"Clear" and "turbid" in Chinese phonology, part 3

« previous post | next post »

[Guest post by San Duanmu.  Please note that San's remarks were written before Sara de Rose's post ("part 2") on the same subject earlier this evening.]

In response to Victor’s request, I am offering some comments on qing 清 (clear) and zhuo 濁 (muddy), two commonly used terms in traditional Chinese phonology. I shall follow the outlines suggested by Victor as well.

  1. When and how did the terms arise?

According to Tang (2016: 32), the terms were used linguistically in a ten-volume book 《聲類》 (Sound Categories) by 李登 (LI Deng) during 三國時期 (Three Kingdoms period, 220-280). The book was later lost, but references to it can be found in other books that survived.

According to YU Min 俞敏, in Li Ji《禮記》 (the Book of Rites), compiled by followers of Confucius (孔子 551-479 BC), the terms were also used to discuss music, as in “长者浊也……短者清也” (long ones give a muddy sound… short ones give a clear sound). If long and short refer to the shape of an instrument, then ‘muddy’ ought to mean a lower tone and ‘clear’ a higher tone. The exact relation between the terms used in music and those in sound classification is open to interpretation.

  1. How do the terms function within traditional Chinese phonology?

In traditional Chinese phonology, qing 清 (clear) and zhuo 濁 (muddy) are used to classify consonants. In addition, each is further divided into two sub-categories. Therefore, there are four categories of consonants, shown in the table below, with samples in IPA (International Phonetic Alphabet).



Samples in IPA

qing 清 (clear)

Voiceless unaspirated stop/affricate


ci-qing 次清 (less-clear)

Voiceless aspirated stop/affricate


zhuo 濁 (muddy)

Voiced stop/affricate


ci-zhuo 次濁 (less-muddy)

Nasal and approximants

[m l r w j]

Fricatives do not contrast in aspiration, so there are only two categories of them. It is not clear which one they ought to be grouped with. Tang (2016) groups voiceless fricatives with qing 清 (clear) and voiced fricatives with zhuo 濁 (muddy). However, [s] is clearly closer to the fricative in [tsʰ] than that in [ts].

  1. How do the terms correlate with concepts in modern linguistics?

There is a common view that, in modern terms, qing means ‘voiceless’ and zhuo means ‘voiced'. It is true that the consonants in the qing and ci-qing categories are all voiceless, and those in the zhuo and ci-zhuo categories are all voiced. However, it is hard to understand why [pʰ] is ‘less voiceless’ than [p], or why [m] is ‘less voiced’ than [b]. If qing means ‘less sound’, then we can explain why [p] has more qing that [pʰ] (because [pʰ] has more noise from the aspiration). However, if we interpret zhuo as having more sound, then there is a problem: cannot explain why nasals and approximants have ‘less sound’ than stops, because clearly, nasals and approximants are much louder that stops.

The conclusion, then, seems to be that classic Chinese linguists did not have a clear notion of voicing, and that even though qing 清 (clear) and zhuo 濁 (muddy) have something to do with some properties of the target sounds, they are mainly categorical labels, rather than precise phonetic descriptions.


唐作藩 (TANG, Zuofan). 2016. 《音韻學教程》(第五版). 北京: 北京大學出版社.


Selected readings


  1. ~flow said,

    December 5, 2020 @ 5:05 am

    Are there any cross-cultural studies how people in various languages describe phonetic/phonological distinctions like voiced/unvoiced and so on? We have gained great insights from such studies when it comes to the how and why of color words and can now even make some sense of the color terms used in ancient texts like the Illiad, so there's hope that a historical, cross-linguistic overview may shed some light onto 清濁.

  2. Chris Button said,

    December 5, 2020 @ 7:25 am

    Pulleyblank treats "zhuo" in Late Middle Chinese as having voiced breathiness /ɦ/. So, for example, EMC /b/ would be /pɦ/. He compares this to the case of Wu Chinese, which Ladefoged & Maddieson (1996:64-66) in their "Sounds of the World's Languages" describe as "determined not so much by whether the stops are voiced or voiceless during closure, but rather by a phonation difference at the release of the stops. The so-called voiced stops have slack voice offsets, a quality detectable in the following vowel …"

  3. ~flow said,

    December 5, 2020 @ 10:16 am

    @Chris Button—do you know of a website that present contrastive examples of breathy voice? I've looked around but couldn't find any.

    For those not familiar with Chinese traditional phonological terminology—there's always a chance that whoever chose a given (set of) term(s) for a particular phenomenon did so in an self-referential way, meaning that the terms themselves are examples of the phenomena such labelled. At least that's what I think was implied here. Note though that e.g. the terms for the 四呼 'four enunciations(?)' only fit that hypothesis to 50%, as the readings of 開 and 齊 are examples of what they describe, but 合 and 撮 are not (at least not in modern Mandarin), so that's far for being a certainty as far as 清 and 濁 are concerned. Would be intriguing if it could be shown to be true, though!

  4. David Marjanović said,

    December 5, 2020 @ 12:34 pm

    I wouldn't be surprised if stiff voice & slack voice were indeed the key to this mystery.

  5. San Duanmu said,

    December 5, 2020 @ 1:34 pm

    I'd like to add a few words to my note:
    As categorical labels, qing 清 (clear) and zhuo 濁 (muddy) have served the purpose quite well. On the other hand, finding accurate terms for phonetic or phonological properties remains a challenge even in the present day, and terms like ‘voiceless’ and ‘voiced’ are not without problems. For example, should voicing be defined in terms of articulatory gestures (such as ‘slack vocal cords’, as proposed by Halle & Stevens 1971) or their acoustic effects (such as the periodic vibration of the vocal cords, as assumed in popular textbooks)? Should the English [b] and the French [b], which are phonetically different, both be called ‘voiced’? Should English nasals, which do not contrast in voice, be specified as ‘voiced’? Answers are unlikely to come from statement of classic scholars, but require new research.

  6. Michael Watts said,

    December 5, 2020 @ 2:26 pm

    Should the English [b] and the French [b], which are phonetically different, both be called ‘voiced’? Should English nasals, which do not contrast in voice, be specified as ‘voiced’?

    I don't understand these questions. Should the English [b] and [d], which are phonetically different, both be called "voiced"?

  7. ~flow said,

    December 6, 2020 @ 2:34 am

    @Michael Watts "Should the English [b] and [d], which are phonetically different, both be called "voiced"?"

    English [b] and [d] are different in place of articulation, not in manner of articulation, which is what voice/voiceless is all about. Conversely, the manner of articulation between English [b] and French [b], and English [p] and French [p] does differ—or so goes the argument—so that's the relevant point here, that when you arrange the two English plosive bilabial stops and the two French plosive bilabial stops on a line from 'ideally voiced and unaspirated' to 'ideally unvoiced and aspirated' you'd get four points, not two. And even if we conflate 'voicing' and 'aspiration' to the one dimension of voice onset timing (VOT), there's still breathiness to take into account for languages that make that distinction (not to speak of ingressive sounds and so on), so it's a potentially multi-dimensional space.

  8. Michael Watts said,

    December 6, 2020 @ 7:48 am

    English [b] and [d] are different in place of articulation, not in manner of articulation, which is what voice/voiceless is all about.

    This doesn't explain why we might not want to call English nasals "voiced".

  9. Chris Button said,

    December 6, 2020 @ 9:16 am

    @ Michael Watts

    The point is about contrastive voicing relative to a voiceless, or more likely preaspirated (so partially devoiced), nasal.

    For me, this all goes back to an argument I've made a few times elsewhere on LLog that obstruent vs sonorant is a far more useful distinction than consonant vs vowel.

RSS feed for comments on this post