Tuesday, February 05, 2008

You only think I said that

Tomorrow's topic in phonetics is the phoneme. Fun word, phoneme. It's kind of like fun-neme, or even fun-meme, but not. Phonemes have to do with the fact that what we hear in speech is often not what was said.

Speech is this highly dynamic, noisy acoustic signal, and it can be categorized, sliced and diced, in a lot of different ways. This happens most clearly with vowels. Vowels are really wishy-washy things. When people say the sound "ee", they, generally speaking (ba-dum-bump), raise their tongue towards the roof of the mouth and push it forward a bit behind the teeth. But there's no "ee" grooves keeping your tongue exactly on target, and sometimes it's a little higher, sometimes a little lower, sometimes a little more back, etc. But as long as it's close enough we hear it as the sound "ee" and hear the word "whee" and not "whoa" or "whoo" (or actual words). But, what the heck does "close enough" mean? How close do you have to be?

A lot of it depends upon the language. English happens to have another vowel that's very close to "ee" but different in a way we care about. Compare the vowels in "heat" versus "hit". The second vowel's pretty close to the first, but not quite as high and not quite as front. And, most importantly, it makes a new word. If you don't hear the difference between these two vowels then you can't tell these words apart. This is the (traditional) idea of the phoneme. They are categories of sounds, here two vowels, that make a difference in meaning. In many other languages there's no vowel like the 'i' in "hit". The next vowel is further away. And so in those languages, often, sounds which we English speakers hear as two different vowels collapse to be just one sound. Japanese is an example of this. There is no need to distinguish the "hit" vowel from the "heat" vowel and so they don't. To them, there's only one phoneme "ee" and "hit" sounds just like "heat". If you listen to a Japanese pop idol singing in English, you can often hear vowels that should sound like the one in "hit" being pronounced like "heat". And that's why.

So a phoneme is a mental category; it's a perception in your mind. In my vowel example, English had two phonemes in the area that Japanese only has one, but they are both pretty close acoustically. However, phonemes aren't just a matter of categorizing very similar things slightly differently. We hear some items that acoustically, physically are really different as the exact same thing, the exact same phoneme.

Moving to consonants, there's a group of consonants that are all made by putting the tip of your tongue to this ridge behind your teeth. Sounds made here include [t], [d], [n], [s], and [z], among others. Consonants made here are 1) really popular in the world's languages (there was a time when linguists thought that all languages had a [t], but it's not true; Hawaiian has no [t] sound) and 2) really prone to move around and vary. Focusing on [t] for a second, try saying the word "top". If you hold your hand in front of your mouth, you are likely to feel a little puff of air after the [t]. Now say "stop" with your hand in front. If there's any puff of air at all, it's very small. We actually say two different types of [t] based on whether it's the first sound in the word or the second, but we only hear [t]. Now say "cat" but don't open your mouth at the end of the word, just stop there. Still sounds like a [t] and yet this [t] is different from both of the last two [t]s because you don't even open your mouth at the end. In fact, this [t] is almost nothing but silence.

And [t] moves around even more than this.

Try saying "butter". If you have a posh British accent, your [t] there is probably pretty close to the one in "top". But if you have an American accent, your [t] likely just changed from a [t] to a little tap where you just pop your tongue a single time against that ridge and let it go. As far as phonetics is concerned, this isn't a [t] at all. Yet it still sounds like one to an American English speaker. Now try "cotton" and "button". If you articulated it very carefully, you may have just said [t] still. However, the far more common way to say this is to not use a [t] at all. Instead, you shut off all air down in your glottis/thorax/adam's apple/voice box when the [t] would occur. You might be able to hear this by saying "cotton" and stopping right at the silence of the "t" and before the [n] comes. A [t] is made, remember, by putting your tongue at the ridge behind your teeth. But if you stop yourself before the [n], which is made at the same point as [t], you will discover that your tongue didn't need to move to the ridge at all to make that consonant we call [t].

The point of all this is that phonetically, physically, there were five different sounds there, one of which never used the [t] place in your mouth at all, and yet they all sound identical to an English speaker. They all sound like the phoneme [t].

I don't know if you will believe this last one, but it's true. [n] moves around a lot as well. [n] is made at the same place as [t], with your tongue on that ridge. The difference between the two is that air goes through your nose on [n], while it doesn't on [t]. (Say a normal [n] and let it go for a while. Now say an [n] holding your nose closed. Can't keep it up for long, can you?) However, [n] loves to move to wherever you make the next consonant. If you just say just the word "ten", then your tongue will go right to the ridge. But now look at the word "boys". A [b] is made with the lips coming together and doesn't use the tongue. Now say "ten boys". Again, speaking carefully, you will put your tongue on the ridge for [n]. But if you speak at a normal pace, most people, most of the time, never move their tongue to the ridge at all for the "n" of "ten boys". Instead, they just close their lips like they need to do for the [b] that follows and let air out the nose. What's particularly weird about this is that closing your lips and letting air out the nose is in fact an [m]. (Sing, mmmm, mmm mmmm; now say mmmm with your mouth open. Can't). So we say the [m] sound, but hear the [n] phoneme. We can do this because, as English speakers, we sort of "undo" the process that moves the [n]. We know that [n]s move to the lips before [b] and so we undo that step and just hear [n].

Someone says one thing and we hear another thing. Learning how to warp reality is part of learning your native language. Learning how to undo your own warping and hear what actually occurs is linguistics training.

10 comments:

Robin S. said...

OK. I love these posts. Keep 'em coming, please. And glad you're feeling better - I'm guessing you are, or you wouldn't have the energy to be doing all of this!

I tried the 'top' and 'stop' sound/air checks. Yp- air from the first, hardly any from the second.

I tired cotton and butter - and paid attention to my tongue placement. My tongue didn't get closing to touching the 't ridge', as you mentioned. In fact, my t sounds sounds more like a d sound in these words - my guess is, it's a regional thing?

'Learning how to warp reality is part of learning your native language.' Hadn't occured to me to think of it this way. Thanks for the fresh perspective!

SzélsőFa said...

I liked this post, too!
I tried your suggested exercises. I liked 'try saying 'm§ with lips apart' the most :)))

I tried saying 'top'/'stop' however, and I had almost the same amount of air coming out of my mouth.
What was I doing wrong?

pacatrue said...

Well, I don't know Hungarian, szelsofa, but this difference is peculiarly English thing, meaning there's no reason one must pronounce the [t]s differently; it's just that English speakers learn as children to do so. Hungarian probably doesn't have such a rule and so you are actually the consistent one here. English speakers have learned to be inconsistent.

SzélsőFa said...

Oh yes, you are probably right with that Paca. Hungarian tends to say the same vowel/consonant the same way.
That, and also that most English teachers prefer using the Brit accent, posh or not, when eaching English to Hungarians.
I also think that British pronounciation and intonation is much easier for a Hungarian.

Robin S. said...

Blog alert...

the blogless troll we know now has blogging comments. I'm in shock.

ril said...

Fascinating. My surname has a double "T" in it. I find that when I visit the USA and check into a hotel, I invariably need to pronounce it as a double "D" to be understood, or they can't find my reservation. Or used to. I just hand over my credit card now and let them read it.

In London, the "TT" in my name ends up as a glottal stop in the back of the throat. Other regions pronounce the "TT" with a little tap, as you say.

Now, don't get me started on the trouble I have with my Rs.

pacatrue said...

Ril, how widespread is that glottal stop for "tt" in the middle of words. Cockney's always given as the classic dialect to say things like "bu'uh" (where uh stands in for 'er' with no r). It's not in the standard British english is it, called received pronunciation in a lot of books as practiced by the BBC.

ril said...

It's quite regional. The "Cockney" dialect really only accounts for a small area (officially, a Cockney is someone born within the sound of the Bow Bells). The "TT" glottal stop is generally associated with what's known as "estuary speak" for the region following the Thames to its estuary. It's quite prevalent in the counties around London, and generally, to some degree, class based as well (less associated with so-called middle/upper class).

Received pronunciation was a standardised English particularly prevalent in the earlier days of the BBC (1950's). It's kind of home counties, I think, but I'm getting a little out of my depth here. The use of received pronunciation, or BBC speak, is fading away and boradcasters are favouring regional accents much more recently.

McKoala said...

I love linguistics!

The glottal stop is very common in many Scottish dialects as well.

ril said...

Yeah, I was just thinking along those lines last night. As you move North, the glottal stop can move as well.

In the North East of England, the "TT" sound is pronounced with the toungue tap; in fact, if anything, the tap is more pronounced than further south. The glottal stop, in that region, has moved to the "K" or "Hard C" sound, so the phrase "Walk in a circle, pet" would have two -- one at the end of walk and another in the middle of circle. The click of the "K" sound is further back in the throat than with a more neutral accent.