily anglicised varieties of English,
evaluating them more positively.
The Welsh speakers’ reactions to WE have been found more inconsistent than
those of the Anglo-Welsh; while their attitudes towards WE were positive in the study
by Bourhis et al (1973), another investigation revealed that bilingual pre-adolescents in
Carmarthenshire considered WE speakers more snobbish, unintelligent, and/or selfish
than their Welsh or RP speaking counterparts. A proficient command of English was
valued more than any sort of national affinity expressed through WE (Price et al. 1983
and Giles & Johnson 1986, summary in Giles, op. cit.: 265-266).
English listeners’ reactions to WE have also been tested, and in these studies the
WE speaker was typically regarded as less suitable for socially high prestige positions,
less intelligent, less egalitarian, less ambitious, active, happy, and good-looking than the
RP speaker, although more religious (Giles, Wilson & Conway 1981, Giles & Marsh
1979 and Brown et al. 1985, all summarised in Giles, op. cit.: 270-271). In all fairness,
most British accents posited against RP would have produced a very similar result.
Giles (1990: 271) concludes that the prestige of WE, both inside and outside Wales is
low in terms of overall perceived ‘vitality’, which does not bode well for the future of
the variety, but he also points out that perceived vitality is dependent on the factors of
time, space and elicitation technique. Results gathered twenty or thirty years ago may
by now be outdated in some respects (cf. §8.1.1).
34

Cardiff, as a large conurbation, forms a dialectal and sociolinguistic area of its

own. In the studies by Coupland, Garrett and Williams (1994, 1999) it receives low
marks for perceived Welshness. In the 1970s, Bourhis and Giles (1976, also cited in
Giles, op. cit.: 268) conducted a matched-guise experiment on the English-speaking
and the bilingual Cardiffians’ behavioural responsiveness to broad and mild WE,
Welsh, and RP. They found that the English speakers responded equally favourably to
a request presented in mild WE and RP, whereas the broad WE request was not likely
to gain cooperation. The Welsh speakers responded to the Welsh language request,
whereas none of the English requests were very influential. The result goes to show
where the English and Welsh speakers’ in-group loyalties lie. The English-speaking
born-and-bread Cardiffians do not relate with North Wales. They are more likely to
consider themselves British than strongly Welsh, although English they do not confess
to be. Coupland (1988: 98) indicates that Cardiffians do not in fact hold their local
accent in very high regard at all; this is a trait shared by the speakers of other urban
British speech forms.
Those Welshmen whose families used to speak Welsh a few generations back
have perhaps been the sorest spot for the Welsh speakers. The coalmining valleys are
the areas which have received the strongest critique from the defendants of Welsh,
because according to the more aggressive views, not only did these people lose ‘their
language’, they did not learn to speak English properly, either. This sentiment is
expressed e.g. by Ned Thomas (1973: 116) in The Welsh Extremist, one of the
spearheads of the Welsh language activists of the time:

As the educated Welsh-speaker looks at the new affluent working-class of South Wales
he is bound to see people who have lost a culture and gained only a higher standard of
living, people made particularly vulnerable to commercially-fostered pseudo-values by
their own rootlessness, people who have lost the dignity of a language and acquired a
despised and comic dialect of English.

The English dialects of South Wales, particularly those of the industrial Valleys,
have indeed been regarded as comic. This is evidenced, e.g. by the popularity of John
Edwards’s best-selling ‘Wenglish’ books, Talk Tidy (1985) and More Talk Tidy (1986),
which present an entertaining albeit exaggerated picture of the local dialect (see
Williams 2003 on the accuracy issue). Penhallurick (1993: 32-33), in fact, finds the Talk
Tidy books damaging to WE, as they inadvertently turn the genuine speech form of
the region into something as trivial and laughable as ‘Wenglish’. What is perhaps
overlooked in observations such as those of Thomas or Edwards is that no speech
form is without a history. The past forces moulding the Valley dialects, as well as other
varieties of WE, will be examined in chapter 8 of the present study.
It is plain that the Welsh speakers’ national identity rests heavily on the
language. For the monoglot English people of Wales, the matter is more complicated.
The recent resurgence of Welsh has perhaps placed more emphasis on national identity
being dependent on language, but a counter-reaction can also be detected. Professor
Heini Gruffudd, from the department of Welsh at the University of Wales, Swansea,
implies that the Welsh identity is becoming less dependent on the Welsh language (p.
c., November 1999). As an example he mentions that whereas Anglo-Welsh literature
 35

used to form a threat to the Welsh language literature, and perhaps it still does, the
attitudes of the Anglo-Welsh writers have changed. Previously they might have turned
their backs on Welsh, having ‘lost’ the language themselves, but today they are often
very nationalistic: it has become acceptable to defend Wales in English. Plaid Cymru
has also moderated its official stand on the language issue. Instead of stressing the
Welsh language agenda, the party now strives to gain the support of the whole of
Wales. Individual party members’ opinions are likely to vary on this matter.
There is no denying that the language of the Welsh speakers ties them directly
to Wales, whereas the English speakers connect linguistically not only to Wales or
Britain but to the entire English-speaking world, most notably today to America.
Thomas’s (1973: 105-107) words ring true to an extent when he describes the Welsh-
speaking communities as more coherent, organised, focused and culturally aware
compared to the Anglo-Welsh, whose identity is necessarily more problematic and who
find it difficult to form a national language community because the connotations of
their language pull them towards England and the social, cultural and historical aspects
that England entails. This view may be dated and biassed, but not one which can be
dismissed entirely, even at the beginning of the twenty-first century.
The picture which emerges of the sociolinguistics of Wales is a complex one,
with the fates of Welsh and English intertwined. Welsh as the indigenous language has
the moral upper hand today and the monolingual English speakers are left to defend
their right to speak their mother tongue without having to feel less Welsh for it. In the
light of history, the situation is ironic, to say the least. Yet, the Welsh speakers are
guardians of a threatened language, which justifies their tenacity in defending it and
even attacking the dominant English language culture. There is some tension between
the two language communities, as both the Welsh speakers and the monoglot English
speakers can be quite sensitive to the language issue.