11

The various Wikipedia articles covering Standard Chinese all seem to agree that Mandarin does not have voiced plosives, fricatives, or affricates except for [ʐ] / [ɻ], written in Pinyin as "r".

But many other sources cite voiced IPA symbols and when I get my native Mandarin speaking Chinese friend who lives in the Beijing region to demonstrate the alveolar, retroflex, and alveolo-palatal series, it really seems to my ear that he is making voicing distinctions where Wikipedia indicates only aspiration distinctions.

Further, when I try to produce the sounds with only aspiration differences he corrects me until I give in and add voicing.

I must make it clear that I'm just an armchair linguist with no official training.

But what might I be observing? Could it be regional differences in Mandarin away from Standard Chinese? Could it be due to differing analyses of Chinese phonology? Could it be that Wikipedia is listing archiphonemes which commonly do have voiced realizations? Could it be that Wikipedia is just plain wrong? Or is it all down to my uneducated ear?

Improve this question
7
  • 1
    I'm on my 5th beginner Mandarin lesson and having the same problem. It seems to me as if the teacher is quite consistent with the plosives (although not entirely) but seemingly randomly voices or unvoices the fricatives and affricates. This is hard because I can't hear the aspiration there very well.
    – Dominik Lukes
    Nov 7, 2013 at 7:17
  • @DominikLukes: user2619's answer does seem pretty good though. Hopefully it will help us (-:
    – hippietrail
    Nov 7, 2013 at 7:20
  • There’s an info box on the Pinyin Wikipedia page with suggestions and diagrams for pronouncing Mandarin fricatives/affricates. It makes no mention of voicing, but pretty clearly stresses aspiration (ji/zhi/zi are “gently released”, but qi/chi/ci are “blown away forcefully”.) These are definitely more challenging than the stops.
    – neubau
    Nov 7, 2013 at 8:30
  • 1
    Anecdotally, when I was learning Mandarin and practicing it with my Thai phonologist friend (who had many more years' experience with Mandarin), he was quick to correct my production of initial /bdg/. "It sounds too voiced!" he kept saying. Only when I made an effort to make them voiceless-unaspirated was he satisfied. This suggested to me that he mapped the Mandarin /bdg/ onto his own Thai voiceless-unaspirated category.
    – musicallinguist
    Nov 7, 2013 at 17:46
  • 1
    Yes, in fact Lao has high falling and low falling tones. There are two ways to spell words with the falling tone in Thai – for example, /na:/51 can be written หน้า (meaning ‘face’) or น่า (meaning ‘-able’ in words like น่ารัก ‘lovable’. Historically, this reflects a merger of two tone categories which remained separate in Lao. The falling tone also pops up in odd places when rendering foreign words, for example in the last syllable of ‘central’ – no idea why. As for being 'sweet', well it's kind of looping, that might qualify. A more pleasant sound to my ear than the one in Mandarin.
    – neubau
    Nov 11, 2013 at 5:31

2 Answers 2

Reset to default
10

I don't think it's a regional question. Mandarin b,d,g may often be realized with voicing, but the key distinction is aspirated vs. unaspirated. Here is a lab study of the voicing profiles of Mandarin and German. This comparison with German stops sums it up:

“Phonologically, all Mandarin stop consonants are voiceless; /p,t,k/ are aspirated, /b,d,g/ are unaspirated. In German, stops are usually classified by voicing: /p,t,k/ are voiceless, /b,d,g/ are voiced; aspiration is an optional feature of the voiceless stops. The results of our study show that, despite this typological difference between the consonantal systems of the two languages, Mandarin /b,d,g/, phonologically voiceless and unaspirated, and German /b,d,g/, phonologically voiced and unaspirated, show very similar patterns in their voicing profiles.” [p.2]

Specifically for Mandarin, they found that “83% of realizations of the voiceless unaspirated alveolar stop /d/ in Mandarin are voiced at the beginning of the closure, and 50% are still voiced at the end of the closure.” While the unaspirated stops are more likely to be voiced than the aspirated ones, this is not really surprising. In practical terms, pronounce /b,d,g/ as you might in English, but make sure to aspirate /p,t,k/. It’s not difficult, really, since they are initial sounds, just like English aspirated stops. I don’t think this poses problems for native speakers of English, unlike the tripartite stop distinction found in many languages of mainland Southeast Asia. (Taiwanese also has a 3-way distinction, by the way.)

Here's the link: Chilin Shih, Bernd Moebius, ‘Contextual Effects on Voicing Profiles of German and Mandarin Consonants’ http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.4.3423&rep=rep1&type=pdf

Improve this answer
4
  • Is that the common description for German? I would say that for /p/, aspiration and non-aspiration are in complementary distribution (non-aspirated when preceded by sibilant), but for /b/, voicing is more or less optional.
    – dainichi
    Nov 8, 2013 at 6:30
  • @dainichi – thanks, I wasn’t aware of that. From Wikipedia, it seems that there is a regional difference – in southern varieties of German, /b/, /d/ etc are voiceless, so it’s better to talk about fortis-lenis pairs for German. Note that in the study, a single speaker provided the data for each language. Clearly there is a problem with generalizing in the German case, but that doesn’t necessarily invalidate the generalization that (at least some) Mandarin speakers voice /b/, /d/ etc at least half the time.
    – neubau
    Nov 11, 2013 at 3:05
  • As far as I remember, voicing of plosives in German is only relevant before a nasal release.
    – unknown_person_1000
    Dec 12, 2018 at 20:44
  • See also: Jessen, Michael. 1998. Phonetics and phonology of tense and lax obstruents in German.
    – unknown_person_1000
    Dec 12, 2018 at 20:45
3

There's no rule that says Mandarin unaspirated stops CAN'T be voiced. The point is that there are no minimal pairs between voiced and unvoiced unaspirated stops. The actual phonetic realization- in particular, the voice onset time- can and will vary by speaker, dialect, time, circumstance, etc.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.