Phonology -- whats and hows

A couple of weeks ago, I was discussing the importance of phonology, trying to demonstrate why it should be consciously dealt with in the teaching/learning process, but I took the decision not to include any comments on how to teach it in that article.  Basically, I didn't want to give anyone any grounds to reject my argument out-of-hand.  In this post, I'd like to cover how I believe it should be taught, but remember that this, the how, doesn't affect my argument on the importance, the why.  Reject my methods if you want, but please don't reject phonology as an area of study.

So, what did I establish in the previous post?

• Incorrect pronunciation of an individual phoneme leads to problems in pronouncing clusters with that phoneme.
• Problems in pronouncing certain sequences of phonemes lead to grammatical errors.
• That vocabulary is harder to learn when you're not familiar with the rules of pronunciation in a language.
• That not understanding target language phoneme boundaries makes it hard to understand native speakers.
• That sounds that the learner drops in speech are often matched by a dropping of the corresponding letters in writing.

These are things that I have observed and do not see as particularly controversial.  And yet, my conclusion that pronunciation requires active instruction is rejected by many teachers.  Accent, they say, will take care of itself.  And accent, they say, is a personal thing.  But we're not talking about accent.  Accent is something that is layered on top of phonology.  Phonology is like the basic letter forms in writing, accent is more like individual differences in handwriting.  At school we are taught initially to get the basic forms right, and over the years we develop our own personal "hand".

Can we learn pronunciation from listening?

Some even argue that we learn pronunciation from hearing (and they sometimes add "just like children").  However, as I tried to demonstrate in my recent post receptive skills as a reflective act, there is good reason to believe that we understand language by comparison to our own internal model of the language.  In the follow-up post, I gave a concrete example of mishearing a word on Italian radio, and how my flawed internal model was good enough to understand the message without perceiving every sound.

OK, so that's anecdotal and doesn't prove a general case, but ask yourself this: how many different accents can you understand in your own language?  And how many of those accents can you speak in?

So you can see that simple exposure hasn't given you extra accents.  As I said above, accent is not phonology.  But our brains have learned to ignore accental differences (to an extent) to enable us to understand the widest possible number of people around us.  So if our brain assumes a different phonology is just a different accent, it throws away all the information you're supposed to be learning from.

So I really don't believe it's possible to learn from "just listening", no matter how much you do.

Motherese and exaggeration

Here's the outcome of an interesting study (YouTube video).  It turns out that when we teach kids to speak, we don't expect them to learn from natural speech, but we exaggerate our phonemes, effectively making them "more real than real" or "whiter than white".  And if you think about it, isn't this what we do when speaking to foreigners or people with a very different accent from ours?

The point is that we have to make the differences clear and noticeable, so that one phoneme doesn't blend into another.

I would suggest that this points towards the right answer in language teaching to adults: if even children (who have no preconceptions of what a phoneme is) need extra emphasis to understand the difference between similar phonemes, then us adults (who are biased towards our native language's phonology) really could do with a bit of help.  The brain has to be told that this new information is useful, or it will throw it all away.

Exaggeration of pronunciation appears to help the listener notice the differences.

Learning pronunciation through pronouncing

However, we learn to dance by dancing, and we learn to drive by driving.  In both cases we can pick up a few hints and tips from watching, but we need a heck of a lot of practice.  Why shouldn't this be the case with language?

People are very quick to tell me that language is different from every other skill.  That is a valid opinion, but it is still only an opinion - no-one has ever presented anything to me that demonstrates it to be true, or even likely.  Right now, it's just a theory... and it's one I do not believe.

To me, pronunciation is a muscle skill.  Let's consider some of the extremes sounds that don't occur in English.

Take retroflex consonants.  Retro - backwards; flex - bend.  In retroflex consonants, your tongue bends backwards, and the tip goes behind the alveolar ridge.  This type of sound doesn't occur in English, so a monolingual English-speaker will probably never produce this sound in his life.  If you ask such a person to put their tongue into that position, they won't be able to -- their tongue just can't bend that way.

But then your average person couldn't do yoga postures on a first attempt either -- the yoga teacher will lead them through some simple postures and exercises to encourage the muscles to stretch and strengthen appropriately until they are capable of performing the required movements.

The brain doesn't prepare the muscles just because you've seen the movements; the body prepares the muscles once you've started doing the movements.  Your brain similarly cannot train the tongue as it's just another muscle, after all -- only the body can do that.

So clearly, there are certain sounds that must be taught consciously, or the learner won't physically be able to say it.  But obviously there are also sounds that the learner is physically capable of saying, but isn't in the habit of saying.

This post is starting to get a bit on the long side, so I'll come back to the question of this second category of sounds next time.

How I learned to pronounce retroflex consonants

I had a notion to learn a few words in various Indian languages a few years ago when I was working in IT support.  Our front-line helpdesk was in India and I wanted to try to build a better rapport with my coworkers.

One of the sources I used stated quite plainly that while languages like French and Spanish let you get away with "close enough" pronunciation (not entirely true...) with Hindi, you would simply not be understood if you spoke in an English-speaker's accent.  It described the retroflex articulation and what I did was to start doing a regime of "tongue stretches" -- as I walked to and from work, I would tap my tongue continually off the roof of my mouth, and move it slowly backwards and forwards, to create a sort of silent T-t-t-t-t-t-t-t-t-t-t or D-d-d-d-d-d-d-d-d-d.  Every day I could reach slightly further back, and in about a week and a half I was able to produce a convincingly Hindi-like retroflex for all of the various consonants (except R, cos that's really quite complicated). I was curious about how far I could go, and within another few days I'd got to the point where I could touch the tip of my tongue to my soft palate.


So certain sounds need to be learned physically, and it's something that can be done.  Next time, I'll start looking at sounds that are more a matter of habit, and showing that the boundary between "habit" and "ability" isn't always that clear.