Estuary English: Hybrid or Hype?

J A Maidment

Department of Phonetics & Linguistics, University College London

Paper presented at the 4th New Zealand Conference on Language & Society, Lincoln University, Christchurch, New Zealand, August 1994.

If any of you have read any British newspaper regularly or listened to British radio over the past two or three years, there is a good chance that you have come across the term Estuary English. There have been articles on this topic in The Times Educational Supplement, The Sunday Times, The Guardian to my knowledge. And there has even been a piece on it in The New York Times. There have been items on Estuary English on BBC Radio 4, the BBC World Service and the London Broadcasting Corporation.

The term Estuary English was coined in 1983 by David Rosewarne, who at the time was a postgraduate student of Applied Linguistics at Birkbeck College in the University of London, and first appeared in print in an article by Rosewarne which appeared in the Times Educational Supplement in 1984. This article, together with a very similar, but slightly expanded article which appeared the magazine English Today in January 1994, forms the total (as far as I know) of what Rosewarne has made public in print about his ideas. The other source of information about EE in print is a book by Paul Coggle, called Do you speak Estuary?, published by Bloomsbury in 1993. I shall base my talk on these sources, together with some ideas and concerns outlined in an as yet unpublished discussion document entiltled "Transcribing Estuary English" by John Wells, who is professor of Phonetics in the University of London. The document was circulated to colleagues in the department of Phonetics & Linguistics at UCL in February 1994.

Rosewarne's 1984 TES article was entitled "Estuary English" and bore the headline:

David Rosewarne describes a newly observed variety of English pronunciation

Rosewarne, in this article, defines EE as follows:

Estuary English is a variety of modified regional speech. It is a mixture of non-regional and local south-eastern pronunciation and intonation. If one imagines a continuum with Received Pronunciation and London speech at either end, EE speakers are to be found grouped in the middle ground.

In this talk I shall look at three aspects of the EE phenomenon.

  1. the phonetic and phonological features claimed for Estuary English
  2. the adequacy of the published treatments of Estuary English leading to the question "Does EE really exist?"
  3. the reaction of the British media and public to Estuary English

The phonetic and phonological features of EE

Features absent in EE but normally found in Cockney

A summary of the main claims about the phonetics and phonology of EE can be seen in the above list. We will look at one or two of them in detail, but I am afraid we shall have to reject some of the claims towards the end of the list as being either vacuous or simply untrue.

Glottal replacement:

This is a process which replaces [t] with [?] when it occurs in other than syllable initial position. The rule can be expressed as follows:

substitute [?] for [t] when BOTH preceded by a vowel or /l/ or /n/ AND followed by end of word or a consonant other than /r/

In fact [t] is rarely preceded by /l/ in Cockney or EE because, as we shall see /l/ in preconsonantal position is usually replaced by a vowel.

Examples: bit=[bI?],football=[fU?bOo], belt=[beo?], Cheltenham=[tSeo?n@m], bent=[ben?], Bentley=[ben?li]

but not in: best *[bes?](Not possible in any accent), twenty *[twen?i], water *[wO:?@], mattress *[m{?r@s] (All possible in Cockney)

These examples are from John Wells' discussion document, mentioned before. Speakers who otherwise would be judged to speak undoubted RP can however be heard to produce glottal replacement of exactly this kind and glottal replacement of [t] is common the popular speech of many cities far removed from the supposed domain of EE: Birmingham, Glasgow, Manchester and New York to name a few. So glottal replacement is certainly not a defining characteristic of EE. EE glottal replacement, however, does differ from that in Cockney in being confined to the environments already discussed. Cockney speakers often use [?] for [t] in all environments where it is not syllable initial. Cockney speakers also sometimes extend glottal replacement to affect [p] and [k] as well as [t].

Before leaving the subject of glottal replacement, I would like to mention the claim by Rosewarne that [d] is also subject to this process. This, I think, is simply an error, although the claim appears in both his 1984 and his 1994 articles. The only speakers of British English that I have ever heard who replace [d] with [?] come from Yorkshire. This is a consequence of what is sometimes known as Yorkshire Assimilation which changes voiced plosives to their voiceless counterparts at the end of word before a word beginning with a voiceless consonant. Thus, a word like bad, may well find itself in environment where the final [d] is replaced by [t] and the [t] in its turn may be replaced by [?]. Examples of this would be [b{? taIm] and [gU? TIN]. However, no-one has claimed (as far as I know) there is any connection between EE and the Yorkshire accent. To be fair to Rosewarne, there is one other situation in which a [d] may be replaced by [?], but this affects only the word had in the collocation had to. Many speakers of many British accents pronounce this phrase [h{? tu:]. Such a pronunciation parallels the devoicing of the final consonants in the first word of the phrases has to and have to.


Examples: milk=[mIok], mill=[mIo]

Some possible EE vowel neutralisations before [o]

Some Cockney vowel neutralisations probably not part of EE:

Like Cockney (and some other English accents) EE exhibits l-vocalisation. The rule is:

/l/ is realised as a back, closeish rounded vowel in positions before a consonant with or without an intervening word boundary or in absolute utterance final position before a pause

[o] symbolises a sound very like what most speakers of English produce for w. Indeed, both Coggle and Rosewarne use [w] to symbolise the sound.

The effect that [o] has on preceding vowel qualities and on the loss of contrast between vowels may be restricted in EE. The neutralisations which Cockney exhibits before an [o] are much more extensive, it is claimed.

Diphthong shift:

The diphthongal vowels of FACE, PRICE and GOAT in EE are supposedly as shown below. These are also what would be expected of a Cockney speaker.

GOAT allophony:

In common with speakers of many other accents, EE speakers are said to exhibit salient differences of the quality of the vowel of GOAT in general environments compared to pre-lateral or pre-[o] environments. This type of allophony is not usually recognised to be part of RP.

GOAT vowel --> [QU] before [5] or its reflex [o]


Yod coalescence:

Most accents allow the coalescence of alveolar plosive and following palatal approximant (yod) to produce a postalveolar affricate. RP is an accent which tends to confine this to unstressed environments. So: constitute [kQnstItSu:t], did you [dIdZu] are all right in RP. EE is said to allow such coalescence in a much larger set of environmnents and specifically before stressed vowels as seen below.

The last three items on the list of features claimed for EE are extremely dubious.

/r/ realisation:

I have doubts about this simply because I find it difficult to see what Rosewarne and Coggle are getting at. Coggle seems to think that the usual target for EE /r/ is a labiodental approximant. I say seems to think, though he does not actually say so, because he says that EE /r/ sounds like [w] and also lists a number of celebrities: Jonathan Ross, David Bellamy, Derek Jameson, as users of this /r/ variety. At least the first two of these definitely use a labiodental. But, a labiodental approximant is well-known in speakers of all accents as a minor defect. It really is not clear if Coggle is claiming that the large majority of current EE speakers use it. If so, I simply think he is wrong.

Rosewarne's account of EE /r/ is equally puzzling. In both the 1984 and 1994 articles he describes the production of /r/ so:

the tip of the tongue is lowered and the central part raised to a position close to, but not touching, the soft palate

This seems to suggest a velar approximant articulation. This again is a well-known "deviant" production of /r/ in speakers of many accents. Rosewarne gives examples of EE-speaking celebrities who use this articulation. At least one of these, Paul Merton the comedian, does not, in my experience, use a velar approximant articulation for /r/.

Nuclear prepositions:

Rosewarne claims that EE speakers are more likely to place the intonational nucleus on a preposition. The same example he uses in both his articles illustrates this:

Let us get TO the point

Speakers of many accents of English are likely to use nuclear accented prepositions in what are known as counter-presuppositional utterances. A good example of this sort of utterance is:

Don't phone the fire brigade. The house isn't ON fire.

Here the speaker seeks to counter a presupposition that is guiding the hearer's behaviour. This sort of utterance contrasts with what is known as a counter-assertive utterance (or more simply a contradiction) where the speaker negates the hearer's explicit assertion of a proposition:

Rosewarne ignores the possibility that the examples of EE nuclear prepositions he claims to have observed are simply examples counter-presuppositional utterances.

Another use of nuclear-accented presuppositions which is fairly common is stylistically determined. Announcements, especially over public address systems, seem for some reason to encourage their use:

The train arriving ON platform 14 is the 16.00 FROM Cheltenham Spa.

Rosewarne again ignores the possibility of a stylistic explanation for the supposed EE occurrences of this phenomenon.

Tone and pitch features:

What Rosewarne has to say about EE intonation is so vague and unsubstantiated that it need not detain us very long. He says that:

there is a a rise/fall intonation that is characteristic of Estuary English

One would like to know what exactly marks this characteristic rise fall intonation off from the rise falls one finds in most, if not all, accents of English. Is it the form of the rise fall? Its alignment with the text perhaps? Is it its correlation with syntactic structures? Is it its frequency of occurrence? Its supposed function? Rosewarne leaves us in the dark about all of these possibilities.

The other claim that Rosewarne makes for the prosody of EE is equally vague. This is that:

the pitch of intonation patterns in Estuary English appears to be in a narrower frequency band than in RP

I would be very surprised if this were true for all supposed EE speakers and all RP speakers, but until someone does some objective research into this, perhaps we had better pass on.

To summarise then: Coggle and Rosewarne make a number of specific claims about the phonetic and phonological features which characterise a speaker as belonging to the EE community. The only features that are supposedly unique to EE are very dubious. All the rest are found in Cockney, or RP, or in some cases both.

We now leave behind the phonetics of the matter and pass on to the sociolinguistics of EE. And the first question which needs to be asked of Rosewarne and especially Coggle is: What exactly is EE? Is it an accent of English or is it a dialect? There seems to be a good deal of confusion about this in the writing of both writers. To be fair to Rosewarne he does say that EE pronunciation is generally accompanied by certain vocabulary items, suggesting that he does make a distinction between accent and dialect, but then he makes the claim that EE is marked by a greater use of question tags. This is definitely a matter of syntax and not pronunciation and as such should be a feature of dialect and not accent. Coggle's book is full of examples of supposed EE features which are dialect-based and nothing to do with accent and the confusion can be summed up by quoting one sentence from page 70:

It should now be clear that Estuary English cannot be pinned down to a rigid set of rules regarding specific features of pronunciation, grammar and special phrases.

A much more worrying feature of the description of EE by Rosewarne is its naivety. The impression left by Rosewarne's definition of EE quoted earlierand not dispelled by anything that he writes later in either of his two articles is that the relationship between Cockney, EE and RP is similar to that diagrammed below:


Here EE appears to be marked off from both Cockney and RP by rigid boundaries. It would, with a model like this, be possible to say whether a given speaker is an EE speaker or not, with 100% certainty. I am sure that neither Rosewarne nor Coggle would subscribe to this model, but neither really goes any way to providing a more realistic model of the relationships involved. Such a model must take stylistic and register variation into account.

[I <---Cockney---> F] [I <---RP---> F]
[I <---EE---> F]

Above, we can see that a model more in touch with the realities of accent variation must recognise that a speaker of a given accent has within his or her competence a range of styles from informal (I) to formal (F) and that any overlap between accents may well be as diagrammed.

Let us take a specific example: that of H-drop. In a formal style, a Cockney speaker may avoid dropping /h/, while in a very relaxed, informal style an EE speaker may drop the odd /h/. (Note we are not concerned here with /h/ elision in unstressed pronouns and auxiliary verbs which occurs in nearly all English accents). If this is the case, then the boundary between Cockney and EE becomes extremely fuzzy unless style of speech is controlled for, especially if the same situation obtains for all the variables which supposedly distinguish Cockney and EE.

To take another example, this time at the borders of EE and RP, let us look at /t/-glottaling. There is no doubt that in informal styles RP speakers do this. If supposed EE speakers only avoid /t/ glottaling in formal speech, then we have the same situation of fuzzy overlap.

All this leads to the possibility that EE is no more than slightly poshed up Cockney or RP which has gone "down market" in appropriate situations and that rather than there being a newly developed accent which we should call EE, all that has happened over recent years is that there has been a redefinition of the appropriateness of differing styles of pronunciation to differing speech situations. For example, the perception may be that it is now more acceptable to use informal style in broadcasting.

A further possibility is that the style/register fuzziness at the boundaries between EE and Cockney and EE and RP is only valid for some pronunciation variables. This leaves the concept of EE intact, but recognises that sorting out whether a particular bit of speech counts as EE or whether a particular speaker generally uses EE is no longer a simple cut-and-dried decision.

The final criticism that I have about EE from the sociolinguistic point of view is a relatively trivial one. That is the name itself. Estuary English, if it exists at all, is not only spoken on or near the Thames estuary. There is no real evidence that it even originated there.

Recent research by workers at the University of Reading, Paul Kerswill and Anne Williams, has suggested that the accent of younger speakers in Milton Keynes which is a new city quite a long way from the Thames Estuary has many of the features claimed for EE. Rosewarne and Coggle claim that EE is taking over the southeast of England and is supplanting both RP and the more localised accents of the area. An alternative explanation is that the perception of formality and informality has changed and that, in this post-modern age, it is quite acceptable to pick and mix accents. Perhaps, we ought to call this new trend Post-Modern English, rather than Estuary English. This is a suggestion I make with my tongue only slightly in my cheek.

We now turnfrom the sociolinguistics of the phenomenon to the sociology (or perhaps the politics) of it. In the discussion document already mentioned, Wells writes and I quote:

Rosewarne and Coggle have arguably done a public service by drawing attention to it (meaning EE)

I would say it was very arguable in view of the reaction of the British public to the phenemenon. If Rosewarne can invent new names, so can I. The publicity given to EE has given rise to a severe outbreak of what I shall call the Disgusted-of-Tunbridge-Wells Syndrome, or DTWS for short. For those of you unfamiliar with this, Disgusted of Tunbridge Wells is a mythical figure, very probably ex-military or married to such, retired, living in Tunbridge Wells in Kent, a town, so I'm told, amply supplied with inhabitants of this sort. DTW's main hobby in retirement is writing outraged letters to the local and national newspapers, complaining about everything imaginable, or at least everything that might imaginably be tinged with slightest whiff of left wing, or even moderate political views. One thing that DTW is hot on is the purity of the English language. The furore in the press and on the radio about EE has really given DTW something to rage about. Here are just a few quotes culled from various newspapers, some of them are from Coggle's book, others not.

I could go on with this catalogue, but I simply find it too depressing. At least one of the writers quoted above would, in my opinion, be well advised to consult a psychiatrist at the earliest opportunity. I think that it would be very difficult to justify treating anyone as a "slob" simply because they use the odd glottal stop. There must be millions of decent, honest, intelligent men, women, and children who use vocalised /l/? Are we to treat them with anything less than the respect they deserve simply because of this? And as for "our children", does anyone really believe that they are at risk of being corrupted because they hear people around them saying /r@dZu:s/ rather than /ridju:s/?

Are Rosewarne and Coggle to blame for this? I must hasten to make clear that neither of them express views anything like those that I've just been quoting. Both see EE as a positive or at least neutral, but linguistically interesting, phenomenon. What they have done, however, by giving this purported phenomenon a name and by publicising it in rather simplistic terms is, wittingly or unwittingly (perhaps half-wittingly) built the image of an ogre which threatens the imagined static, pure condition of the English language. Nothing is likely to enrage DTW more than the suggestion that the standard language which he/she holds so dear, the grail of which he/she sees him/herself the guardian, is being usurped by the usage of people who are NOT OUR CLASS. DTW is not going down without a fight, you may be sure. And as in all fights, innocent bystanders are likely to get hurt - innocent bystanders like sociolinguists and speakers of non-standard varieties of English.

The newspapers in Britain, of course, encourage debate of this sort, if it can be graced with the name debate. It sells newspapers and feeds the prejudices of their readers. I guess that the same sort of linguistic phenomena happen in all languages and in all societies of reasonable size and complexity. Is it only in Britain that news of linguistic developments generates such venom?


Placed on the web 1998 11 30

Back to Estuary English home page