Move back one step  Move forward one step  Show themes of website
Larger font Smaller font Default font
  Most recent changes

Current innovations in advanced Dublin English
    Short Front Vowel Lowering
    BACK-retraction
    Ejectives / preaspiration
    GOOSE-fronting
    Syllable-initial affrication

Possible external influences
    American English: California in Dublin?
    British English: cultural attitudes and language

Analysing change and non-change
    Non-participation in change
    The ethnography of variation and change
    Internal arguments for change
    Whither Dublin English?

References

Current innovations in advanced Dublin English

Short Front Vowel Lowering


The major item of sound change in present-day Dublin English is undoubtedly the lowering of short front vowels. This is to be found with the DRESS and the TRAP vowels in particular, with the KIT vowel trailing behind except where it follows /r/ or /l/ as in lid [led] or riddle [redḷ]. Similar lowering is known from varieties of English in North America, e.g. in Canada (Boberg 2005) and in California (Eckert 2008). The question of what triggered this does not have a simple answer, but for the recent data collection the aim was to record speakers who might have this lowering.
   As with the changes of the 1990s, it was obvious from the very beginning that young female non-vernacular speakers are the vanguard of this change. None of the males recorded either in 2009 or 2011 had short front vowel lowering to any appreciable extent.

  DRESS vowel with slight lowering
  DRESS vowel with moderate lowering
  DRESS vowel with extreme lowering
  DRESS vowel with three degrees of lowering in succession

Note. In the above spectrogram F1 and F2 are progressively closer to each other from left to right due to the raising of F1; this is a clear indication of increased lowering of the DRESS vowel from left to right for the three instances from the above three speakers (slight, moderate and extreme respectively).


BACK-retraction


In connection with the lowering of DRESS one should see the current development in recent non-local Dublin English whereby the TRAP vowel is retracted before velars and these are then articulated further back than has hitherto been the case. With advanced speakers this retraction can be almost to a uvular stop with the vowel retracted as well, i.e. back = [bɑq]. This development is counter to the widespread realisation of back which in mainstream Irish English has a forward realisation of /k/ with a movement of the vowel forward as well, i.e. back = [bæk-]. See remarks on ejectives in the next section.

  TRAP vowel with young female speaker from Dublin
  Retraction of TRAP before voiceless velar stop with three non-local female Dublin speakers
  Lack of retraction of TRAP before voiceless velar stop with local female Dublin speaker


Ejectives / preaspiration


Ejectives are formed by a closure of the glottis (the space between the vocal cords) at some point in the supra-glottal tract, for instance, the velum. Given the fact that the glottis is closed, ejectives are voiceless segments. During the articulation of an ejective the glottis is raised somewhat increasing the pressure inside the area between the glottis and the supra-glottal closure. When an ejective is released there is a sudden burst of the air trapped during the articulation. Ejectives thus have a sharper burst than aspirated consonants which show a pulmonic pulse of air on the release of the voiceless stop. There are languages around the world which have phonemic ejectives, e.g. many Caucasian languages such as Georgian, languages of the Pacific Northwest, native languages of South America and some Afro-Asiatic languages of north-central Africa. In English ejectives are rare though some speakers do have them as an idiosyncratic feature as in this moderate example and this extreme example, both incidentally from the same speaker of southern British English. As can be seen from the following spectrogram, there is a considerable intensity peak after the release of the [k] in the second example.

The situation in Irish English is interesting from this perspective. Firstly, many advanced speakers of Dublin English have a central or slightly retracted realisation of the TRAP vowel. This is apparent in the reading style pronunciation of the word BACK (see Word List (2011) in the Recordings module). The pronunciation of BACK is accompanied by a retracted /k/, something like [k-] or even [q]. There would appear to be at least glottal reinforcement for this stop and the release is longer and stronger than that for conservative Dublin speakers (see second instance of BACK in the following spectrogram). What is not present in the advanced pronunciations is a raising of the glottis with increased pressure between glottis and [k-] / [q]. It is not there yet, although advanced pronunciations are moving this way now (late 2011). So the question for future monitoring is whether final /k/ will develop into an ejective, remain where it is, or reverse to a simple aspirated stop. At any rate the embryonic ejective quality of [k-] / [q] is not an idiosyncratic feature of a few individuals but a characteristic of advanced speech with in the community of young non-vernacular females in Dublin at present.
   Note further that the first pronunciation in the following sound file shows slight pre-aspiration before the stop. This has been observed with a number of speakers who terminate the vowel articulation before the closure for the final stop in BACK.

  BACK – advanced female speaker (left) and older conservative (right)


GOOSE-fronting


This term refers to the pronunciation of the vowel in the GOOSE lexical ever further forward in the mouth. A degree of GOOSE-fronting has always been typical of Dublin English and is salient in words which have a short vowel in many other varieties of Irish English, e.g. book [bu:˖k]. In advanced Dublin English among young females the fronting can be quite extreme.

  GOOSE-fronting with young female Dublin speaker – 1
  GOOSE-fronting with young female Dublin speaker – 2

N.B.: GOOSE-fronting does not affect the realisation of the KIT vowel and has not provoked any downward shift of the latter vowel.


Syllable-initial affrication


Among some female speakers of advanced Dublin English one can ascertain a slight affrication of stops in initial position. This is unexpected, given that the lenition which is so common in Irish English is confined to post-vocalic positions. The affrication, which can be heard in the following sound file, is not universal among young females, but it does occur often enough to be viewed as a non-idiosyncratic feature. I noticed this when doing recordings in the early 2000s and in terms of its acoustic prominence it does not seemed to have changed.
   What is remarkable is that this affrication, like strong GOAT-diphthongisation, is definitely a gender-specific phenomenon. None of the males recorded at any time had this feature.

  Slight affrication of /d/ with young female advanced speaker


Possible external influences

American English: California in Dublin?


Short front vowel lowering is known from a number of anglophone locations around the world. One of these is Canada (Clarke, Elms and Youssef 1995; Boberg 2005, 2010) and the other is California. An influence of Canadian English on advanced Dublin English can be excluded as the Irish have virtually no exposure to English from Canada, either through speaker contact or the media.
   The situation with California is not quite as straightforward. It is true that there are not large numbers of Californians in Dublin or Ireland so speaker contact can again be ruled out. However, young Irish people do hear many American accents through the media. Because Irish television buys in many American-produced films there has also been considerable exposure to American accents since at least the mid twentieth century. In the past few years there has been additional exposure due to soap operas from America being shown on Irish television. By far and away the most popular among young female teenagers is Glee a soap opera aimed at precisely this sector of the English-speaking population, first and foremost in the United States but also in other anglophone countries where the programme is broadcast.
   The question of possible Californian influence is not straightforward. True there are, and have been, some actors in certain episodes of this soap who have Californian accents. But not all actors have the (northern) Californian vowel shift, as described by Penelope Eckert, and many of them are from other parts of the United States. Furthermore, there is much singing in this soap, so the amount of speech with a possible Californian accent is actually smaller than would normally be the case in programmes of equal length.
   In addition there is the general question of an influence of accents in the media on spoken language. I have argued that this was a means of transmission for the vowel shift of the 1990s. But the situation there was different: virtually all the announcers and newscasters on Irish television had the new pronunciation in the early 1990s. It was also obvious to young people around Ireland that the new accent heard on television was a fashionable Dublin accent and hence worthy of emulation.
   Coming back to possible Californian influence on present-day advanced Dublin English consider the following points.


Similarities

Advanced Dublin English shares lowering of /e/ after /r/ with California, e.g. dress [drɛ˕s], friend [frɛ˕d]. In Dublin this lowering is also salient after /l/, e.g. left [lɛ˕ft].

  Young female Dublin speaker saying ‘on the left’ (echoing the phrase by the author who says it without lowering)

Advanced Dublin English among young females has considerable GOOSE fronting.

Many advanced Dublin English speakers have the MERRY–MARY merger (vowel length is not distinctive before a syllable-coda /r/ for these speakers) though the MERRY–MARRY merger is not found (vowel height differences are maintained before a syllable-coda /r/).


Differences

There is no centralisation of the PRICE vowel (this would effectively lead to (near-)homophony with the local Dublin pronunciation and hence is out of the question).

The STRUT-vowel, as in cut [kʌt], has not shifted to any noticeable extent at all during the past 20 years in Irish English, i.e. it has been unaffected by the vowel shift of the 1990s and is not affected by the present short front vowel lowering. It remains a central-to-back short mid vowel, occasionally with slight rounding. A slight fronting of this vowel (without any rounding) is found with some south Dublin speakers, but this is indicative of a ‘posh’ accent rather than a ‘cool’ or ‘trendy’ accent.

  Somewhat fronted and unrounded realisation of the STRUT-vowel

There are a number of other widespread features of American English which the Irish do not have, e.g. the COT-CAUGHT merger, pre-nasal vowel tensing as in band [bɛ:n(d)], and [ɛ:n(d)] (stressed), though vernacular varieties do have pre-rhotic tensing and older speakers, especially in rural areas, have pre-lateral tensing as in calf [kɛ:f].

Most importantly, one should point out that the short front vowel lowering of advanced Dublin English is not a chain shift, this distinguishes is from the (northern) Californian vowel shift, the Northern Cities Shift of the Inland North of the United States and the Southern Shift in the southern United States. For instance, the realisation of the STRUT, LOT and FOOT vowels have not been affected by short front vowel lowering, though they may be at a future date.


British English: cultural attitudes and language


Irish English is impervious to ongoing change in British English. Features such as TH-fronting, T-glottalisation (Fabricius 2000), L-vocalisation (though there is L-velarisation, R-labialisation (Trudgill 1988, Foulkes and Doherty 2000) are not found in advanced local Dublin English at all. HAPPY-tensing is not an issue in Irish English as it always occurs.

The most important difference between non-local Irish English and non-local southern British English is that the former is rhotic but the latter is not. Furthermore, such distinctions as that between /æ/ and /ɑ:/ in Received Pronunciation, as in grand [grænd] and grant [grɑ:nt], play no role in Irish English. It is true that there is variation among low vowels – and many of these can be heard in the sound files on this website – but this is not systemic and lies between [æ(:)] and [a(:)] with unconditional retraction to [ɑ:] a complete non-starter in Ireland because it sounds too ‘English’.


Analysing change and non-change

Non-participation in change


Sociolinguistic identity can be construed by not participating in change. And it can be broken down into several smaller sections. Take, for instance, the following situation which I observed while working on A Sound Atlas of Irish English. When this data was being collected (early 2000s) the changes in Dublin English were already fully underway, indeed established for most young people, certainly for females. In a sports club in Carlow (a town about 80km from Dublin) I observed the social and linguistic behaviour of a group of young males, playing hurling, a traditional Irish game (with a wooden stick and a small leather ball) which for many locally rooted individuals in towns and in the countryside is an Irish alternative to playing soccer. Hurling does not have an international following and, importantly, it does not have any of the glamour and money attached to soccer. I observed that none of the players of hurling had any aspect of the new pronunciation: they had no GOAT-diphthongisation, no MOUTH-fronting, no GOOSE-fronting, no L-velarisation and no R-retroflexion (see the sound file given below). And yet they would have been in constant contact with female co-evals who would have had just these features as part of the new pronunciation for non-local speakers.
   The point here is that by not participating in the ongoing change around them these males were projecting their identity as locally rooted individuals.

  Alveolar /l/ with local Carlow speaker
  Lack of AU-fronting with local Carlow speaker
  Lack of R-retroflexion with local Carlow speaker
  Sample sentences with local Carlow speaker


The ethnography of variation and change


There is no doubt but that speakers participate in variation present in their community and engage in this variation to project their social identity, i.e. this variation is an integral part of how individuals present themselves in their community. For instance, among the 106 individuals which I recorded in 2011 it was clearly predictable which of them had short front vowel lowering along with other features of advanced Dublin English such as GOAT-diphthongisation and NORTH-raising. If a young female was stylishly dressed, wearing makeup, perhaps with dyed or tinged hair, with a smartphone and a fashionable handbag then that individual invariably showed the features of advanced Dublin English (or had a switch-over accent). If, on the other hand, a young female was plainly dressed, no makeup, not obviously concerned with outward appearances then that individual would have less advanced features, e.g. slight GOAT-diphthongisation, moderate NORTH-raising and virtually no DRESS-lowering. If such a speaker came from the North Side of Dublin then that individual would be much more likely to use the vernacular accent of Dublin and show T-glottalisation, low rhoticity, lack of CHOICE- or NORTH-raising, etc.
   A frequent feature which correlates with advanced pronunciation in Dublin English is the presence of high rising terminals in declarative sentences (statements). This practice is commonly referred to as ‘uptalk’. Other aspects of more general youth language across the anglophone world, such as the use of quotative like or of augmentative lexical items such as awesome, are equally common in Ireland.

  High Rising Terminal (young female speaker)


Internal arguments for change


If one excludes an external influence, from either America or Britain, on the development of Dublin English in the past two decades (and by extension of Irish English in general) then the motivation for change must have been internal to Dublin. A closer inspection shows that reasons can be put forward for developments which, by coincidence, have parallels elsewhere in the anglophone world.


GOAT-diphthongisation

An obvious similarity between advanced Dublin English and non-vernacular varieties of Southern British English is that the vowel in words like home, road, go is a diphthong with a central starting point and a movement towards a high back vowel, i.e. [əʊ]. Given that this pronunciation in Dublin is a recent development it might be thought that the Irish simply adopted the English pronunciation. However, there are cogent arguments against this view.
   First of all GOAT-diphthongisation in England is part of the long series of vowel changes known as the Great Vowel Shift which began in the late Middle Ages and which led to the original vowel in FACE, /a:/, becoming /e:/ and that in GOAT, /ɔ:/, becoming /o:/ by the eighteenth century. Both these vowels were then diphthongised (rather than further raised to /i:/ and /u:/ respectively). This diphthongisation has been continued in vernacular varieties of London English.


Diphthongisation of FACE and GOAT in southern England in 19th and 20th centuries

In Ireland, the recent diphthongisation of the GOAT vowel is different. Firstly, it bears no relationship to the development of the FACE vowel. Irish English adopted in the late modern period (18th and early 19th centuries) the English pronunciation of FACE with /e:/ replacing the earlier pronunciation of this vowel in Ireland as [a:]. Since then the FACE vowel has remained a monophthong [e:] and shows no sign of changing.
   Secondly, the diphthongisation of GOAT correlates with the degree of raising for the THOUGHT, NORTH and CHOICE vowels: the more the latter are raised, the greater the degree of diphthongisation (see the section on Implicational Scales). This clearly implies that the diphthogisation was provoked by the raising of the vowel (or diphthong onset) in THOUGHT, NORTH and CHOICE respectively. These always had open realisations in Irish English and the new pronunciation of the 1990s was a reaction to the traditional pronunciations. From investigations of the 1990s developments it is know that the raising of the low vowel realisations was the beginning of the original vowel shift. The conclusion from this is that GOAT-diphthongisation was a reaction in the vowel space of speakers to the raised realisations of THOUGHT, NORTH and CHOICE.

Diphthongisation of GOAT in Dublin in the 1990s and early 2000s


Short front vowel lowering

Although this is a common feature of many varieties of English in North America it can be interpreted as an internal development in Ireland. Consider that the movement of the TRAP down and back did not interfere with any other sounds in this area. The downward movement of the TRAP vowel vacated space which was then occupied by the lowering of the DRESS vowel. Furthermore, at present (late 2011) the KIT vowel has not been lowered to any significant extent.

But what speaks most clearly for an internal motivation of short front vowel lowering in Dublin English is (i) the fact that it is not a full chain shift (the LOT, STRUT and FOOT vowels are unaffected by it) and (ii) its apparent triggering by vowel lowering in the environment of /l/ or /r/. The lowering is most obvious for the mid short front vowel in rhotic and lateral environments, e.g. dress [drɛ˕s], left [lɛ˕ft], sell [sɛ˕ɫ]. Recall that the velarised [ɫ] of Dublin English is preceded by a low-central onglide (and not a high-back onglide as in varieties with L-velarisation in England).

  SELL with considerable lowering (non-local speaker)

For speakers of advanced Dublin English, /l/ is also somewhat velarised in syllable-initial position (but not in other non-vernacular varieties which have syllable-final /l/-velarisation). Even KIT-lowering, which is not very evident in Dublin, does occur when the high front vowel follows /l/ or /r/, e.g. lid [led], rip [rep].

An interesting question in this context concerns the trigger for the entire process of short front vowel lowering. From the remarks in the previous paragraph one might think that it occurred first in post-rhotic and post-lateral position and then was extended to other phonotactic environments. However, the position of the TRAP vowel seems to suggest that at least there has been another movement which has been running parallel to the post-rhotic and post-lateral lowering. The recordings from late 2011 suggest that for all speakers who have any significant lowering, the TRAP vowel when it occurs in an environment other than after /r/ shows lowering with retraction. Importantly, this applies to speakers who only have DRESS-lowering in post-rhotic and post-lateral positions, i.e. they would have trap [trap] and back [bak] along with dress [drɛ˕s], but not neck [nɛ˕k].


  Whither Dublin English?


The innovations which appeared in Dublin English in the early 1990s became established as change and then spread outside the capital and have since determined the shape of supraregional Irish English for all young females and many males born after the mid 1980s. These changes are clearly manifest in the raised realisations of back vowels, R-retroflexion, L-velarisation, MOUTH-fronting, to mention the most prominent features.


The fate of recent changes

Some of the changes, such as AI-retraction, were short lived and had disappeared from general non-vernacular Dublin English by the early 2000s. But other innovations have occurred since then and these may well become established as items of linguistic change in the near future. The most likely innovation to become established is short front vowel lowering which while it is still largely confined to rhotic and lateral environments for the DRESS vowel is not conditioned for the TRAP vowel, consider the [bak] / [bɑq] pronunciations which are now widespread. The fact that the TRAP vowel is free of the rhotic/lateral environment would imply that the lowering began with this and that its movement had a drag effect on the DRESS and KIT vowels. This assumption is supported by the fact that some speakers already have unconditioned lowering of the DRESS vowel but none of the 106 individuals recorded in 2011 had unconditioned lowering of the KIT vowel.


Short front vowel lowering and phonological space

TRAP-lowering involves not just a shift down from [æ] to [a] but often a retraction to [a-] / [ɑ] which is why this movement is referred to, in its unconditioned form, as BACK-retraction. However, there is no sign (as yet) of a merger with the LOT vowel. This is a low back rounded vowel in all non-vernacular forms of Dublin (and Irish) English although it is unrounded in local Dublin English. But given that the TRAP vowel has [æ] in the Dublin vernacular there is no merger here either. This means that in all varieties of Dublin English a triad of words such as sacks : socks : sucks shows no mergers although the realisations vary along the vernacular – non-vernacular continuum.

Consider the following sound file from a speaker of advanced Dublin English. It is very obvious that the DRESS vowel is lowered, that the BACK vowel is retracted from [æ] to almost [ɑ] and that the DIP vowel is slightly lowered. However, the LOT vowel and the BUS (i.e. STRUT) vowel show no shift when compared to realisations by other non-vernacular speakers whose speech is not as advanced in terms of short front vowel lowering.

  DIP DRESS BACK LOT BUS (advanced speaker)

If DRESS lowering and BACK retraction are the main movements of the current shift then one might expect them to have a push effect on the LOT vowel and on the STRUT vowel. Indeed one might surmise that, in the manner of a true chain shift, the LOT vowel would be pushed upwards, given that the BACK vowel is acoustically quite close to the LOT vowel for advanced speakers (compare the similar F1 contour in the above spectrogram).
    But there are external sociolinguistic arguments militating against this. A raising of the LOT vowel would make it sound like the raised START vowel which was found in the 1990s and which is still occasionally still ridiculed as ‘Dortspeak’ (for ‘Dartspeak’) and occurs in negative comments about posh, snobbish South Side Dublin accents, e.g. when complaining about people who say ‘Gords’ for ‘Guards’ (the Irish police force) or ‘orts’ for ‘arts’ and the like.
    There is a similar external argument against a possible raising of the STRUT vowel due to a push from below within the vowel envelope of advanced Dublin English speakers: a raised vowel would immediate suggest a local Dublin English accent so such a pronunciation would definitely be taboo, just as a centralisation of the PRICE vowel would be.


Innovations and linguistic change

It is obvious that not all innovations will become instances of linguistic change. Milroy and Milroy (1981: 381) point out that one about 1 in 10 innovations will spread throughout an entire community and become a change. A similar estimate might hold for varieties of English in the Republic of Ireland. For instance, the syllable-initial affrication found with many females in advanced Dublin English does not show signs of either spreading to all females or crossing the gender divide.


Who act as innovators in advanced Dublin English?

In classical social network analysis (as presented, for instance, in L. Milroy 1987) dissemination of embryonic linguistic change in a community is taken to occur via innovators: ‘innovators tend to be marginal individuals at the edges of networks who diffuse innovation via weak ties with others, the persons whom investigators actually identify as being strongly associated with a change are most probably the more socially central early adopters’ Milroy and Milroy (1981: 381f.).

But the type of innovation observable in present-day Dublin English is not triggered by marginal individuals but rather arises through inherent variation within subgroups in society, in this case among young females who vie with each other for status as ‘trendy’ and ‘cool’. To begin with the inherent variation would occupy the initial stages of trajectories which show phonological regularity, e.g. back vowel raising, short front vowel lowering, increasing rhoticity, etc. Young females unconsciously realise the direction of these trajectories and would move slightly along them, pushing the movements to be at the vanguard of the innovation, e.g. showing slightly more short front vowel lowering than others to highlight their status as prominent members of their subgroup. This behaviour incites others to move along the trajectories contained in the inherent linguistic variation within the subgroup, here that of young non-local females in Dublin.

Whether innovations become instances of linguistic change depends essentially on whether (i) the innovations spread to the entire subgroup in which they are to be found – this would seem to be the case for short front vowel lowering but not for syllable-initial affrication, for instance – and (ii) whether the innovations then disseminate further into other subgroups to encompass the entire speech community, here that of all young non-vernacular Dubliners. This happened with the vowel shifts and other changes of the 1990s and it remains to be seen whether it will also apply to the most recent changes recorded in Dublin.


Innovators and different kinds of adopters

The distinction between innovator and adopter applies to current advanced Dublin English: here one can recognise both groups. The innovators are a small number of central members among young females who are active in short front vowel lowering. The adopters, as their name implies, adopt the pronunciations they hear from the innovators. At an early stage of a possible change one can distinguish the different sub-groups, though later when the change is complete or nearly complete, i.e. as far up the S-curve as the change goes, then it it not possible to recognise different types of agents anymore.

Given that the innovators are a small pro-active subset, how can one sub-divide the rest of the group? In my opinion there is a tertiary division into (i) early adopters, (ii) the mainstream and (iii) late adopters. The large mainstream lacks momentum and only engages in change when it is obvious that this is establishing itself. Both the early and the late adopters show linguistic insecurity and are concerned to involve themselves in embryonic change. However, the difference between the two subgroups is that the early adopters are more adventurous and immediately engage in change while the late adopters only do so after the mainstream has decided to go with it. For the early adopters, it is a gamble: maybe the change will not be accepted by the mainstream and they will have to retreat later on. But retreating from a change is actually not that unusual: recall the AI-retraction of the 1990s which was abadoned later on and now is very rare.

To illustrate the conception of the different subgroups consider the following metaphor: A linguistic change is like a train journey. The innovators own the train and run it, they have decided the direction it is taking. The others see the train and consider whether or not to travel on it. The early adopters are adventurous and jump on. Being the first on, they get the best places. Then the mainstream follow and take nearly all the other places, filling up the train. The early adopters have clout within the group because they have the best places, so the gamble has paid off. The late adopters are the last on. They have taken least risk and waited to be certain about the change. This matches their personal behavioural pattern, but they have no clout among the train passengers and get the worst seats.

The gamble pays off for the early adopters if the mainstream gets on the train too. If they do not, then the early adopters have to get off again as the train does not have enough passengers to travel and there is no journey, i.e. the linguistic change, signified metaphorically by the train journey, does not take place. What one is left with is variation (the prospect of a journey) but in the end everyone stays put (no established change).

One could expand this metaphor: the innovators might run several trains with different journeys (representing different items of variation). The early adopters might decide to travel on each train, whereas the mainstream might say: we’ll only take one or two journeys (so not every item of variation becomes established).

A journey could involve various distances (the relative distance from the starting point corresponding to the point on a trajectory for scalar features). Some passengers might go for part of the journey and others the whole distance. The length of a journey would correspond to the length of a trajectory in the mean vowel space of speakers in a group.

The distinction between innovators and early adopters is a valid one, consider the situation in Dublin in the early 1990s again. There were people in Dublin who had introduced the vowel shift upwards along a back trajectory. These were the innovators. The early adopters were those who recognised this and decided to go with the new pronunciation. Importantly, outside Dublin there were early adopters who quickly picked up the new pronunciation from television and radio. These individuals then diffused the innovation in their own environments. That way the change spread quickly outside the capital.


References


Boberg, Charles 2005. ‘The Canadian shift in Montreal’, Language Variation and Change 17.2: 133-154.

Boberg, Charles 2010. The English Language in Canada : Status, History, and Comparative Analysis. Cambridge ; New York: Cambridge University Press.

Cheshire, Jenny, Sue Fox, Paul Kerswill and Eivind Torgersen 2008. ‘Ethnicity as the motor of dialect change: Innovation and levelling in London’, Sociolinguistica 22: 1-23.

Clarke, Sandra, Ford Elms and Amani Youssef 1995. ‘The third dialect of English: Some Canadian evidence’, Language Variation and Change 7.2: 209-228.

Di Paolo, Marianna and Malcah Yaeger-Dror 2011. Sociophonetics. A Student’s Guide. London: Routledge.

Eckert, Penelope 2004. ‘Adolescent language’, in: Edward Finegan and John Rickford (eds) Language in the USA. Cambridge: Cambridge University Press, pp. 361-374.

Eckert, Penelope 2008. ‘Ethnography and linguistic variation’, in: Nikolas Coupland and Adam Jaworski (eds) The New Sociolinguistics Reader. Basingstoke: Palgrave Macmillan, pp. 136-151.

Fabricius, Anne 2002. ‘Ongoing change in modern RP: Evidence for the disappearing stigma of t-glottalling’, English World-Wide 23: 115-136.

Foulkes, Paul and Gerard J. Doherty 2000. ‘Another chapter in the story of /r/: “Labio-dental” variants in British English’, Journal of Sociolinguistics 4.1: 30-59.

Foulkes, Paul and Gerard J. Doherty 2006. ‘The social life of phonetics and phonology’, Journal of Phonetics 34: 409-438.

Hickey, Raymond 1999. ‘Dublin English: Current changes and their motivation’, in: Paul Foulkes and Gerry Docherty (eds) Urban voices. London: Edward Arnold, pp. 265-281.

Labov, William, Sharon Ash and Charles Boberg 2006. The Atlas of North American English. Berlin: Mouton de Gruyter.

Ladefoged, Peter 2003. Phonetic Data Analysis. An Introduction to Fieldwork and Instrumental Techniques. Malden, MA: Blackwell.

Milroy, James and Lesley Milroy 1985 ‘Linguistic change, social network and speaker innovation’, Journal of Linguistics 21.2: 339-384.

Milroy, Lesley 1987 [1980]. Language and Social Networks. Second edition. Oxford: Blackwell.

Preston, Dennis R. and Nancy A. Niedzielski (eds) 2011. A Reader in Sociophonetics. Berlin: Mouton de Gruyter.

Thomas, Erik R. 2011. Sociophonetics. An Introduction. Basingstoke: Palgrave Macmillan.

Trudgill, Peter 1988. ‘Norwich revisited’, English World-Wide 9: 33-49.

Van Hout, Roeland and Hans Van de Velde 2001. ‘Patterns of /r/ variation’, in: Hans Van de Velde and Roeland Van Hout (eds) R-atics: Sociolinguistic, Phonetic and Phonological Characteristics of /r/. Special issue of Études et Travaux 4, pp. 1-10.