Accent: are we bovvered?

Many teachers worry about what the best model accent should be, and whether their own accent serves as a suitable model. My argument is that the premise of the question is wrong – there needn’t be a single model accent, and that the teacher’s own accent will usually be the best model, providing that the teacher is an intelligible speaker of English. A variety of accents, both native and non-native, can and should be tolerated. In order to reach this conclusion, we will look at the phenomenon of accent from four different angles – the space dimension, the social dimension, the time dimension and the English-as-a-lingua-franca dimension.

The Space Dimension

It is well known that accent varies from region to region. This gives rise to popular debates about, for example, whether scone rhymes with cone or con. This, however, is a relatively trivial variation given that it affects only one individual lexical item. From the point of view of language teaching, there are much more important systematic variations. For example, the vowel sound in big is very different in London and Glasgow. To a London ear, the word big said by a Glaswegian sounds like bug. This difference was at the heart of a comedy sketch in the BBC Scotland series Burnistoun. Two Scottish men enter a lift which is operated by voice recognition. They say eleven (sounding like ulleven) and the automated voice in the lift simply says, I’m sorry, I don’t understand, could you say that again please? We can entirely sympathise with the men’s growing frustration - surely the lift can’t be so stupid; from the context, it must be expecting a number, and ulleven is not so different from eleven! We would expect a human interlocutor to be much more tolerant of minor accent variations like this. From the perspective of pronunciation teaching, we need to remember that phoneme such as /I/ represents a range of sounds which includes the first vowel in both eleven and ulleven. None of the sounds in this range is any more ‘correct’ than another, they are simply different accents.

The Social Dimension

Many Glasgow speakers pronounce Scottish as Sco’ish – where the apostrophe represents a very short silence known as a glottal stop. This is a very common substitution for the /t/ through out the British Isles, and yet it is often singled out as a symptom of careless speech. It carries stigma; it is socially marked. Yet people who criticize it very often use it themselves, unwittingly. For instance, before /n/ - in a word such as kitten, the glottal stop is more common than the /t/.

Try saying the rhyme below, first with all the bolded letters replaced by a glottal stop, and then with them all fully pronounced:

I know a little bit about kittens

I got bitten by a kitten last year

A certain little kitten in Britain

I’ve not forgotten that it bit me on the ear

The first version may indeed sound colloquial and regional, but the second is also marked: it sounds overly pedantic, perhaps, like patronizing speech to a child or someone hard of hearing. From the perspective of pronunciation teaching, the significance of this is that speakers can control certain features in their own speech. For instance, most British people can increase or reduce their use of glottal stops, and often do so depending upon how formal or informal the context is. In the context of pronunciation teaching, teachers will probably naturally modify their own accent to reduce the features that they know are likely to be less widely intelligible, and this is entirely appropriate. At the same time, however, they should make students aware of accent variations. For instance, students studying in Britain will need to be aware of the glottal stop. (See my post on glottal stops on pronpack.com)

The Time Dimension

David Crystal and other historical linguists have tried to reconstruct the accent that William Shakespeare and his contemporaries might have spoken. They have labelled this as original pronunciation, and when you hear it, it serves as a reminder that accent is not fixed in time; it evolves. One notable feature of original pronunciation is that it is rhotic – the ‘r’ is pronounced in words like ear. This ‘r’ has gradually disappeared in most English and Welsh accents, although it remains in the far south west, as well as Ireland and Scotland. It also remains in North America, and rhotic English is probably the most common globally since many learners prefer to pronounce the ‘r’ wherever it is written. Even in varieties of English where the ‘r’ in ear has disappeared, it remains psychologically, like a phantom limb, and it is always ready to reappear when the context demands, for instance in the phrase ear ache.  Given all these facts, it would be peculiar for any teacher working toward a non-rhotic model such as English English to outlaw a rhotic production on the part of their learners. Pronouncing the ‘r’ in ear does not make a speaker less intelligible – indeed, it might make them more intelligible, globally. In cases like this, teachers should be flexible – make students aware of both varieties and let them choose the one that suits them best.

English as a Lingua Franca

Pronunciation teaching has traditionally worked on a symmetrical model as regards the productive skill and the receptive skill. The idea has been that students all learn to speak in the same way, for instance General American or Received Pronunciation, and they all learn to understand that same variety – and no other. The model can be represented thus:

 STANDARD PRONUNCIATION ------ STANDARD UNDERSTANDING

I would argue that this monolithic model needs to be replaced by a asymmetrical model, especially in view of the emerging role of English as a global lingua franca. Students learn to speak with a locally appropriate accent, be it native or non-native, but they learn to understand and tolerate a wide variety of accents in addition their own. This model can be represented thus:

 LOCAL PRONUNCIATION ------ GLOBAL UNDERSTANDING

The implications of this for pronunciation teaching are that we do not need to be obsessed about the target model. Neither the teacher nor the students need to hit an idealised and perhaps unattainable target such as received pronunciation. On the other hand, we do need to make sure that students are aware of accent variation and are capable of being tolerant listeners. Everybody's got an accent. Are we bovvered?

(Based on Mark Hancock's presentation at IATEFL 2017, Glasgow)

Another article about models in pronunciation teaching here, and a series of posts on the topic of accents here.