Pinyin pronunciation: zh as z, sh as s (retroflex)


Senior Member
Based on what I've read so far, the Pinyin sequence "zh" is supposed to represent the voiceless unaspirated retroflex affricate, i.e. [ʈʂ], but in the word "zhōng" (as in "Zhōngguó"), my Chinese instructor pronounces it as the voiceless unaspirated alveolar affricate, i.e. [ts], represented by the letter "z". Thus, to me, it sounds as though he were saying "zōng". Now, I've already looked this issue up and in some other threads, I've read about how in certain dialects (e.g. in Taiwan), the retroflex consonants are reduced to their alveolar equivalents, but my Chinese instructor doesn't reduce all retroflex consonants in this way. For example, he pronounces "zhè" with a retroflex affricate. Then again, he also pronounces "shuǐ" (water) as though it was written "suǐ".

So, my question is, why does he substitute retroflex consonants with alveolar ones only in some words; what is the underlying pattern here? Moreover, will my pronunciation be correct if I always use consonants in places where they're written, including in "zhōng" and "shuǐ"?

Thank you in advanced
  • fyl

    Senior Member
    Mandarin Chinese
    Hi, first of all, shui (water) is the correct way to write and sui is non-standard (I guess that's a typo?). It is always correct to pronounce retroflex consonants as Pinyin shows. In standard Mandarin, zhong, shui are both retroflex.

    As for your teacher pronounces some retroflex consonants with non-retroflex ones, it can be a dialectical issue.
    There are 4 groups of relevant consonants in Middle Chinese, the 精 group, the 莊 group, the 章 group and the 知 group. For simplicity let's call them c1, c2, c3, c4 respectively. Below is how they are pronounced in several Mandarin dialects (of course, this is the general rule and there are exceptions).
    1. Many southern Chinese pronounce all of them with non-retroflex consonants when they speak Mandarin.
    2. In standard Mandarin, c1(精) is non-retroflex 'z', 'c', 's'; c2(莊), c3(章), c4(知) are retroflex.
    3. In some Mandarin dialects (e.g. my mother tongue), c1(精) and c2(莊) are non-retroflex, some c3(章), c4(知) are also non-retroflex, and other c3(章), c4(知) words are retroflex. I have found people from places very far away also speak with a similar pattern. (For example, zhong1 and shui3 are also non-retroflex in my dialect.)
    4. There are many other ways too. In some Mandarin dialect, all 4 groups are retroflex. In some other dialects, they use alveolo palatal 'jqx' (Pinyin) in places of retroflex consonants in case 3.
    So if your teacher was affected by some Mandarin dialects, he or she can pronounce some retroflex words with non-retroflex 'zcs'.

    Note: In many Mandarin dialects (including the standard Mandarin), some c1(精) words, including 精 itself, have developed into 'jqx' (when followed by an /i/ or /y/ sound), so you won't found them having 'zcs' consonants.
    Last edited:


    Senior Member
    zh and z, ch and c, sh and s, all the three pairs may be pronounced incorrectly, mistaking one with the other.
    So shui becomes sui. ....


    Senior Member
    why does he substitute retroflex consonants with alveolar ones only in some words; what is the underlying pattern here?
    What you are talking about concerns language transfer (i.e., L1 interference), that is, applying linguistic features from the native language to a second language. Take the Min speakers as an example: Since <zh> (i.e., the retroflex affricate [ʈʂ]) is absent in their phonological inventory, many speakers of that language would naturally substitute it with a sound that they are familiar with. /ts/ is close to /ʈʂ/, and so it becomes the typical substitute.

    水 "water": Min /sui/ or /tsui/ vs. Mandarin <shuǐ> ==> Min /s/ (fricative) is closer to <sh> (i.e., fricative /ʂ/) than /ts/ (affricative) is. Hence, <sh> is typically substituted with /s/.

    In the Min language, /tsi/ is palatalized:
    這 "this": Min /tsit/ (palatalized as [tɕit] and voiced dialectally as [dʑit]) vs. Mandarin <zhè>.
    遮 "intercept": Min /tsia/ (palatalized as [tɕia] and voiced dialectally as [dʑia]) vs. Mandarin <zhē>.
    Also 折 (Mandarin zhé vs. Min /tsiat/), and so forth.
    Mandain schwa triggers palatalization and, dialectally, voicing effects in language transfer, it thus gives you the impression that your teacher "pronounces zhè with a retroflex affricate". Mandarin <o> (/ɔ/) in 中, a back vowel, fails to trigger palatalization, and so the unpalatalized /ts/ sounds "as though he were saying zōng".
    Last edited: