The Average English Accent

Someone made a comment the other day which really got me thinking. They said that the Mancunian dialect (I would mainly call it an accent but it does have dialectal features) was the easiest to imitate, with which I broadly agree right now, before I’ve done research I can remember. I too would characterise the Mancunian accent as the most “average”, if that’s the same thing, and this is based on some knowledge of English accents. But what’s the truth? When it comes down to it, what does it mean to call an accent “average” and which English accent is closest to that average?

I’m going to start by discussing and defining terms. When I talk about the average English accent, I’m talking about the accents of this nation rather than North America, Australasia, Ireland, Scotland, Wales or anywhere else where English is a widespread mother tongue. This is a little unfortunate as in the US there is an accent corresponding closely to General American, spoken in a rectangle including southern parts of Illinois, where apparently call centres tend to be based in order that customers are most likely to feel they’re talking to someone they can trust because they identify with them. Something similar happens in Britain, where call centres apparently tend to use Scots, although this isn’t even an English English accent, or a single accent at all.

The mention of General American brings Received Pronunciation, also known as the Queen’s English, to mind. I’m not sure at all, but I wonder if overseas English speakers think we all have that kind of accent in this country, or if not that a generic “home counties” middle class kind of English. Although this isn’t so, features of Estuary English, referring to the Thames Estuary and therefore Kent and Essex coasts, have spread well beyond their original homeland around London due substantially to the popular soap opera ‘Eastenders’. Similarly, many of us use high rise terminals nowadays because of Australian soaps. This is where the pitch of the voice rises towards the end of a declarative sentence in a manner previously associated with a question. I first heard this in about 1980 from the American Kelly Monteith, and the first time I heard it in “real life” was from a woman from Perth in Western Australia in 1989.

Approximate Cockney might be the modal English accent due to the concentration of people in the Southeast. In my perception it extends as far as Luton, which sees itself as being in the Midlands, and the population of the combined official regions of London and the Southeast, whereof the latter surprisingly excludes Essex and Herts, is about sixteen million compared to a total English population of fifty-five million. There are variations within that region. I always used to feel at home when I got to Ashford on the train and heard the Kentish accent of the station announcer, and to me the Sussex accent sounds more like Hampshire, although both of these have been considerably eroded by London’s linguistic encroachment, in Kent partly fuelled by the railways themselves. This subregional variation also means there isn’t really a Home Counties accent at all, although they do tend to resemble each other quite closely. But if we’re talking modes, that is, the largest number of people who speak a particular way, the average English accent is, rather disappointingly, that of the Home Counties.

Even so, those sixteen million people speaking this way still leaves thirty-nine million who definitely don’t, and in particular the “Rose Cities” along the belt between the Mersey and the Humber constitute something like ten million people who definitely don’t talk like that, but also don’t necessarily talk like each other either. There are two other measures of central tendency than the mode, namely the median and the mean. It’s hard to imagine how one might go about determining the median accent as it would involve lining up everyone in England in order of accent, which is effectively impossible because accents vary in two rather than one dimension, but that does perhaps suggest some kind of weighted average which finds the centre of the English population and determines it according to how our language is spoke there. Where, then, does the average English person live? Not in the Lake District or the Isles of Scilly for a start. There are also at least two answers to this question: in terms of population or in terms of physical geography.

Like most countries (not, for example, the Maldives), England has a centroid. This is like the centre of gravity for a territory, the point where a cardboard map of England would balance perfectly on the end of a knitting needle, as it were. Our centroid is Fenny Drayton, birthplace of George Fox, co-founder of the Society of Friends with Margaret Fell, near Hinckley. This is one of those cases where you kind of regret flipping the coin, because there are several lines passing through England which separate the short and long A’s, the rounded and the unrounded short U’s and the rhotic and non-rhotic accents, and at least a couple of decades ago some of them passed through Leicestershire. Here’s the map for rhoticity:

By en:User:Unoffensive text or character – uploaded by en:User:Unoffensive text or character to en.wikipedia, CC BY 2.5,

This is actually less rhotic than I remember from my youth, as I seem to recall that Mancunian was definitely rhotic at the time. Just in case you don’t know what rhoticity is, it’s the tendency to pronounce the R after a vowel without a following vowel, as in “fur” and “force”. My own accent was rhotic for a while but isn’t any more.

Here’s a map of three isoglosses (lines delineating similar pronunciations):

From here. Will be removed on request

It probably isn’t obvious from this map to anyone who doesn’t leave in this area, but Leicestershire nestles in the tangle of isoglosses on the east side. These lines are not static either. Generally they move in a roughly northwesterly direction at a rate of a few kilometres a decade for fairly obvious reasons.

It’s notable that rhoticity is different from the other features because it follows an east-west divide rather than a north-south one. It surprises me rather that Northumberland and northern Cumbria are non-rhotic, but the isogloss along the Solway Firth is said to be the sharpest in the English-speaking world. Getting back to the point at issue though, the centroid is not very far from the kink in the red isogloss just southwest of Northampton. It would mean that the accent in Fenny Drayton is a mixture of northern and southern characteristics, as it should to be representative of the country as a whole, but were it just a few kilometres in several directions from it, it would be very different, and I feel like it’s just chance really that it happens to be there.

There are some other places which can be considered the centre of England. Meriden is traditionally considered the centre, and is in Solihull. Solihull, like the rest of the West Midlands, has quite a distinctive accent, which however varies dramatically between the Black Country and Birmingham, and is considered to be the stereotypical Midlands accent. It bears little resemblance, though, to the East Midlands accents and Solihull people in particular often have a markedly southeastern sound to them in my experience. Also, the Ordnance Survey proved that Meriden was not the centre of England in the ‘noughties.

By Mat Fascione, CC BY-SA 2.0,

The crossroads of the ancient Fosse Way and Watling Street is at High Cross, and is also considered by some to be the centre of England, possibly for that reason. Since this is associated with Roman Roads, both of which I’m familiar with by living on or near the first and having the second pass through my home town as a child, it’s probably in order to mention that because of rising sea levels and reclamation of land from the sea, the centre of England as defined in this way moves around, and strictly speaking must drift somewhat through the course of any twenty-four hour period due to the tides. Speaking of the sea, another issue is that of islands considered part of England, in particular the aforementioned Scilly Isles, which are around sixty kilometres from the rest of Cornwall. Berwick and Monmouthshire, with their changing and in the latter case indefinite status, constitute another issue, as would Cornish nationalism.

The point farthest from the sea in Great Britain is Coton In The Elms in Derby. Goitres, caused by underactive thyroids and iodine deficiency, used to be called “Derbyshire neck” because of the lack of iodine in the soil in this area, which is due to the fact that it’s been far from salt water for such a long time. However, this is Great Britain rather than England, although notably all of these places are quite close to each other.

It’s probably more sensible to consider the centre of population. This is the point nearest the largest number of people, and can be defined in several ways, once again due to the measure of central tendency used. The mean centre is the centroid, but instead of being the centre of gravity on a normal map, it’s more like the centre of a massive “spider” of lines radiating from a point out to everyone in England, or, put another way, the centre of gravity of a map of England whose density at different points reflects the population density of this nation. This can be found mathematically by finding the point with the lowest sum of the squares of the distances to each person in the country. However, this point does not represent the minimum distance to each person in England. This could, however, be more closely approximated by taking the median latitude of the population and combining it with the median longitude of the same. Finally, the geometric median does locate the minimum distance of travel and takes the form of the smallest possible sum of distances, which is unfortunately not calculable over a finite number of operations. The point would come, though, when the distance from the real value became trivial.

The centre of population for the whole of Great Britain is gradually moving southeast, and this is of course not England. In 1901 it was in Rodsley, Derbyshire, and a century later was in Appleby Parva, Leicestershire, a distance of around thirty-five kilometres, meaning that it’s been moving at about three hundred metres a year, or a metre a day, to be facetious. The Great British centre of population moves unusually slowly compared to many other areas, and this is thrown off in particular by the central belt of Scotland and the diaspora into England, which resulted in my own birth in this country along with several friends and acquaintances. Even so, there is also a similar southeasterly movement in the English centre of population since Scotland can only make a maximum difference of about twelve kilometres, assuming that everyone in Glasgow and Edinburgh moved to the English Midlands.

While this centre moves slowly southeast, the influence of London on our accents moves northwest more rapidly, meaning that in all probability the average English accent is becoming more like Cockney more quickly than the movement of the population might suggest.

None of these ideas of locating the centre of England, either by population or physical geography, correspond to my original idea of how to determine the average English accent. The distance of travel necessary to a particular accent is not the same as the same quantity in kilometres. There are also a few features shared very widely by non-RP accents, such as H-dropping, which one might therefore expect to hear from such a speaker, and other very rare features such as the erstwhile insertion of L’s in Bristol after final vowels which would doubtless be absent. It made most sense to me to consider the following features:

  • H-dropping.
  • Rhoticity.
  • The pronunciation of the U in “bus”.
  • Short or long A in “bath”.
  • Closer or more open short O.
  • Rolled, tapped, uvular, retroflex or labialised R.
  • WH-voicing.
  • Voicing of intervocalic T/glottal stop for intervocalic T.
  • Distribution of clear and dark L’s/vocalisation of dark L’s.
  • Labialisation of TH.

In writing that list, it occurs to me that it may be rather Home Counties biassed, and it doesn’t take account of suprasegmental features such as the insertion of R sounds in vowel-final position and the odd “flowing” nature of the Leicestershire accent which I can’t put my finger on, but this is complicated enough as it is.

Starting near the end, the labialisation of TH (“fink” for “think”) is confined to the Southeast, but another common feature of that area, H-dropping, is widely distributed, so one might expect an average accent to drop aitches, possibly hypercorrect haitches in some positions, but not to labialise TH. Labialisation of R is unique to the Southeast as far as I can tell, and although it seems quite common, I have strangely never seen a professional linguist or phonetician comment on it. Judging by the first map, rhoticity is present in Cornwall, Devon, Gloucestershire, Bristowe (Bristol), Somerset, Wiltshire, Herefordshire, Oxfordshire, Dorset and Lancashire. It has gone AWOL in Cheshire and Salop, possibly due to the use of the Welsh Knot and the influence of Walian English, which is non-rhotic due to Sassenach cultural imperialism. That’s around eight million people, so it doesn’t make it as part of “the” English accent.

Next come the classic North-South divide vowels A, U and to a lesser extent O. The distinctively Northern ways of pronouncing these are practically universal throughout Yorkshire, County Durham, Northumberland, the Northwest and are also found in the northern Midlands. This is over fifteen million people, although it’s difficult to quantify due to the isoglosses cutting across the middles of counties. The most southerly of the northern features is the absence of the A-like pronunciation of short U in certain words, which might swing the “oo” pronunciation just into the majority. Besides any of these is the treatment of long A, which is an open E in the Northwest and a closed long E in other parts of the North and certain parts of the Southwest. The long O, which has recently become a diphthong beginning with schwa in the South, and the Middle English long U which has become a long A in the South and an “ow” sound elsewhere, are probably in a small enough minority not to be considered average.

Hence the “average” English accent probably has the following features:

  • H-dropping.
  • Voiced WH.
  • Rounded short U.
  • Long A in “bath”.
  • Non-rhotic. R’s retroflex.
  • Diphthongised long O starting with schwa.
  • Unvoiced alveolar intervocalic T.
  • Separate clear and dark L’s with no vocalisation.
  • Open short O.
  • Diphthongised Middle English long U.

Such an accent might currently be expected to be found in the East Midlands, possibly in Northamptonshire away from Corby, which has a large Scottish population.

Presumably then, the easiest accent to hear and speak for the majority of English people is quite close to the centroids and geographical “centres” of England, on the same latitude but further to the east. This surprises me because I really expected it to be Mancunian, but in fact it isn’t.


2 thoughts on “The Average English Accent

  1. Americans certainly seem to think the ‘English’ speak in RP (which, btw, is an ‘invented’ form of speech, which no-one speaks ‘naturally’). Susan Boyle gets subtitles on American television: as do the Gallagher brothers, I’d imagine. The film ‘Kes’ caused widespread incomprehension when it was released in America: reportedly, literals no-one could be found who could ‘translate’ the dialect for the subtitles. I’ve often wondered whether people from Alabama, etc, get subtitled when they are interviewed for North American television – I’d imagine that would happen.

    I strongly dislike Australian Questioning Intonation – or the ‘moronic interrogative’ as Rory Bremner once called it.


    1. I’m prepared to believe that there’s enough variation in English for subtitles to be necessary between sufficiently different accents and dialects.

      Just on high-rise terminals, it’s a linguistic universal that question sentences involving intonational changes always have them at the end, and it occurs to me that the likes of the Finnish “ko”, Japanese “ka” and German “oder” as markers of questions also occur at the end of the sentence, so sufficiently widespread use of high rise terminals may obscure that function. Then again, maybe questions will be signalled by a drop in intonation one day!


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s