Pronunciation Models and False Choices

(This article first appeared on the Cambridge ELT Blog)

Which English do you want – American or British? This is a question which students (or their parents and other stakeholders) typically face at the beginning of a language course. As regards pronunciation, this usually means a choice between the two ‘standard’ accents General American (GA) or Received Pronunciation (RP). It’s as if these accents are products on a supermarket shelf and it’s simply a matter of choosing. But I think presenting a choice in this way is misleading, and here’s why.

Pronunciation teaching, not accent training.

I think some people assume that pronunciation teaching is about helping people to ‘improve’ their accent, so they sound ‘better’. That’s not how I see it. That’s what I would call ‘accent training’ or ‘elocution lessons’. To me, pronunciation teaching is not about sounding ‘better’, it’s about being better understood. In other words, it’s a question of intelligibility. To be widely intelligible, it isn’t necessary to acquire any specific accent. For example, you can be clearly understood while still having an accent which is non-‘standard’ or non-‘native’.

A goal, not a target

A target, such as a darts board for example, is a good metaphor in accent training. A shot at the target may be good, but if it doesn’t hit the bull’s eye it’s less than perfect. If your phonemes are not identical to those in the target accent, they are imperfect according to this metaphor.

In pronunciation teaching, I think a better metaphor is a goal, as in football. The player may aim to kick the ball into the centre of the net, but if it goes into the corner of it, it’s still a goal, and no less ‘perfect’. A goal is a goal. If your intended phonemes are recognised as such by your listener, they’ve succeeded, no matter whether or not they match the phonemes of any particular model.

Reference models or attainment standards

Published materials, including dictionaries, often give pronunciation guidance in the form of audio recordings and transcriptions. Obviously, these can only realistically represent one or two accents, and typically GA and/or RP are chosen. However, we must remember that these are reference models rather than attainment standards – you can’t say that somebody whose pronunciation does not match these models is ‘incorrect’. Other alternatives may be perfectly intelligible. A reference model provides something for students to aim at, but if their shot goes wide, it’s not a problem providing that the ball still goes into the goal.

A false dilemma

Returning to the question ‘American or British?’, I think this is a false dilemma. It implies there are these two alternatives and no others. It conceals the fact that there are many different accents within the USA and Britain, and many more outside of these two countries, and many of these would be equally valid if students need a model to aim at. The choice of a model is much less crucial than is often implied – it may not even be a real choice.

Process, not product

I think that the problem with over-emphasising the importance of a model is that it focuses on product at the expense of process. Let me explain what I mean. It is widely acknowledged that very few language learners end up sounding exactly like a ‘native’ speaker. Why is this? I think it’s because teachers can’t simply put pronunciation into the head of the learner like some kind of implant. The learner’s accent is something they grow for themselves through their own efforts, albeit with a teacher’s help. This process is influenced by a range of factors such as their motivation, the range of language samples they are exposed to, their teacher, and last but not least, their own L1. The exact combination of factors will be different for each individual learner, and the accent that emerges as a result will be very much their own. And this is the case whatever the model they aimed at.

English as a Lingua Franca

It’s often assumed that the best model is a ‘native’ variety. I can see how that may seem an obvious choice. For example, if you are learning Turkish, a ‘native’ variety makes sense – there may not be any alternative available in fact. However, English is different in kind. It is a global Lingua Franca; it doesn’t ‘belong’ to any particular nation. And crucially, there doesn’t appear to be any evidence that any ‘standard’ national variety of English, native or otherwise, is intrinsically more intelligible. For this reason, the evaluative connotation which is often attached to the terms ‘standard’ and ‘native’ in English pronunciation teaching contexts seems entirely inappropriate. That’s the reason for the quotation marks around those terms in this post.

The elephant in the room

In discussions of pronunciation models in English teaching, I feel that there is an elephant in the room – a factor which is critical and yet ignored. I’m talking about the teacher. Surely, of all the models that learners are exposed to, the teacher is number one. So what happens if the teacher does not speak with one of the main ‘standard’ accents – GA or RP? That’s the majority of us, surely!

The teacher is the model ‘American or British?’ is a false choice, because in the end, what the students get most exposure to is the teacher. The teacher doesn’t suddenly become American or British just because the student made that selection at the start of the course. Most teachers will teach in their own normal English accent, and I think we should accept and embrace that reality. If a teacher is competent and intelligible, it doesn’t matter what their accent is.