voiced/unvoiced - the real story?

<b>Forum for the discussion of Applied Linguistics </b>

Moderators: Dimitris, maneki neko2, Lorikeet, Enrico Palazzo, superpeach, cecil2, Mr. Kalgukshi2

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

voiced/unvoiced - the real story?

Post by woodcutter » Mon Mar 10, 2008 12:59 am

Now class, a typical pronunciation exercise:

Put your hand on your throat and make the sound "t" (tuh). Your throat does not vibrate. This is an unvoiced sound.

Keep the hand there and make the sound "d" (duh). Feel the vibration? This is a voiced sound.

Now, whisper softly
"Teacher is a cheat. He said a short vowel only after the d!"

Now, you have just whispered a sentence. When we whisper, we use an unvoiced airflow. Und yet, the sentence you made, full of so-called "voiced" sounds, was perfectly easy to understand. I put it to you that any language in which you can whisper and be easily understood - every language? - is a language in which voicing is basically used just to provide volume. You can voice/unvoice randomly through a sentence and be perfectly understood.

You can run through the alphabet demonstrating the sounds
A-ah B-buh C-kuh D-duh
all in a whisper, and each letter will be distinct. This is because there are no consonants distinguished from others only by voicing. The position of the tongue and the mouth, the tensing and the timing all show slight variations each time. If you keep any of these constant through a sentence, your ability to communicate will be impeded. Saying everything with an unvoiced airflow, on the contrary, only makes you quiet.

Now with z/s, v/f and th/th, (all sounds which we can continue for a length of time) voicing is helpful to distinguish. However, it isn't crucial. You can whisper zoo/sue, thy/thigh or vie/fie, and we can hear the difference.

Strangely though, we usually use stops to demonstrate voicing. Stops, of course, stop the airflow. There is thus nothing to voice. A voiced stop is a contradiction in terms. What matters is how we release - whether we have "aspiration". If you say tan/Dan, you can make them perfectly distinct in a whispered voice. However, if you try and say tan without a hissy explosion after the t, without aspiration, you will see it is almost identical to Dan. Aspiration is certainly the key feature here.

So the exercise with which we began our lesson is a confusing, blind-leading-the-blind kind of exercise, no?

blackmagicABC
Posts: 37
Joined: Mon Feb 18, 2008 4:37 am
Location: Taiwan

Post by blackmagicABC » Mon Mar 10, 2008 6:07 am

For now, unfortunately, there isn't an easier way to explain this to elementary school kids. Trying to explain it in too much detail is counter productive.
An interesting experiment I did led to some interesting results for me. I own a school so doing this was my decision. I have two classes at the same level. They started in September last year and are mostly grade 1 students with almost no prior knowledge of English. When we did "oo" in phonics I explained the difference to one class and not to the other purely to see the difference a week later when they had class again and needed to read a list of words and short sentences.
Both classes were about equal in correctness. It is almost as if the sound (long or short) happens naturally when you don't force it. I by no means consider this a valid experiment. It was just interesting.

fluffyhamster
Posts: 3031
Joined: Tue Oct 26, 2004 6:57 pm
Location: UK > China > Japan > UK again

Post by fluffyhamster » Mon Mar 10, 2008 6:15 am

Hi woody, did you write that? If so, I kind of know what you mean (I wondered sometimes at the distinctions that were made regarding the sounds of Mandarin, when I was first learning Pinyin).

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

Post by woodcutter » Mon Mar 10, 2008 7:17 am

Yes, I wrote it. I put in some extra effort, so maybe it is not typical of my style! (my mother considers me illiterate on the basis of what I e-mail off to her, by the way).

Blackmagic, I'm not really arguing about the worth of teaching phonics in general, matters such as the length of vowels, though it is worth discussing, and is relevant here in that we do tend to make an overly big fuss over just one factor.

I'm saying that we teach and talk about voicing in particular in a ridiculous way, and that even the best teachers don't even know themselves what they are on about. We talk about voicing when we should talk about aspiration - I'm all for white lies, but the truth in this case would be simpler for any student!
Last edited by woodcutter on Mon Mar 10, 2008 7:26 am, edited 1 time in total.

User avatar
Lorikeet
Posts: 1374
Joined: Sun May 18, 2003 4:14 am
Location: San Francisco, California
Contact:

Post by Lorikeet » Mon Mar 10, 2008 7:25 am

Well I will certainly admit to teaching "voicing" by having students put their hands over their throats, but I do it with the sounds /s/ and /z/, /f/ and /v/. A lot of students are able to make the voiceless but not the voiced of one of those pairs, and can carry it over to the second pair.

I tried your whisper thing, and I'd suppose that a /t/ or /d/ in initial position (as well as other stops) could be differentiated by the aspiration more than the voicing. (You will notice there is more "confusion" between a /t/ with no aspiration and a /d/ than an aspirarted /t/m and I'm not so sure if you whisper them if they aren't essentially the same.)

As for voiced stops at the end of words, the most important part is the length of the the previous vowel; not the "voicing." So if you whisper "cat" and "cad" you will hear the second vowel to be much longer, thereby clueing you in on the voiced status.

The same goes for the voiced continuants (or whatever the word is that is used these days) at the end of a word. In fact, even when not whispering, the final /z/ sound really trails into a /s/ sound unless there is a vowel after it; however, the length of the preceding vowel lets you know it was voiced.

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

Post by woodcutter » Mon Mar 10, 2008 7:39 am

I wonder which learners of English have trouble with voicing, if we look at it in the way I have argued?

I zaw a cat. I velt bad.

Does anyone make these kinds of errors?

By the way Lori, you say "more than the voicing". In what way can a stop have voicing? All that means is that voicing begins as soon as you release. And the only alternative to that happening, unless you are whispering, is a breathy explosion, aspiration. Non-voicing/aspiration seem to me to be precisely the same thing. If you whisper you can hear a contrast between the breathy explosion of the stop and the normal unvoiced airflow of a vowel after an aspirated consonant, and that works pretty well too.

blackmagicABC
Posts: 37
Joined: Mon Feb 18, 2008 4:37 am
Location: Taiwan

Post by blackmagicABC » Mon Mar 10, 2008 3:47 pm

I wasn't disagreeing with you. I agree. I just find it the easiest to explain to a grade one or two kid who asks me why one sound is long and the other short. That was really the reason for the experiment. It really does happen kind of naturally and I suspect (have no proof) it has to do with the aspiration and that it happens more naturally if you don't think about it.
I used to be a pro magician but now ... I am not. Point though is, ventriloquism is an interesting study as well. You can get some very accurately bad (can not think of a better way to put this) sounds by playing with aspiration and voicing.

thethinker
Posts: 49
Joined: Thu Sep 08, 2005 5:49 pm

Post by thethinker » Mon Mar 10, 2008 4:06 pm

In fact it's not even aspiration that allows us to distinguish between what are termed voiced and unvoiced stop sounds in English, it's the delay between the release stage and the voicing stage.

There are three stages with a stop sound in English - there's the closure stage, when the air builds up, then there's the release stage where we release the air, and then there's the voicing stage, where we actually use our vocal cords. In a sense it's true that a stop sound can't be voiced as such a sound by nature involves stopping the airflow (and hence any voicing), but in many languages there is a voicing stage before the closure stage, and if there is vocalisation immediately before and after the stop/release stage then I think the sound is categorised as being voiced.

But in English (or at least in RP) there's no voicing before the closure stage in either voiced or unvoiced stop sounds, and what distinguishes the two is that 'voiced' stop sounds have vocalisation simultaneously with the release, whereas 'unvoiced' stop sounds have a delay (caused by aspiration) between the release and the vocalisation. But it seems that it's the delay we perceive, not the aspiration, as there have been experiments in which sounds have been manipulated to actually remove the aspiration but preserve the delay, and native speakers still perceive these as being unvoiced (and aspirated).

And of course we're only talking about stop sounds found at the beginning of words, as those at the end have no voicing, and those in the middle may or may not depending on the stress pattern of the word.

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

Post by woodcutter » Tue Mar 11, 2008 12:29 am

I agree, but it seems a bit superflous to say that, since you can't really produce a delay without the aspiration (without technology).

As to "onset of voicing", there is a fair bit of talk about that in phonetics, but I'm not sure if I believe it matters as much as some people make out. In English you might make a bit of a voiced buzz before the word "gummed" or you might not, just as you might make a slight (voiced) vowel sound after the d or you might not. It still sounds like "gummed".

By the way, concerning contrasting Z and S and such like while touching the throat, this artificial exercise removes the element ot time. Isn't time the most crucial element here?

Blackmagic, I suppose one could consider whispering in the same vein as ventriloquism (and I'm sure that ventriloquism has lots to teach us about phonetics), but it seems to me that while a ventriloquist cunningly overcomes problems and sounds a bit odd, a whisperer encounters no problems at all.

thethinker
Posts: 49
Joined: Thu Sep 08, 2005 5:49 pm

Post by thethinker » Tue Mar 11, 2008 5:30 pm

Yes, I wasn't really suggesting that the fact that the delay rather than the aspiration has any teaching implications - it's just that it's quite interesting. But I don't completely agree with what you say about inserting a 'buzzing' before 'gummed' - there have also been experiments that showed the native speakers perceive a 'fully' voiced stop sound like /g/ (by 'fully' voiced I mean with vocalisation before the closure stage) as being recognisable as a /g/, but at the same time sounding strange, so it does make a difference (albeit a small one since it doesn't impede communication).

User avatar
Lorikeet
Posts: 1374
Joined: Sun May 18, 2003 4:14 am
Location: San Francisco, California
Contact:

Post by Lorikeet » Tue Mar 11, 2008 6:56 pm

woodcutter wrote:I wonder which learners of English have trouble with voicing, if we look at it in the way I have argued?

I zaw a cat. I velt bad.

Does anyone make these kinds of errors?
Actually, the errors are more "I want to sip the sipper." I have never heard someone correctly make the voiced and not make the voiceless; only the other way around. (But of course I am sure there are some languages some where that do that--just not the ones my students speak. ;) )
By the way Lori, you say "more than the voicing". In what way can a stop have voicing? All that means is that voicing begins as soon as you release. And the only alternative to that happening, unless you are whispering, is a breathy explosion, aspiration. Non-voicing/aspiration seem to me to be precisely the same thing. If you whisper you can hear a contrast between the breathy explosion of the stop and the normal unvoiced airflow of a vowel after an aspirated consonant, and that works pretty well too.
I was trying to explain that the lengthening of the vowel prior to the voiced stop was more an indication of the voiced consonant than the so-called voicing itself, which I don't think is as significant. If you whisper "cat" and then whisper "cad", I think you will find the length of the vowel is longer in "cad" than in "cat". In initial position, I said the aspiration is the key to understanding whether it is "tab" or "dab", not the voicing. Do you disagree?

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

Post by woodcutter » Wed Mar 12, 2008 12:24 am

No, we agree Lori, since you throw the "so-called" in!

Which first language speakers say "sip the sipper"?

I did notice that voicing either end of gummed makes me sound a bit like a speaker from "Zummerzet" in the South-West of England. I wonder about the implications for differentiating words though - I don't know that it necessarily follows that there will be any.

Surely a language which made serious use of such differentiation would be one in which people would say "wow, you are a good whisperer!", as we might tell Blackmagic that he is a talented ventriloquist.

lolwhites
Posts: 1321
Joined: Wed Jul 16, 2003 1:12 pm
Location: France
Contact:

Post by lolwhites » Wed Mar 12, 2008 8:48 pm

Aspiration can't be the distinguishing feature either. The /k/ in skill, the /p/ in speech and the /t/ in stick aren't aspirated - if you don't believe me, say the words with a sheet of paper in front of your mouth and see how much it moves.

In fact, if you record a native speaker saying the words, and then digitally remove the /s/, you'll hear something that sounds much more like peach, gill and word I won't get past the swear filter!

woodcutter
Posts: 1303
Joined: Sat Jun 19, 2004 6:14 am
Location: London

Post by woodcutter » Thu Mar 13, 2008 12:26 am

If spelling were rational we would spell those words sgill and sdick. We should certainly do that when writing the phonetic spelling, but people usually don't.

I don't think it has any bearing on my argument, therefore.

thethinker
Posts: 49
Joined: Thu Sep 08, 2005 5:49 pm

Post by thethinker » Thu Mar 13, 2008 5:44 pm

No, the words 'skill', 'speech' and 'stick' are spelt perfectly logically (or at least the first two letters of each are). Try actually saying 'sbeech' or 'sgill' - it's not easy. So while stop sounds after /s/ might take on the characteristics of their voiced equivalents, from a phonological point of view (i.e. how our brains actually process the information) they are the /k/, /p/ and /t/ phonemes, but the proceeding /s/ means that they are not aspirated. A phonetic transcription of 'speech', 'skill', and 'stick' would begin /sp.../, /sk.../ or /st.../. However, a phonetic transcription of 'peach', 'kill' and 'tick' would mark the stop sounds /k/, /p/ and /t/ with a small 'h' to show aspiration. In a phonemic transcription there would be no difference in the transcription of the stop sounds since they are the same phoneme in each word pair - the details we're looking at here are alophones (ie. differences in a how an individual phoneme is realised), not different phonemes.

As for the argument that the fact that there's no aspiration in /t/, /p/ or /k/ after /s/ meaning that aspiration can't be the distinguishing feature between the voiced and voiceless pairs, that's also inaccurate. The fact is that the phonological system in English doesn't allow /sg/, /sb/ or /sd/ at the beginning of a word, so there's nothing to distinguish. Our brains know that /st/ at the beginning of a word is /st/ whether or not the /t/ is realised with aspiration, as /sd/ combinations simply don't occur. And in any case, our brains understand that the reason that the /t/ in /st/ isn't aspirated is not because it's actually a /d/, but because the /s/ has had an effect on the /t/ and caused it to be realised without aspiration.

Post Reply