Like many people, I’ve been doing some language study during the lockdown. On my daily walk, I’ve been listening to a Greek course delivered entirely through audio, with no reference whatever to the written form. It remains to be seen if this has worked, since I haven’t had the opportunity use it. But one thing I can be fairly sure of is that my Greek pronunciation won’t be subject to spelling-induced error.
Learning through the eyes
Most learners of English, certainly the ones I’ve come
across in my teaching career, have not learnt entirely through the ear as in my
Greek experience. Most have learnt through the eyes too. They’ve used books,
not just audio, and most have been both guided and misled by the written form. We
shouldn’t underestimate the importance of spelling in teaching pronunciation,
and not only for the productive (speech) skills but also for the receptive
If you are learning a language which shares an alphabetic
writing system which resembles that of your mother tongue, the first problem
you’re likely to encounter is L1 interference. If you’re an English speaker
learning French, you may pronounce an /s/ in Paris simply because you’re following your L1
spelling rules. Most learners quickly become aware of how the target language
rules differ from the L1, however, so this tends to be more of a beginner’s
issue, relatively quickly overcome.
A more serious issue is when the spelling of the target
language doesn’t ‘match’ the pronunciation, as in the case of English. Learners
may wonder what kind of sadist invented spelling patterns like ‘ough’ in words
like rough, through and bought, for example. And who
thought that it would be a good idea to have a ‘k’ in knife and a ‘b’ in
lamb? On the other hand, there’s often a lot more reason in the madness
of English spelling than first appears, and learners can be taught (or simply
pick up) rules and patterns such as the ‘magic e’ which distinguishes made
from mad, or the doubled consonant that distinguishes bitter from
A hybrid language
English spelling is also complicated by the fact that so many words have been imported from other languages, either historically or more recently. It helps to know that ballet comes from French if you are to avoid pronouncing a /t/ at the end of it, and you could be forgiven for thinking that quiche is pronounced like quickie if you don’t know the word is imported.
But serious though some of the above difficulties are, I
think people tend to be generally aware of them. The most serious problems are
the ones you don’t know about, and here I make a special mention of connected
speech and the difficulty that it creates for listening.
Learners, and sometimes teachers too, tend to assume that
spoken English is like written English, but audible. Not so. In the spoken
version, words tend to get mangled and distorted, especially the high-frequency
grammatical ones. So, for example, the phrase ‘cooks and meals’ can
surprisingly end up being a perfect homophone of ‘cook some meals’, despite the
middle word looking completely different in the written form.
Hidden in plain sound
Although features of connected speech pervade the spoken
language, many people are only aware of those which have found representation
in the written form. These include ‘official’ contractions such as don’t or
would’ve, but also more ‘unofficial’ folk spellings like wanna or gimme.
However, these are really and truly only the tip of the iceberg. On closer
inspection, these kinds of reduction are to be found throughout the ‘sound
substance’ (Richard Cauldwell’s term) of spoken English, but speakers of the
language are very often unaware of them.
The power of knowing
So what to do? My feeling is that we very much need to make
learners (and teachers) aware of the features of connected speech, bearing in
mind that this is mainly for receptive purposes. Learners who are unaware tend
to blame their listening difficulties on their own ears, and this is the road
to despair. But when they learn the reality of linking, elision, assimilation
and so on, the reaction tends to be something like,‘Oh, now I understand why I
don’t understand!’ That to me sounds like the beginnings of empowerment.
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.
I argue for a more democratic use of the IPA symbols in this article, which first appeared in Modern English Teacher 29.3 (July 2020).
There seems to be a love-hate relationship between the
English teaching community and phonemic symbols. Some teachers love them,
others won’t use them. A third category uses them under duress, but do not feel
comfortable in doing so. This article is for anybody, but especially teachers
in this third category.
A teacher’s feelings about the material they’re teaching with
are often perceptible to the class. If a teacher uses phonemic symbols with
confidence and pleasure, that will communicate itself to the class. If on the
other hand a teacher uses them but dislikes doing so, it’s likely that this
dislike will be part of what the learners take home from the lesson. I would
like, if at all possible, to help teachers feel more at home with the symbols –
to feel as if they ‘own’ them.
What is the IPA?
To begin with, we need a little background on where the
phonemic symbols come from – the IPA (International Phonetic Alphabet). This
alphabet was constructed by the International Phonetic Association, which was
founded in Paris
in 1886. The full set of symbols was intended to be able to represent all
meaningful sounds in all of the world’s languages. The use of IPA symbols specifically
in English language teaching only came later.
When symbols are repurposed to represent meaningful sounds
in one specific language, they are known not as phonetic symbols but instead,
phonemic. A phonetic symbol can be thought of as objectively representing a
specific sound. A phonemic symbol, on the other hand, is more subjective – it
represents what the speakers of a specific language think is a specific sound, even though it might in fact be a range
of sounds. For example, speakers of English may think all instances of the
phoneme /t/ are the same sound, whereas in fact the /t/ in top is slightly different from the /t/ in stop. A phoneme may vary across accents too – the /t/ in city will be different in the accents of
London and Los
Angeles. The same /t/ symbol can be used to represent
all of these sounds.
The IPA as symbols of power
When the symbols of the IPA were first repurposed to
describe British English, the assumption was that the phonemes should represent
the accent of power at that time – namely RP (Received Pronunciation). It seemed
obvious that anybody who was learning the language would want to aim for that
specific accent, since it is the one which British listeners would find most
acceptable. And it wasn’t only language learners who might aim for RP –
speakers of regional accents may also want to “improve” themselves by
acquiring it. Hence the popularity of the kind of elocution lessons which were
the subject of Shaw’s play Pygmalion, written in 1913.
From the start, then, a strong bond was formed in people’s
minds between the IPA and prestige accents such as RP: the letters of the English
phonemic alphabet became in effect symbols of power. We should avoid the
temptation of thinking that such accents are somehow “better” though. As John
Wells (1982, p34) points out, ‘A standard accent is regarded as a standard not
because of any intrinsic qualities it may possess, but because of an arbitrary
attitude adopted towards it by society’.
Updating the symbols?
Over time, accents have evolved, as have social attitudes.
RP has acquired a dated and elitist flavour, and linguists have sought to
update it. For example, Geoff Lindsay in his book English After RP (2019) describes how the standard Southern British
English has changed in recent decades, and he proposes some important
modifications to the phonemic alphabet we use in teaching it, to bring it more
in line with phonetic reality.
Lindsay is persuasive in his argument for moving beyond RP
and updating our description of the accent which may be considered as ‘standard’.
However, there remains the assumption that pronunciation teaching is about
getting learners to conform to a specific standard native accent – in other words,
that it is accent training. Do learners really need the degree of phonetic
detail that he is suggesting?
Nativeness versus intelligibility
According to John Levis (2019), pronunciation teaching can
be based on one of two contrasting positions which he calls the Nativeness
Principle and the Intelligibility Principle. The kind of accent
training envisaged by Lindsay seems to be based on the Nativeness Principle,
the assumption being that learners should try to sound like native speakers. Teachers
who follow the Intelligibility Principle, on the other hand, take the view that
the accent doesn’t matter as long as it is intelligible. Improving
pronunciation is not about sounding “better”, whatever that might mean, but
about being better understood.
As Jennifer Jenkins (2000) has pointed out, English today is
different from most other languages in that it is a global Lingua Franca, and
pronunciation teaching needs to take this into account. We need to break the
link with prestige native accents such as RP and instead focus on
intelligibility. If we are to use IPA symbols in our teaching, then the
perceived link between these symbols and RP needs to be broken too.
I wonder if the perceived IPA connection to RP is actually a
confusion of phonemic and phonetic. As I have said, unlike phonetic symbols, phonemic
symbols do not represent specific sounds but a range of sounds. In that case,
they can be flexible enough to accommodate a fair amount of accent variation.
Let me illustrate that point with a specific example.
A case study: the duck vowel.
As a trainee teacher, I had trouble with the phonemic symbol
for the vowel in duck. They symbol is
an upside-down V, known as the turned V.
I was told that the vowel sound in my Manchester
accent did not match this symbol. I think that was wrong, at least if we are
following the intelligibility principle. The turned V is both a phonetic symbol
and a phonemic symbol, and that may be the source of the confusion. As a
phonetic symbol, it represents a specific sound which is not present in my
accent (and nor is it present in RP, in fact – which may come as a surprise to
some teachers). As a phonemic symbol however, the turned V can represent
the duck vowel in both RP and
accent. It is not specific to either. The idea that it only represented the RP
version was merely an arbitrary convention.
A phonemic chart as a set of oppositions
If we are following the Intelligibility Principle, I think
the best way to view the phonemic chart is not as a set of sounds, but as a set
of oppositions. The turned V represents the vowel in duck as opposed to the one in dog
or the one in cat, for example. The
actual quality of the vowel sound in duck
doesn’t matter much, as long as those oppositions are maintained.
Some teachers worry that a sound in one accent my overlap
with a different sound in another. For example, the vowel in New Zealand pat
sounds like the one in UK
pet. Does that mean that a New
Zealander can never distinguish pat
from pet? No it doesn’t, because the
vowel in New Zealand pet in turn sounds different (like the
one in UK
pit, in fact). This means that the
distinctions are maintained, and speakers with these accents are mutually
intelligible once the ear has adjusted. The same phonemic symbols can be used
for the equivalent vowels in both varieties of English, even though they sound
Where problems may arise
IPA symbols can not easily accommodate all accent variations
however. There is a problem when two phonemes are conflated in a given accent.
For example, for many Northern English speakers, there is no difference between
the vowel sounds in putt and put, so for this accent, we have a
redundant symbol in our set. Fortunately for the purposes of intelligibility,
few words are distinguished by this vowel opposition anyway, so it is not a big
issue. However, when using the chart, a Northern English teacher would have to
explain that in their accent, those two symbols represent the same sound.
This raises the question of phoneme conflations in our
learners’ accents. If, for example, our learners pronounce the same vowel sound
in both ship and sheep, is that a problem? We can’t know for sure. It may be the
case that our learners can make themselves pretty well understood without making
that particular opposition. The problem is that with every opposition lost, the
space for misunderstanding grows. It’s like pixels in a digital photo – every
pixel lost makes the image a little less clear. So, as a rule of thumb, I would
suggest that we aim to help our learners produce and distinguish all of the
oppositions in the IPA phonemic chart, unless we have evidence to suggest that
a given opposition doesn’t matter.
The exceptional case of schwa
There is one important phoneme conflation which I think
requires a special mention and that is the overlap between the schwa and the duck
vowel. I mentioned above that the turned V symbol represents a wide range of
possible sounds, depending on the speaker’s accent. Among these is one which is
identical to the schwa. The presence of both the upturned V and the schwa
symbol on phonemic charts can raise awkward questions in the classroom, and
teachers can waste a lot of time trying to demonstrate a difference that
My pragmatic solution to this problem is to treat the schwa
as a special case – as a vowel sound which is not of the same kind as all the
others. I would present it as a potential variant of any vowel when it is
reduced, in an unstressed syllable. Take the name London, for example. The first vowel
sound is the duck vowel and the second is the schwa, and the difference
is not in the sound quality, but rather in stress – the first is stressed and
the second is unstressed. In other words, I suggest that being unstressed is
the defining feature of the sound represented by the schwa symbol.
The IPA as symbols of empowerment
By way of summing up, I would like now to return to the big
picture. When the IPA symbols are attached to a specific prestige accent such as
RP, they are symbols of power. Learners and teachers whose accent does not
match will feel that they are deficient even when they are perfectly
intelligible. If we can break the assumed link with RP (or indeed any specific native
variety), then the symbols can be used more flexibly. They can match up to the
phonemes of any intelligible speaker of English no matter what their accent. A
non-standard or non-native speaker can make the symbols their own and teach
pronunciation with no sense of inferiority. The IPA letters can become symbols
Jenkins, J. (2000) The
Phonology of English as an International Language Oxford, OUP
Levis, J. (2018) Intelligibility,
Oral Communication, and the Teaching of Pronunciation Cambridge: CUP
Lindsey, G. (2019) English after RP: Standard British
Pronunciation Today Palgrave Macmillan
Wells, JC (1982) Accents of English 1: An Introduction
Cambridge: Cambridge University
Why the myth about ‘native speakers’ being better language teachers? If you want to learn a language, you’d be well advised to look for a qualified teacher who shares the same first language as yourself, or at least knows it well. I can testify from my own experience that teaching pronunciation to someone whose language I know nothing about is like wading across a river when you can’t see the bottom: you just don’t know where the slippery boulders are! I will call this situation, ‘teaching blind’.
For a Spanish speaker, it seems perfectly natural for a [w] sound to be closely associated with a [g]. The Spanish dictionary entry for hueso, for example, gives the pronunciation as [(g)wéso], the bracketed (g) indicating that it is optional. Similarly, there’s an optional [g] at the beginning of huevo (egg).The g may be explicit in the spelling too. For example, the loan word for whiskey is often written güisque. For an English speaker, by contrast, this g comes as a surprise, and teachers are mystified as to why Spanish learners sometimes insert a g before words that begin with a w.
Delighted to announce the upcoming release of PronPack 5: Pronunciation of English for Spanish Speakers! This new addition to the PronPack collection takes a different approach by focusing sharply on the needs of learners from one specific language background – namely Spanish. Do your learners drop the consonants at the end of words? Put a ‘g’ before ‘w’? Confuse ‘b’ and ‘v’? Switch ‘chicken’ and ‘kitchen’? The lessons in this book focus on issues like these. The objective is to help learners make themselves more intelligible.
Our estimated publication date is September 1st 2020, available in paperback on Amazon and as an eBook on iBooks and KOBO.
If you’re an English native speaker in the domain of ELT,
you have the privilege of being sought after. In matters of the English
language, you are seen by many people to be the best kind of expert there is, and
there’s a market for that. Such people will defer to your judgement on what’s
wrong or right. They will prefer you as a candidate for teaching jobs. They
will display an unusual level of interest in the minutiae of your home culture.
They will marvel at effortless vowels. They will pay you for being able to
” I have been using at my school the series PronPack with excellent results as the activities do not just shift the focus from the traditional teaching of grammar and vocabulary to phonology, but also the worksheets are fun, engaging and meaningful. And above all, students love them! “
(MA TESOL, PhD Applied Linguistics, ABD) is a teacher, teacher
trainer, researcher and school owner, promoting alternative and
innovative approaches to teaching and implementing them in her own