Apologies for the long hiatus, due to a sudden rush of courses, talks and workshops in various parts of the world. Since I last wrote, 2012 has turned to 2013. We find ourselves half a century from a watershed in British social history.

Nothing symbolized this watershed more than the appearance on the self same day in 1962 of the Beatles’ first record Love Me Do and the first James Bond film Dr. No. Before long there was hardly anyone in the world more famous than the four Liverpool lads and the working-class Scot Sean Connery.

While Dr. No was bringing to the screens of 1962 a new and gaudy mixture of sex and spies – condemned by a special Vatican communiqué – a real-life mix was already brewing. In a racy milieu where ‘good time girls’ were introduced to members of the ruling class at stately-home parties, Christine Keeler allegedly started affairs both with the Secretary of State for War John Profumo (who was a Brigadier and a Baron, and married) and also with a Soviet naval attaché, who may have been using Keeler to get nuclear secrets at the height of the Cuban missile crisis.

The ensuing scandal, inflamed by the daring new sensation of TV satire (the BBC’s That Was The Week That Was ran 1962-3), not only brought down the dissembling Profumo but shifted public attitudes against the upper-class establishment typified by the then Conservative government. Prime Minister Harold Macmillan was briefly replaced with the even posher Alec Douglas-Home – of whom it was once said “in his heart of hearts, Alec believed that the work of government should ideally be carried on by the members of twelve upper-class families” – before Labour won the next election, led by state-educated Merseyside M.P. Harold Wilson, a man already on cosy terms with the Beatles. Andrew Marr’s neat video summary of the period is here.

1962 also saw the first appearance of A.C. Gimson’s Introduction to the Pronunciation of English. Gimson, known widely as “Gim” gɪm, was Daniel Jones’s pupil and successor at University College London. He was one of the greats in UCL’s phonetics-teaching tradition; I was lucky enough to attend lectures of his in the late 1970s and flattered that he was happy to chat about phonetics with a keen undergraduate. His 1962 Introduction, focussing of course on British “Received Pronunciation” (RP), was clear, packed with detail and more approachable in style than Jones’s Outline of English Phonetics. It established itself as a basic text and would go into several editions, more recently revised by Alan Cruttenden.

The book’s most influential feature was its broad transcription of Received Pronunciation vowels, which for convenience I’ll refer to as Gim 62. This system displaced the earlier more Jonesian systems which had been very broad. They had shown vowel contrasts largely with typographically familiar letters, preferring to leave precise IPA notation to specialist work. Here are some examples from earlier systems of the simplest type (for further discussion, an invaluable paper by John Wells is here):

TRAP	        a		START		 
LOT		o		THOUGHT	         
FACE		ei		PRICE		ai		MOUTH		au

Gim 62 took a narrower approach to broad transcription, sacrificing economy and typographical ease in order to represent vowel qualities more precisely through extensive use of IPA symbols, e.g.

TRAP		æ		START		ɑː 
LOT		ɒ		THOUGHT		ɔː 

So Gim 62 was a relatively accurate, fine-grained snapshot of the social elite’s vowel qualities, taken just as the upper class was to fall from grace and Britain was to enter a period of exceptional social and phonetic change. The great irony of Gim 62 is that this snapshot became increasingly established as it became increasingly out of step with the standard accent.

We can see how rapidly things started changing if we compare the discussions of RP trends by Gimson in 1962 and by John Wells in his Accents of English of 1982. The earlier discussion (“The Present-day Situation”, p. 83) distinguishes three types of RP: conservative, general and advanced. Advanced RP is

used by young people of exclusive social groups – mostly of the upper classes, but also, for prestige value, in certain professional circles. In its most exaggerated variety, this… would usually be judged ‘affected’ by other RP speakers, in the same way that all RP types are liable to be considered affected by those who use unmodified regional speech.

That quote seems to place Gimson on the far side of an historical divide (rather like his references to the influence of BBC radio, television not rating a mention). Gimson is writing in a world where the future of the reference accent seems to lie in the mouths of posh-sounding young things who if anything are liable to sound posher than their elders.

A mere 20 years later, the world seems to have turned upside down. John Wells on Gimson’s “advanced RP” speakers:

social changes in Britain in the 1960s and 1970s… robbed them of their position as the most admired and imitated group. Over the last quarter-century all the signs are that the covert prestige of working-class speech is acting as a more potent source of innovation than the overt prestige of advanced RP. Mainstream RP is now the subject of imminent invasion by trends spreading from working-class urban speech, particularly that of London (Accents, p. 105-6)


With the loosening of social stratification and the recent trend for people of working-class or lower-middle-class origins to set the fashion in many areas of life, it may be that RP is on the way out. (p. 118)

In the period when John was writing Accents of English, there was considerable talk of low-born speech models. John himself quotes posh writer Jilly Cooper on the lowering/backing of the TRAP vowel:

All those female interviewers talking about bunk bulences and Ufrica. I suppose they all grew up in the Sixties… when… working class became beautiful…

In talking of perceived working-class beauty and imminent working-class invasion, I think both Jilly Cooper and John Wells were slightly overstating the case. Working-class lifestyles, attitudes and behaviours never really stopped being looked down on. Arguably more basic in separating the world after the ‘1962’ watershed from preceding decades (indeed centuries) was the stigmatization of privilege.

Before this sea change in social attitudes, privilege was a fact of life to be envied and deferred to. It was a fact of life that Prime Minister Harold Macmillan was an Old Etonian; that his predecessor Anthony Eden was an Old Etonian; that his successor Douglas-Home was an Old Etonian. (Baron Profumo, like Eden’s predecessor Churchill, was an Old Harrovian.) When Old Etonian author Ian Fleming created his great wish-fulfilling hero in 1952, it was inevitable that James Bond would be an Old Etonian too.

A decade later, the North American producers of the first 007 film clearly sensed the winds of change when they chose for the role an ex-milkman from Edinburgh; the character would surely have had less appeal if he had sounded as posh as his creator. Today, for the first time since Douglas-Home, we have a Prime Minister who is an Old Etonian, but his privileged background is a public-relations liability, repeatedly pointed out and mocked – e.g. reminders that Cameron, like Profumo, belonged to the infamous Bullingdon Club. Privilege has never stopped conferring advantage; but after ‘1962’, the perception of being privileged became undesirable in new ways.

And Received Pronunciation, from its very conception, was the accent of privilege. Daniel Jones in 1918:

It is convenient for present purposes to choose as the standard of English pronunciation the form which appears to be most generally used by Southern English persons who have been educated at the great English public boarding schools.

John Wells in 1982:

Typically [those who speak RP] belong to families whose menfolk were or are pupils at one of the ‘public schools’ (exclusive private schools standing outside the state education system).

After the sea change circa ’62, the motor of phonetic change for young members of the higher social classes was not so much an urge to sound Cockney as the desirability of shifting away from the traditional sound of privilege: a letting-go of brakes that allowed a rapid and wide-ranging assortment of changes, many of which shifted towards London speech while others shifted away from it. The lowering of DRESS and TRAP, for instance, was not driven by Popular London, which maintained traditional RP-like values; nor do we need to see it as a part-imitation of the Beatles or other northerners; young privileged speakers simply let their front vowels fill the vowel space more naturally, establishing a different sound from the characteristic close-front congestion of posh RP.

One hint in Accents of English at the post-1962 stigmatization of poshness was the new concept of “upper-crust RP” or “U-RP”, something for which Jones and Gimson had had no need. John Wells associated U-RP with stereotypes such as the dowager duchess, the Terry-Thomas cad, the elderly Oxbridge don and the jolly-hockey-sticks schoolmistress at an expensive school. Crucially, these stereotypes were figures of fun: they and their accent were essentially ridiculous. Where the traditional relationship between social status and phonetic prestige had been fairly simple and linear, it was now becoming increasingly possible to sound too posh for your own good.

John’s observations on U-RP were a sign of the times; so was the mere fact that he made them. Comical duchesses and Terry-Thomas cads were around when Gimson was writing his Introduction; but back in the deferential world pre-That Was The Week That Was, it would probably have seemed inappropriate, perhaps even impudent, to discuss them in a scholarly phonetic work.

But despite all the change he was discussing, John chose to conserve Gim 62’s RP vowel symbols, modifying only ɑʊ to for MOUTH. With RP DRESS, for example, John continued to describe it, like Gimson, as closer to cardinal e than to cardinal ɛ. While pointing out that true in GOOSE was actually U-RP, John retained it as the broad symbol for GOOSE in ‘mainstream’ RP too. Clearly phonetic shifts were underway – by 1982 I myself was probably hearing close-mid DRESS realizations as markedly posh – but for broad transcription it was felt that Gim 62 was not yet worth changing.

On the contrary, it was during this very period, just as Gim 62 was beginning to show its age, that the system was being canonized as never before. Peter Roach, author of the famous English Phonetics and Phonology (soon to be an enhanced ebook from Cambridge University Press), recently corresponded with me about the need for updated English vowel symbols. Peter has kindly allowed me to quote from his informal impression of historical developments (with the proviso that this not a properly researched academic statement):

Some time around 1980 the major British publishers (or some of their senior editors concerned with English phonetics) agreed among themselves that they would use Gimson’s symbols in publications. Certainly I was told by CUP to change the symbols in my draft of ‘English Phonetics and Phonology’…

So began the real entrenchment of Gim 62, this rapidly ageing system soon attaining an established status that had never been enjoyed by any other system. More recently Gim 62 has received a further enormous boost – again, not driven by academic/scientific arguments – through the internet, with the planetary dissemination of online dictionaries, blogs, Wikipedia transcriptions, etc.

Of course, there have been tinkerings with the system since 1962, but these have been minor and have tended on the whole to make it worse. Updating ɑʊ to for MOUTH was certainly an improvement; but changing ɛə to for SQUARE was a phonetic step in the wrong direction. Some have changed the PRICE vowel from to ʌɪ, a dubious move for various reasons (to show the backness of PRICE, ɑj or ɑi is preferable). At least one recent book has changed MOUTH to æʊ, presumably due to mistaken faith in the currency of Gim 62’s æ for TRAP. Also unfortunate was the addition to Gim 62 of two extra phonemes which never really existed, namely i and u (I’ve discussed i in detail in another post).

For an audio-illustrated account of how the RP vowels notated by Gim 62 differ from the vowels of the contemporary standard/reference/teaching accent, you can work through a long, earlier post here. In the remainder of this post, I’ll just give three examples to show the different ways in which the persistence of Gim 62 causes problems.

First, here is part of the vowel system for London English as notated in 1982’s Accents of English (p. 304):

KIT		ɪ		FLEECE, happY	ɪi 
TRAP		æ		START		ɑː 
GOAT	ʌʊ		GOOSE		ʉː 

The problem here is the adherence to Gim 62’s ɪ and ʊ for the second elements of closing diphthongs, e.g. PRICE, CHOICE, GOAT. Gimson’s symbols may have been appropriate for pre-1962 RP, in which such diphthongs ended less than close, even in accented final position; listen to plays and May:

But how then does one show the closing nature of the London FLEECE/happY diphthong? John is forced to make a distinction between Gimsonian ɪ-closing diphthongs like CHOICE and the i-closing diphthong of FLEECE/happY – a distinction which surely no language needs to make. Further, although he refers in the text to “the GOOSE diphthong”, John chooses like Gim 62 a monophthong for its notation. The result is that the London long vowel system is presented as far messier than RP’s, when in fact it is neater, with front-closing diphthongs (like FLEECE and CHOICE), back-closing diphthongs (like GOOSE and GOAT), and just one group of long monophthongs which all trigger linking r (like START). This is especially worth pointing out today because, in this respect, the London 1982 system essentially is the contemporary standard BrE system.

Second, here’s a clip from a BBC educational video telling learners to pronounce the SQUARE vowel with a marked change in mouth position during the vowel:

To my ear, the utterance which we’re told to watch at the end might be a disyllabic pronunciation of the name Ayer; it certainly doesn’t sound to me like a standard pronunciation of BrE air. The presenter’s SQUARE vowels sound considerably more normal (i.e. more monophthongal) when she’s speaking English rather than having to demonstrate the symbol “/eə/”. Eg carefully and beware

A pronunciation textbook published only last year similarly encourages mirror practice for “/eə/” on the same kind of model, with photographs indicating different mouth shapes at the start and end of the vowel (while the accompanying audio contains monophthongal realizations in phrases like dairy fare). Why nowadays would one teach learners to ensure that their mouths change shape during the stressed vowels of dairy fare? The basic answer, I suspect, is that Gim 62 says so.

Lastly, Wikipedia. If you go to its entry for the International Phonetic Alphabet you will find at the top:This, presumably, is a transcription of the initialism IPA as said in some standard accent of English. It’s clearly intended to be a narrow transcription: there are square brackets, and aspiration is marked on the p. How then do we account for the vowel symbols? You wouldn’t be very likely to hear these qualities from a British newsreader, though in North America they may not be so far from standard. But what in particular are we to make of that last diphthong, gliding from [e] to [ɪ]? As demonstrated in Wikipedia’s own audio clips, [e] sounds closer than [ɪ], with first and second formants measurably further apart – making [eɪ] an opening/centralizing diphthong. Here it is, crudely spliced together:

If you check out other phoneticians’ IPA demos you’ll find that Jill House’s pronunciations somewhat resemble Wikipedia’s; by contrast Peter Ladefoged says [e] with a slightly opener auditory quality than [ɪ], and with formants less far apart; while for good measure John Wells makes [e] and [ɪ] practically identical. In narrow phonetic terms, then, [eɪ] is hardly a sensible diphthong: its two elements do not differ consistently in a way that any language would make use of, without involving another factor such as vowel length.

Wikipedia is a priceless asset and I do appreciate the work of this entry’s creator(s). But why would one present this confusing pseudo-diphthong as a head example of narrow IPA? I may be mistaken, but it looks like yet another misguided use of Gim 62 as phonetic gospel.

Three examples, then, which show the legacy of Gim 62: hampering phonological analysis, encouraging dubious pedagogy, muddying the IPA.

Gimson’s 1962 Introduction and its broad transcription system are still interesting and valuable, but as historical objects. Gim 62 was a fairly accurate snapshot of upper class pronunciation which appeared at roughly the historical moment when upper class pronunciation started to be something to dissociate oneself from. There are those who refer to today’s British standard/reference/teaching accent as “modern RP” or “current RP”; but they keep transcribing it as if it were old RP, i.e. with Gim 62. “RP” and Gim 62 have become inextricably intertwined. They refer to the standard accent of BrE as it was on the other side of an historical watershed, not as it is today.

15 replies
  1. Aleph
    Aleph says:

    Wonderful, wonderful, wonderful post.

    I am so in love with every discussion you dedicate to the Empire Speech that I am seriously hoping that you one day write a whole book about it: only U-RP, and nothing else. No other accent discussed.

    But I do have a few questions: back in the day, say in the last years of the 19th century and the first ones of the 20th, did people really say [ɑʊ]? What about [ɛə], [ɛɐ] and [eə]? Which one of those came first? Which one was considered more upper class? Or did they all co-exist?

    Where did you find that clip of King George V using [eə] when opening the Kingston power station?

    • Geoff Lindsey
      Geoff Lindsey says:

      Thanks for your very kind words. I doubt that I know enough about old U-RP to fill a book!

      The vowels you mention certainly all existed. I can’t answer your questions about historical order, but I think one can make a generalization that centring diphthongs tend to end with a more open quality when in absolute final position, esp. if accented, and to have less of a diphthongal movement when the syllable is closed by a consonant, esp. if unaccented. This can be related to the allophony of commA, for which in both U-RP and contemporary London/Estuary we get ɐ or a in absolute final position.

      ɛə for SQUARE is still common in accented final position in standard BrE, as long as the centralizing offglide is slight enough; but I think the same is true of for NORTH and ɑə for START.

    • Ed
      Ed says:

      The transcription [eə] is often used for the New Zealand pronunciation of SQUARE. I don’t think that any English people would use this. To English ears, a New Zealand pronunciation of “pair” sounds like “peer”, of “bear” sounds like “beer”, etc.

      • Aleph
        Aleph says:

        Thank you, Geoff, so much for everything: the posts, the generous replies, the audio files! I still hope you will one day write a book about the long-gone accent. The only one I know that covers it that has been published in recent times is a book by Edda Sharpe and Jan Haydn Rowles.

        Ed, King George VI used it. He’s English. No?

        • Ed
          Ed says:

          I presume that you’re talking about the second voice clip here. That sounds like [ɛə] to me.

          However, in the first clip on that page, I agree that Michael Thomas says [eə] when he first says “prepare”. I can hear a difference between these two examples. I think that Michael Thomas’s pronunciation is clearly not that of an English native.

          Geoff has said before that he finds [e] and [ɪ] difficult to distinguish, and I’m sure that many other people do. This is why English people often mistake the New Zealand [eə] in SQUARE for their /ɪə/ in NEAR.

          • Aleph
            Aleph says:

            Michael Thomas says [eːr], King George VI says [eə]. Notice the tenseness of the whole vowel. Then compare it to Edith Evans or Vita Sackville-West on YouTube.

  2. Aleph
    Aleph says:

    And also, why ‹ɪ› in , ? What justification for it? What was the reasoning? Phonetically, what is the exact sound the upper-class members used?

    • Geoff Lindsey
      Geoff Lindsey says:

      There’s a reasonable argument for ɪ and ʊ in diphthongs, namely that the diphthongal trajectory rarely reaches the full ‘corner’ quality of i/j or u/w. There’s a general phenomenon of “undershoot”: as speech becomes less careful and more rapid, diphthongs become less diphthongal. A phrase like don’t take my toys said at normal tempo may have very little diphthongal movement. I think undershoot is pretty universal.

      An additional factor is that in old RP many speakers ended their diphthongs noticeably short of i/j and u/w even in emphatic speech, which as I pointed out in another post is something that entertainers sometimes mimic when making fun of old posh RP, just as they mimic the use of ɪ in happY and æ or ɛ in TRAP.

      What matters in diphthongs is the direction in which they seem to be pointing. There are three main directions which languages choose: towards i/j, towards u/w, and towards schwa. The use of ɪ and ʊ in the broad transcription of diphthongs of any language strikes me as “spuriously accurate”: it implies that a language might distinguish ɪ-ending diphthongs from i-ending diphthongs, which I doubt ever happens. Incidentally, even Gimson (1960) acknowledged that the diphthongs could be written with j and w instead of ɪ and ʊ.

  3. Aleph
    Aleph says:

    Oh, well, since I’m not getting my answers, let me just point to the new, eighth edition of Gimson’s Pronunciation of English, which will be published on 28 March 2013 – such a coincidence that this post surfaces exactly at the time of the new edition! Amazing.

    You can find it on Amazon, but the publisher’s website still has no information.

  4. Geoff Lindsey
    Geoff Lindsey says:

    The book is here. I haven’t had a chance to go through it in detail but it seems to be clear and simple with an appealing visual design, and the audiovisual speech material seems to provide generally good models. My comments in the post were not intended to constitute a negative review.

  5. Ed
    Ed says:

    I’ve wondered if we’re moving towards a similar situation to that in the USA in terms of accent and privilege. The diversity in accents in not what it used to be (if we exclude accents from recent immigrants). In some areas, the local dialect has completely died out and been replaced by this standard. In other areas, there is still a distinct local accent but it is much closer to the standard than it was a generation or two ago. However, the standard accent is loosely defined and it is not a great impediment to a career if you have some regional features to your speech. This all sounds like the situation in the USA with General American. I think that our class system has come to resemble the USA’s, and this has affected the status of our accents.

Leave a Reply

Leave a Reply

Your email address will not be published. Required fields are marked *