The changing accents of British English

« previous post | next post »

King’s English and Cockney replaced by three new accents, study finds

Britons depart from overtly class-based post-war speech epitomised by either clipped vowels or working-class dialects

By Charles Hymas, The Telegraph, Home Affairs Editor 

I vaguely recall an earlier study from about ten years ago that came to similar conclusions (including the emergence of a "multicultural" accent).  It's not surprising that differences would gradually diminish, especially under the influence of enhanced, pervasive mass communications and increased population mobility.

What we see, though, is that, as the older, established accents wither away, new ones arise among various shifting cultural, ethnic, and social regroupings.

Remember the Valley Girl accent, which people used to talk about a lot ten or twenty years ago?  Where is it now?

The King’s English and Cockney have all but disappeared among young people as three new accents have emerged, research has found.

Voice analysis found the two accents had been overtaken by standard southern British English (SSBE), as articulated by Ellie Goulding; estuary English, as spoken by Adele; and multicultural London English, as voiced by Stormzy.

The research, using computer algorithms to analyse voices of adults aged 18 to 33, shows how far Britons have departed from the overtly class-based post-war accents epitomised by the clipped vowels in the King’s English of BBC presenters and the working-class cockney of film stars such as Michael Caine and Barbara Windsor.

The voices of nearly 200 people from across south-east England and London were recorded, then analysed by a specially designed algorithm that listened to how they spoke and grouped them by how similarly they pronounced vowels in different words.

Around 26 per cent of those surveyed by the researchers at Essex University spoke estuary English, which has similarities with Cockney, but is more muted and closer to received pronunciation, euphemistically known as the King’s or Queen’s English.

Estuary English speakers pronounce words such as “house” more like “hahs”, but not as extreme as found in Cockney.

The accent, so named because it has extended out along the Thames estuary, is spoken across the South East, but particularly in parts of Essex, and is similar to how Stacey Dooley, Olly Murs, Adele or Jay Blades speak.

SSBE, which is typically perceived as a prestigious, “standard”- or “neutral”-sounding accent, is a modern, updated version of received pronunciation.

SSBE speakers comprised 49 per cent of the sample analysed by the researchers, led by Dr Amanda Cole, a lecturer in language and linguistics at Essex University.

Such speakers tend to say words like “goose” with the tongue further forward in the mouth – sounding a bit more like “geese” – than what would be expected in received pronunciation, according to Dr Cole.

This drift away from received pronunciation was evident in the changing accent of Queen Elizabeth II over her lifetime, spanning the changing diction of the nation from post-war Britain to the birth of the 21st century.

I find it particularly fascinating that the changing fashions of British accents could be traced in the person of Queen Elizabeth II over the course of her long life and rule.

A feature of SSBE and estuary English is the glottal stop. In estuary English, it replaces the ‘t’ in wa’er (water), be’er (better) and ci’y (city). But in SSBE, it only replaces the ‘t’ at the end of the word, such as wha’ (what), or bu’ (but).

Speakers include Ellie Goulding, Josh Widdicombe and potentially even Prince Harry.

Around 25 per cent of the young people used the third accent, multicultural London English.

That's a lot!  It says something highly significant about the diversifying demography of England.

They pronounced the vowels in words like bate and boat with the tongue starting at a point higher up in the mouth compared to standard southern British English, so that they might sound a little bit more like “beht” and “boht”.

They tended to be Asian British or black British and many were from London, but there were also people from across the South East who spoke with elements of a multicultural London English accent.

England footballer Bukayo Saka, and rappers Little Simz and Stormzy are examples of people with these features in their speech.

However, gone from among those surveyed was any measurable evidence of the King’s English or Cockney. “Cockney, the working-class London accent of Barbara Windsor or Michael Caine, and received pronunciation, which some call Queen’s English (or perhaps now King’s English), did not appear in our analysis,” said Dr Cole.

Truly amazing that they are gone, since they were the two quintessential types of British English we knew for the past century or so.

“That’s not to say that there aren’t any young people in our sample who might have spoken these accents but, if so, they were too few and far between for the algorithm to identify.”

Dr Cole said the shift in accents was a result of increased movement of people, resulting in greater contact between dialects, the growth of universal education and literacy and people buying into the idea that there is a “correct” or “standard” way of speaking.

“Standard southern British English and estuary English are not as different from each other as Cockney and received pronunciation. This could be evidence of what’s known as dialect levelling – where young people from different parts of the region now speak more similarly to each other than their parents or grandparents did,” she said.

Speakers of standard southern British English and estuary English generally tended to be white British, and women were more likely than men to speak the former, which is closer to the King’s English.

“It’s not surprising to find that women speak in a more socially prestigious way, as much previous research suggests women are often more chastised for speaking with regional accents than men,” said Dr Cole.

“Attempting to prevent accents from changing is like sweeping back an incoming tide with a broom – fruitless and defying nature. Instead, we should embrace linguistic diversity, work to combat accentism (discrimination based on a person’s accent), and accept that accents will always continue to change,” said Dr Cole.

From the time I was a wee lad, I was always enchanted by English accents.   I could recognize them immediately as coming from the British Isles, although it was sometimes difficult to distinguish them from Australian and New Zealand accents.  It would be sad if they slowly disappeared, perhaps partly under the onslaught of our bland American accent.

Selected readings


Captured from the 6pm news on BBC Radio 4, and from the Daily Telegraph.  To my ear, the lead researcher, Dr Amanda Cole, pronounces some /θ/s as /f/s ("all aged thirty-three", at approximately 01:15) and some /ð/s as /v/s ("computer algorithm", at approximately 01:17).

(courtesy of Philip Taylor)


  1. Ben Zimmer said,

    November 3, 2023 @ 3:52 pm

    Two old LL posts from Mark Liberman are relevant:

    "Jafaican," Apr. 12, 2006

    "What 'Multicultural London English' Sounds Like," Apr. 15, 2006

  2. Daphne Preston-Kendal said,

    November 3, 2023 @ 4:35 pm

    While the evidence (not only phonological, but lexical as well, at least) does seem to bear out an unfortunate decline in the diversity of regional dialect in England, I don’t see that it follows that these changes are due to the ‘onslaught of our bland American accent’. None of the features seem particularly American-influenced. No resurgence of rhoticity, for example, and the spread of t-glottalization is a distinctly different development to AmE t-flapping.

    (I just noticed that the Wikipedia article for t-glottalization has the delightful headnote ‘"Bo’oh’o’wa’er" redirects here. For the item the term mimicks, see Bottled water.’)

  3. Jim Breen said,

    November 3, 2023 @ 5:03 pm

    My accent is rather RP-ish. When I'm in the US, France, etc. people mostly assume I'm British, although once a taxi driver in Boston picked me as Australian (he was a fan of old Australian films.) When I'm in the UK people usually can't identify my accent, however some do notice what one woman called "a faint trace of an Australian accent."

  4. David Marjanović said,

    November 3, 2023 @ 6:14 pm

    “It’s not surprising to find that women speak in a more socially prestigious way, as much previous research suggests women are often more chastised for speaking with regional accents than men,” said Dr Cole.

    Isn't it more likely that men find themselves more easily in social situations where non-posh ways of speaking have what's been called "covert prestige"?

    There is further discussion in this thread.

  5. mg said,

    November 3, 2023 @ 7:35 pm

    Did the researchers not look at the north of England at all? Their discussion seems very much centered on the southern part of the country.

    The class part reminds me of a childhood friend whose family moved to England when we were in high school in the 1970s. She said her American accent (which sounded very English to me by the time we were adults) had been helpful by putting her somewhat outside the class system.

  6. Chris said,

    November 3, 2023 @ 11:00 pm

    This post only refers to England (no mention of Scotland or Wales), so the title should be "The changing accents of English English". It doesn't actually deal with Britain as a whole (and there are interesting things to say about the changes in the other two countries as well). Indeed, as mg said, the north of England might have been worth a look-in as well.

  7. Rob said,

    November 4, 2023 @ 3:16 am

    What everyone (including Victor) seems to be missing is that this was a survey of “nearly 200 people from across south-east England and London”. So it’s not just the North – or Scotland, Wales and NI – that’s not covered, it says nothing about accents across England in general – Midlands, South West, North East etc. There are probably similar changes occurring across these regions as well – less ‘extremes’, more homogeneity, possibly due to increased media consumption.

  8. Peter Taylor said,

    November 4, 2023 @ 6:27 am

    Rob said:

    it says nothing about accents across England in general – Midlands, South West, North East etc.

    There is at least one exception: it mentions Josh Widdicombe as a speaker of SSBE. He's from Devon, so either SSBE really is southern and not just south-eastern or he's adapted his south-western accent.

  9. David Morris said,

    November 4, 2023 @ 7:07 am

    Tangentially related: a watched a series of videos by a young US couple hiking the Appalachian and Pacific Crest trails. I noticed he pronounces 'mountain' with a glottal stop, but can't remember whether it is 'moun(-)ain' or 'mou(-)ain'. How widespread is glottal stopping among young US men (of Asian heritage, judging by his appearance)? How common is glottal stopping when the 't' is part of a cluster?

  10. bks said,

    November 4, 2023 @ 7:40 am

    I have traveled (and travelled) by British rail pass twice: 1970 and 2017. On the earlier trip I found everyone north of Oxbridge to be incomprehensible. On the latter trip, even Inverness was mostly intelligible. Also, what happened to the use of "shall" which seemed very common in 1970 (in UK but not USA)?

  11. Ralph J Hickok said,

    November 4, 2023 @ 7:57 am

    I wonder what "bland American accent" the study refers to. Is it the "bland American accent" of Maine, of Boston's Beacon Hill, of South Boston, of Cape Cod, of New York, of Alabama, of Texas, of northern Minnesota, or somewhere else?

  12. Richard Hershberger said,

    November 4, 2023 @ 8:39 am

    "Remember the Valley Girl accent, which people used to talk about a lot ten or twenty years ago? Where is it now?"

    Ten or twenty years ago? Frank and Moon Zappas' "Valley Girl" was released in 1982. It lampooned what was presumably a real accent, but one of limited scope before the song popularized it. The accent's prominence was an 80s and 90s thing. I don't know of anyone speaks it natively today, now that the fad has died out.

  13. Philip Anderson said,

    November 4, 2023 @ 8:42 am

    Yes, the survey covered one region, but the London press treat the results as applying to Britain as a a whole, and people read those headlines.

  14. ADAM ROBERTS said,

    November 4, 2023 @ 9:27 am

    I'm a late 50s Londoner/SE England geezer (born/raised in Sydenham/Croydon, moved to ast Kent as an older child) I have two idioms: the basically RP manner of my middle class professional and home life, and the estuary english into which I code-switch when eg I meet up with my old school mates, all of whom speak that way. One thing that slips through into my RP speech, and which my (posher-speaking) wife always hauls me up on, are the words "window" and "yellow", which she pronounces "window" and "yellow" but which, in unguarded moments, I still articuate as "winder" and "yeller".

  15. Adrian Bailey said,

    November 4, 2023 @ 9:30 am

    A good guide to what's going on in English pronunciation is Geoff Lindsey

  16. J.W. Brewer said,

    November 4, 2023 @ 9:31 am

    The Telegraph piece is paywalled, so I don't know whether it includes a link to the actual underlying research piece, which is here:

    As noted above, all 193 speakers studied were "from Southeast England," with the specific meanings of "from" and "Southeast England" explained in section 2.1. The vast majority were either students or staff at the university where the research was being done, so this is in large part what I think is politely called a "convenience sample," and does not reflect any attempt to find a genuinely meaningful statistical cross-section of the younger cohort of Southeast England speakers.

    The other thing that may be noteworthy is that they apparently managed to sort 100% of the speakers they studied into one of the three accent groups. This seems intuitively implausible, but I think what it means is that they did some sort of computerized analysis and found that via some sort of mathematical model you could meaningfully assign all 193 idiolects to one of three clusters of features, with any apparent outliers that were meaningfully different to the ear from the prototype examples of any of those three accents at least being in some mathematical sense least distant from the central tendency of one cluster as compared to the central tendencies of the other two. I suspect that if you were doing this by ear there were probably some speakers where you'd focus more on how the individual sounds unlike a prototypical speakers of variety A rather than on how the speaker is at least marginally more similar to a prototypical A-speaker than a prototypical B-speaker or C-speaker. There might also be analyses that would break one or more of the clusters into meaningful subclusters.

  17. Jonathan Smith said,

    November 4, 2023 @ 10:00 am

    @David Morris
    This is "General American" I think and relates to /t/ before [n̩]… though my ear tells me "we adults" mostly use e.g. [ˈbʌʔtn̩] whereas among the youngs (kids/teens) [ˈbʌʔn̩] etc. is taking over. Don't know if phonetic study would confirm…

  18. Victor Mair said,

    November 4, 2023 @ 10:05 am

    @Ralph J Hickock:

    None of those, of course, but thank you for mentioning such colorful varieties, which still do exist to a certain degree.

    When I went to college in the early 60s, many of my classmates — who came from all over the country — had noticeable regional accents. In graduate school during the early 70s, fewer of my classmates had detectable regional accents. When I began to teach in universities in the late 70s and early 80s, fewer still of my students had regional or local accents. By the mid-80s and 90s, it got to the point that students from places like Atlanta and Austin spoke the kind of Standard American English (SAE) that I describe in the next paragraph. When I asked them why they had no accent, they would tell me that their parents would forbid them from speaking that way, and their teachers would correct them if it was too obvious.

    I was thinking of General American English or just General American; Standard American English (SAE), also called Academic English (AE); Mainstream American English (MAE).

  19. J.W. Brewer said,

    November 4, 2023 @ 10:30 am

    To another of the points upthread, the underlying research article does not mention Josh Widdecombe or suggest that his was one of the 193 idiolects studies, so that's presumably a journalistic embellishment. Saying that these three accents dominate Southeast England (on a specific definition of the boundaries intended) is not a claim that none of them are also found elsewhere. And indeed if you did good dialectology and produced a map with a bunch of isoglosses showing the northernmost extent of the distinctive features of "Estuary English" you would not expect that accent's range to stop cleanly and immediately at the Essex/Suffolk border or any other political boundary. Yet for purposes of research design it was simplest to define "Southeast England" as London plus such-and-such list of shires even presumably either some of those included were marginal in southeastness, or the next one north or west that was excluded would have at least some claim to southeastness.

    As to American accents, as I'm sure Prof. Mair is aware, students from places like Atlanta and Austin that go to high-prestige private universities in the northeast are as a general matter grossly unrepresentative (in terms of wealth, social class, parental level of formal education etc.) of the general populations of Georgia or Texas so you would expect them to have more "national" and less "regional" accents for those reasons. Drive 50 miles in any direction from either Atlanta or Austin and listen to some high-school seniors who aren't planning on going to college at all, and you may hear something different.

  20. David L said,

    November 4, 2023 @ 12:05 pm

    @David Morris:

    The glottal stop that you heard in 'mountain' seems to be increasingly a part of the way younger Americans speak. I don't think it has anything to do with ethnicity. I remember hearing it from black news reporters in the DC area (e.g. when saying Bill Clinton) and I also heard it recently from a young white woman on local TV in Maine, where I live now. She wasn't a local, I'm guessing from her voice. I don't know if anyone had surveyed its emergence and distribution.

  21. Jarek Weckwerth said,

    November 4, 2023 @ 5:05 pm

    @ David Morris and David L
    The glottal stop is indeed on the rise in the US and Canada. However, because it doesn't have the social stigma that is (used to be) attached to it in the UK, there has been less research into it. If you're interested, I would suggest starting from work by David Eddington. In one of his studies, he found about 30% of prevocalic /t/'s being glottalized, which was actually more than a comparable study had found for SSBE. The speakers were from the Pacific Northwest, if memory serves. A colleague of mine did a study on the Buckeye Corpus (Ohio) and found lower rates (but on the whole quite a lot). It's in Laboratory Phonology, thus open access. Yay!

  22. David C. said,

    November 4, 2023 @ 8:23 pm

    The 2010 David Eddington paper cited in the link above indicates that young Westerners tend to glottalize more. My personal observations align with that and I find that only a subset of Western American English speakers have very noticeably strong glottalization in words like mountain and Martin – almost like a pause in the middle of the word ("MAR-[ʔ]-n").

  23. Michael Watts said,

    November 5, 2023 @ 1:01 am

    I noticed he pronounces 'mountain' with a glottal stop, but can't remember whether it is 'moun(-)ain' or 'mou(-)ain'. How widespread is glottal stopping among young US men (of Asian heritage, judging by his appearance)? How common is glottal stopping when the 't' is part of a cluster?

    As at least one other person has noted, the cluster has nothing to do with it. The /t/ in "mountain" obligatorily becomes a glottal stop because it is followed by /n/. Kitten, mitten, Britain, flatten, and every other word following this pattern are affected.

    Note that this demonstrates a phonemic difference between /t/ in "mitten" [glottalization required] and /d/ in "midden" [glottalization impossible], even though intervocalic /t/ and /d/ are not supposed to be distinct. The simplest resolution appears to be to say that in contexts like this /ən/ is actually /n̩/.

  24. Michael Watts said,

    November 5, 2023 @ 1:02 am

    Hm, I put a syllabic diacritic on that final n, but it doesn't seem to have come through.

  25. Rodger C said,

    November 5, 2023 @ 10:52 am

    I see it.

  26. Andrew Usher said,

    November 5, 2023 @ 10:55 am

    It's obvious it was meant. The cluster actually makes glottaling less likely, as /tn/ is often 'unpacked' when another consonant precedes – as 'mountain' and 'Martin' – but practically never when a vowel precedes, as 'button' and 'mitten'. When the /n/ is no longer syllabic, the glottal stop is not so common.

    k_over_hbarc at yahoo dot com

  27. Steve Hartman Keiser said,

    November 5, 2023 @ 11:36 am

    I'm a Midwesterner, late 50s, and I find glottalization of /t/ before syllabic /n/ unremarkable in words like 'mountain', 'button', 'important', etc. I'm pretty sure this has been my dialect my whole life.
    What I have noticed in younger Midwestern speakers (i.e. my children) is that /n/ is no longer syllabic but is produced more like [ɪn]. So they say 'button' as [bʌʔɪn] while I say [bʌʔn̩]
    And occasionally these younger speakers now flap the /t/ so they say 'important' as [ɪmpɔɹɾɪnt] while I say [ɪmpɔɹʔn̩t].

  28. Andrew Usher said,

    November 6, 2023 @ 7:01 am

    It's astonishing to me to think of such a pronunciation as [bʌʔɪn] with an intervocalic glottal, as American. That's something I've only ever associated with British and Scottish English (yes, Scotland is part of Britain but everyone uses 'British' English to refer to an accent of southern England).
    I can't believe I use it with any frequency and I could never consider it standard (the quality of that weak vowel is questionable as well).

    On the other hand I can think of many possibilities for 'important'. Maybe my drawing a distinction between the two reflected my idiolect more than I thought.

  29. /df said,

    November 7, 2023 @ 7:49 am

    I'm not sure if this has been discussed in LL, but one weird innovation in British, in fact English, English accents is what I think of as "well-to-do young woman's accent". All the vowels are strangely off from RP or SSBE, maybe a little Scandinavian: "coast" almost rhymes with "baste", "book" is heading towards "berk". For an example listen (3 minutes in), to the "2nd most influential woman-scientist-on-Twitter" Prof Alice Roberts (but not all the women with these vowel sounds can have come from Avon): check "loads", "about", "to", "knew", "knows", "enjoyed", "bones", "previous", and so on.

  30. Jarek Weckwerth said,

    November 8, 2023 @ 10:28 am

    @/df I don't find this particularly unusual. Her variants are just more advanced (i.e. "modern = extreme") realizations on trajectories of change that have been discussed for quite some time. The central unrounded FOOT, the front offglide of GOAT, etc., all of that has been the topic of many of Geoff Lindsey's videos on YouTube, and he's an absolute king of the genre. If you're not familiar, you can start e.g. here.

  31. Pamela said,

    November 8, 2023 @ 11:04 am

    I wonder how great this shift is in comparison to the early Hanoverian period, when upper class Londoners starting knocking -rs off the end of words and saying "eye-ther" instead of "ee-ther" in imitation of their German overlords. It would be nice if class and pronunciation could be uncoupled, and maybe in Britain they will be, but it seems that Americans will always hear some British accents as "plummy"–in recent years I've heard Americans refer to Britons who spoke with slightly elevated but not at all upper-class university accents as "plummy." I'm pretty sure living Americans rarely if ever encounter Britons who speak genuinely plummy, since every year fewer of these people are even alive. On the other hand, none of these southern accents that have been in the news for the past decade seem to actually get attention in the media, and particularly not television and films, where everybody seems to have a northern accent–like brummy is the new plummy.

  32. Philip Taylor said,

    November 8, 2023 @ 1:05 pm

    Far be it from me to dare to challenge anything that the great Geoff Lindsey opines, but following Jarek’s link two comments above I see on-screen at about 02:13 a vowel chart with those vowels that take a so-called "linking R" on the right, and those which do not take a linking R on the left. But all of Geoff’s example words on the right have an explicit "r" towards the end of their spelling (hear, care, purr, cure, score, star) while of his examples on the left (see, pay, cry, toy, do, go, vow) only one contains an explicit "r" in its spelling (cry) and that in the onset rather than towards the end. It is therefore hardly surprising (to me, at least) that none of his left examples gain a linking R when the -ing suffixed is added. So why does Geoff attribute the linking R phenomenon to the set of vowels rather than to the simple fact that all of his examples which do take a linking R have an explicit "r" towards the end of their spelling, while none of his examples that do not take a linking R have one ?

  33. Jarek Weckwerth said,

    November 8, 2023 @ 5:07 pm

    @ Philip Taylor: He's arguing from the point of view of phonology, not spelling.

    Today, in a non-rhotic variety of English, the r is only there in the spelling, not in pronunciation. A child acquiring that variety will not know that until they learn to read. Thus, during the important stages of language acquisition, the only evidence of that historical /r/ will be in the inflected forms. Therefore the generalization the child usually makes is that it's the vowels that call for a linking /r/ if another vowel follows. And, as a result, you get it not only in storing etc. but also in drawing. Peppa and George is typically /pepərən…/.

  34. Philip Taylor said,

    November 9, 2023 @ 4:03 am

    Yes, I understand Geoff’s perspective, Jarek, but with the greatest of respect I disagree with his conclusions. In "hearing", "caring", "purring", "curing", "scoring" and "starring", I would argue that the linking R which can be clearly heard in almost everyone’s speech is a function of the explicit "r" (or "rr") in the spelling. If some (but not all) base their pronunciation of "drawing" on the "scoring" model, then this is simply an error of speech, attention to which was drawn by my teachers during my formative (primary school) years. For Geoff to demonstrate that his hypothesis is correct and mine is false, he would need to adduce words in his left group which contain an explicit "r" yet which have no linking R when an -ing suffix is added, and this he fails to do.

  35. Andrew Usher said,

    November 11, 2023 @ 9:52 am

    Yes, pronouncing the R in such words in mandatory, but that's not evidence if it being caused by the spelling – it's sufficient that the R has always been there, and one before a vowel in the same word is never dropped (The fact that the first R in 'forever' can be dropped shows that it is two words phonologically).

    Geoff is accurately describing how it works for the great majority of speakers, even though he himself doesn't seem to pronounce unwritten R, and you may not, it's a fact that almost all English people a good fraction of the time, at least, do, and he explains why. I've learned by now that he can be trusted more than any other authority when it comes to SSB.

    Unfortunately he can't be for American English, but it would perhaps be too much for me to digress into why.

RSS feed for comments on this post