What's the difference between voiced and "unaspirated" consonants and how can I check if I got it correct?

Question

I heard from somewhere that there are no voiced consonants in Mandarin. Wikipedia says that consonants like the pinyin "d", "b", "g" are "unaspirated", whatever that means...

Although my first language is Mandarin, I think I spoke too much in English and might have forgotten how do the "unaspirated" consonants sound. For example, I am not sure whether I am pronouncing the "unaspirated" pinyin "b" or the voiced consonant "b" in English when I say something like "不".

Wikipedia says that the pinyin "b" is the same as the p in the word "spy". But this didn't really help me understand "unaspirated" consonants. I have always been told that, in English, when "s" is followed by a "p", the "p" usually turns into a "b" sound. I always thought that the "p" in "spy" is the same as in the "b" in "but".

In addition, I can't seem to distinguish between voiced and "unaspirated" consonants by hearing. So I need another way to check if I pronounced it correctly.

What are the difference pronouncing a pinyin "b" and an English "b"? How can I check if I pronounced it correctly?

Answer 1

Voice-Onset Time

As you probably already know, the distinction between voiced vs unvoiced-unaspirated vs unvoiced-aspirated is the relative timing of articulation and voicing (called voice-onset time or VOT):

• Voiced stops have a negative VOT (i.e., the voicing begins before articulation)
• Unvoiced-unaspirated (sometimes called "lenis") stops have approximately zero VOT (i.e., voicing and articulation co-occur).
• Unvoiced-aspirated stops have positive VOT (i.e., voicing begins after a delay relative to articulation).

In IPA, we use discrete symbols to indicate these distinctions (e.g., b/p/pʰ). Lumping VOT measurements into "negative", "zero", and "positive" is just a convenient simplification, though. In fact, there is a range of possible values that it can take on, and languages do, in fact, differ from one another in terms of how they realize "voicing" or "aspiration". However, any given language only distinguishes at most 3 stops on the basis of VOT alone**, so IPA (e.g., b/p/pʰ) is sufficient to describe things.

Measuring VOT

This is all great, but you say that you're learning Japanese (which has very negative VOT on its voiced stops), and as a speaker of Mandarin (zero VOT) and English (slightly negative VOT), you're not sure whether you're producing voiced or lenis stops. You can in fact measure VOT on an audio recording of yourself. Here's a tutorial on using praat to measure VOT on audio recordings.

If that sounds too complicated, you put your hand on your throat and feel when you produce voicing. To practice making voiced consonants having very negative VOT, try to keep your lips closed while starting vibrate your vocal chords. It's hard to do this very long, as your mouth starts to fill up with air, but it can help you see how it feels to begin voicing earlier than releasing a stop.

* It's probably even more complicated than that, but I'm going to assume that VoT is good enough to explain everything.

** Hindi "voiced aspirated stops" are breathy voiced, to respond to the obvious rebuttal.

Answer 2

It's a fairly technical problem to describe these sounds; for some of the details see the wikipedia article on Fortis and lenis. As a native English speaker, I find the aspirated unvoiced consonants (au) very easy to distinguish, but I may have trouble distinguishing between voiced unaspirated (vu) and unvoiced unaspirated (uu). If you listen to a Hindi movie, you can hear the fourth possibility, voiced and aspirated (va) ; a very unusual sound to people who are not native speakers of such a language.

How can one distinguish vu and uu? Two suggestions; Put your finger on your adams apple; vu has voicing almost immediately after release, while uu has a delay. the apple begins vibrating later after the -p- in spy than it does after the b in buy. In English, the b- lasts longer than -p- There is also more pressure or energy with the -p- than with the b- (very subjective, I know).

Note that in Chinese, when uu is intervocalic, it becomes voiced. For example, in the word baba 'father', the initial b is uu, but the second b is vu. Once my attention was drawn to this, I found that the two did indeed sound and feel different.