Does full devoicing occur?

[5jj]

I am trying to decide how I know. I'll come back when I have a helpful answer - if I find one.

Sorry, I haven't found a convincing answer. I just 'know' that I can hear the vibrations of the voiced sounds.

It's not always easy or obvious - I remember that when I did a refresher phonetics course at UCL a couple of years ago, some of us were not particularly good at this at the start, but the only real 'training' was practice.

I did listen to your samples more than once before I committed my view to paper - to screen, I should [STRIKE]say[/STRIKE] write..

 

[BobK]

:up: I think you either can or you can't, and 'can' refers to both ability and knowledge (as is pointed to by the word's close relation 'canny' [=knowing/crafty...]). The knowledge in this case is a knowledge of English phonology (in particular, English phonemes). We (native speakers) have been programmed since birth to hear the voicing, and the only sure-fire way of telling if it's really there or not is to look at a sound-spectrogram.

b

 

[5jj]

You're right, Bob. Despite some training, I would not be so confident in my views, initially at least, on the sounds in a completely unknown language.

One example of this is the word-initial /p/ phoneme in Czech. Like the English /p/, it is not voiced. However, unlike the English /p/ it is lenis and unaspirated. Until I tuned in, I heard it as /b/.

 

[ancor90]

Thank you for not forgetting me

I found the problem is not only perceiving it correctly, the problem is also drawing the border between two sounds at the correct place, if that makes sense.

If you have two similar sounds, the first (1) at the left, the second (2) at the right, there's a linear continuum between them.

1 ........ | ....................... 2

Where you draw the border (|) might be different in two languages.

In my native language it's usually okay to substitute /z/ with /s/. In fact, the German letter z is almost always pronounced /ts/, so "Zoo" is pronounced /tso/ in German, while the German letter s is sometimes /s/, sometimes /z/, depending on the dialect. This is why, to my ears, both the English /s/ and /z/ sounded like /s/ to me (with the exception of a really strong /z/). I, my girlfriend, and my parents pronounce most German words that have the letter s in them with an /s/, while my German dictionary says for the same words that it's a /z/, and if I pronounce it with an /z/ it still sounds German. It usually doesn't matter, my ears became more or less indifferent to the difference.

But when I finally payed attention and notice how an English native speaker pronounces the /s/ phone, it struck me that it is indeed different from even an only slightly voiced English /z/ phone.

So, I can now hear the /z/ in the first audio example of the woman I attached in an earlier post. Which makes me happy. :-D

I perceived it all the time, I just "drew" the border between /s/ and /z/ at a different place than you guys do. I just need more practice to hear the difference effortlessly.

The /ʧ/ vs /ʤ/ at the end of a word still causes problems for me, however. I can hear the difference at the beginning of a word, so I can distinguish chew vs Jew, but at the end of a word where /ʤ/ may be devoiced in normal speech, I find it difficult to hear.

To illustrate my problem I again have attached two audio samples. One has an /ʧ/, the other a devoiced /ʤ/. Please try to listen to each sample only once before you make up your mind on which is which. I'd be interested if you hear a difference at the first attempt.

I can make out which is which, but I doubt I could do it in normal conversations.

 

[5jj]

On one listening only, I could hear no difference.

 

[ancor90]

On one listening only, I could hear no difference.

See, that's my problem. I had some native speakers identify which is which, while some others couldn't. I'm somewhat glad I'm not the only one who's having problems with this.

I had the same problem with the 2nd and 3rd sample from my previous post (Fringe and charge). People tend to devoice voiced consonants if they are at the end of a word. This is often not a problem, however, because vowel lengthening, context, and other cues will often make it unambiguous which is meant.

When it's at the beginning of a word, however, it is really, really obvious to me.

Out of curiosity, can you identify which is which after listening to it more than once?

 

[5jj]

Out of curiosity, can you identify which is which after listening to it more than once?

Having stated so confidently in post #21, "I just 'know' that I can hear the vibrations of the voiced sounds", I am now going to claim that I have, suddenly and inexplicably, been struck deaf. This means that I cannot answer such potentially embarrassing questions.

 

[birdeen's call]

I didn't hear the difference either.

By the way:

One has an /ʧ/, the other a devoiced /ʤ/

What's the difference between /ʧ/ and devoiced /ʤ/? Do you mean that devoiced is more voiced than voiceless?

 

[5jj]

What's the difference between /ʧ/ and devoiced /ʤ/?

If /ʤ/ is fully devoiced, then there is no difference, assuming that there are no fortis/lenis or aspiration factors involved. A Martian phonetician, hearing the underlined sound in "I spoke to Judge Thomas yesterday, might record it as /ʧ/; a native speaker of English might hear it as a devoiced /ʤ/, but would, I believe, record it as /ʧ/.

 

[thatone]

I can't hear the difference either, but I tried to make it easier to see.

Here's a spectrogram of the culprit sound. The first two graphs show the audio as-is, the second two show the audio amplified by 10.6 dB.

Also have attached the amplified audio files. 0a and 1a are the full audio files with amplified /ʧ/ and /ʤ/, 0aa and 1aa contain only the latter, isolated.

To me, 1aa sounds more like a /ʤ/, what do you think?
I'm really surprised some people can distinguish the two sounds in normal speech though.

 

[5jj]

I have to ask - who decided that one was /ʧ/ and the other devoiced /ʤ/?

 

[birdeen's call]

Also have attached the amplified audio files. 0a and 1a are the full audio files with amplified /ʧ/ and /ʤ/, 0aa and 1aa contain only the latter, isolated.

To me, 1aa sounds more like a /ʤ/, what do you think?
I'm really surprised some people can distinguish the two sounds in normal speech though.

I hear some difference now, but I wouldn't call any of these sounds /ʤ/. They're both /ʧ/'s to me.

 

[birdeen's call]

If /ʤ/ is fully devoiced, then there is no difference, assuming that there are no fortis/lenis or aspiration factors involved. A Martian phonetician, hearing the underlined sound in "I spoke to Judge Thomas yesterday, might record it as /ʧ/; a native speaker of English might hear it as a devoiced /ʤ/, but would, I believe, record it as /ʧ/.

So the only difference is in perception, not in real sounds? I'm asking these questions, because I'm not sure what the OP is saying. Is it that there is a real objectively perceptible difference between the sounds and the question is whether we can hear it? Or is it that there's no perceptible difference and the question is whether some people will tend to hear one anyway?

 

[ancor90]

What's the difference between /ʧ/ and devoiced /ʤ/? Do you mean that devoiced is more voiced than voiceless?

No. As I learned, both /ʤ/ and /ʧ/ differ in other aspects besides voicing, which include but are not limited to [1]:

- higher/lower amplitude of burst or frication
- length of the constriction duration
- length of the preceding vowel.

A devoiced /ʤ/ is shorter and less loud than a /ʧ/.

If you really pay attention, you can hear that in the samples I provided in an earlier post.

The first one is a devoiced /ʤ/, the second a /ʧ/.

So the only difference is in perception, not in real sounds?

No, there is a real difference and some native speakers I asked can hear it right away, others can't.

I'm asking these questions, because I'm not sure what the OP is saying.

I'm saying, or rather asking, if such full devoicing occurs in normal speech; and if native speakers are maybe unaware of that and hear voicing although it's not there. Because sometimes I just can't hear voicing where one's supposed to be.

[1] http://www.ualberta.ca/~tnearey/Ling205/Week3/WBPhonemesAndAllophonesEngCs4Up.pdf -- page 8

 

[5jj]

I'm saying, or rather asking, if such full devoicing occurs in normal speech; and if native speakers are maybe unaware of that and hear voicing although it's not there. Because sometimes I just can't hear voicing where one's supposed to be.

This is why phoneticians use spectograms. Very few people have the ability to hear sounds apparently possessed by such early phoneticians as Henry Sweet and Daniel Jones. If spectograms had been around in the days when they were working, the machines might have revealed that these phoneticians were not as perceptive as everybody thought.