Update: I apologise effusively for writing ɤ instead of ɣ, which I had intended all along.

Preface: I represent the voiceless velar fricative with its IPA phone [x],
and the voiced velar fricative with [ɣ].

  1. Is this phonetic inventory for Mandarin wrong? It excludes [x] and [ɣ], and includes only the voiceless uvular fricative [χ].

2. The Wikipedia page for [x] includes Mandarin, which it exemplifies with (I omit the links):

Chinese  |  Mandarin  |  河/hé  |  [xɤ˧˥]

3. but the Wikipedia page for [ɣ] excludes Chinese,

  1. and similarly Wikipedia's IPA for Mandarin page excludes [ɣ].

Will someone please clarify and explain these contradictions?

  • Though /ɣ/ (the voiced velar fricative) is not really a phonemic initial in standard Mandarin, it is attested as one possible realisation of the null initial, especially after final nasals, e.g. 棉襖 mián'áo. Yuen Ren Chao even wrote a paper on this.
    – Michaelyus
    Commented Oct 30, 2015 at 14:21

3 Answers 3


Voiceless: Pinyin h is standardly the voiceless velar fricative [x], although it is often written [χ] for some reason — Chinese IPA developed its own transcriptional traditions, for instance the use of [ɒ] where [ɑ] might be more usual, in the mid-twentieth century. However, there's no systematic contrast between [x] and [χ] in standard Mandarin, and systematic contrast is the important thing. Besides that, for many speakers, the aspiration of stops is also expressed with the velar fricative rather than pure aspiration: tāng: [tˣɐŋ], kuí: [kˣu̯ɘi], etc.

Voiced: In standard Cantonese, no. And in the major representatives of Yuè ("Cantonese") dialects, as described in Zhān Bóhuì 詹伯慧's Guǎngdōng Yuè fāngyán gàiyào 广东粤方言概要, voiced velar fricative [ɣ] is not attested.

There are some cases of [ɣ] in regional languages. Conservative Wú dialects, for instance, voice /x/ to [x͡ɣɦ] in low-register tones. Hokkien final /-k/ may take on voicing and approximant quality before the noun-suffix á: /sái ha̍k-á/ 'outhouse': [sai₅₅ ha₃₅ ɣa₅₂], and so on, although it is also often heard as a fully voiced stop [sai₅₅ ha₃₅ ga₅₂]. [ɣ] is reconstructed for the medieval initial xiámǔ 匣母 in most environments by most hands, too. But no dialect groups have anything like the degree of standardization we encounter in Mandarin.

  • 1
    Dr. Branner, welcome to Chinese StackExchange. Here, have an upvote! Commented Oct 30, 2015 at 5:35
  • Where did you see the voiced /x/ in Wu transcribed as the complicated [x͡ɣɦ] rather than simply [ɦ], and by the way, I thought /x/ in Wu is really [h].
    – Fan Zheng
    Commented Oct 30, 2015 at 23:57

There are two issues here:

[x] or [χ]?: Who cares? You could pronounce pinyin "h" in a variety of ways without impeding understanding (e.g., [h], [x], or [χ]). I think [χ] is a better description of my experience with pinyin "h", but this isn't a big deal.

[ɤ] is the voiced velar fricative, right? Nope! You're confusing two very graphically similar symbols. [ɣ] is the voiced velar fricative, and [ɤ] is a close-mid back unrounded vowel. Chinese lacks the former and has the latter (it's the "e" in "hē", "kè", etc.).

  • 1
    Really? I think pinyin "h" is [x] at most.
    – user58955
    Commented Oct 29, 2015 at 22:03
  • @user58955 Hence my description "could pronounce... without impeding understanding" as opposed to "the way native speakers pronounce it". Commented Oct 30, 2015 at 5:32
  • I was confused by your next sentence: `I think [χ] is a better description of my experience with pinyin "h"'. I thought you were referring to how it was pronounced by native speakers...
    – user58955
    Commented Oct 30, 2015 at 13:03
  • @user58955 In that sentence, yes. To paraphrase, "you'd be fine saying it any of these ways, but I would describe native speakers as producing uvular voiceless fricatives." Although, TBH, it's probably velar or uvular depending on the vowel. Commented Oct 30, 2015 at 16:13

Standard Chinese does contain voiceless velar fricative [x] but no voiced velar fricative [ɣ]. Actually Standard Chinese has only one voiced consonant [ʐ].

However, many Mandarin speakers especially those from southern China would pronounce [x] as [h], and speakers of northern China would tend to practice [x]. If you take [h] for [x], people will not misunderstand you.

[ɤ] is not a consonant; it's a vowel. It is different from [ɣ]. They are different symbols and represent different entities!

Responses to your questions:

  1. yes. errors occurred.

  2. correct.

  3. Please be careful !!! This page is about [ɣ] but not about [ɤ] (which your post originally confounded with [ɣ]). They are different symbols!!!

  4. no. you missed something. that page does include [ɤ], in the vowel table.

    ɤ | e | roughly like American sir

  • "Only one voiced consonant": but the liquids are all voiced, aren't they? Commented Oct 29, 2015 at 22:50
  • And the nasals. Commented Oct 29, 2015 at 23:05
  • More correctly one should say, "(according to this partiular analysis) a single voiced obstruent, [ʐ]". Then again, this analysis not universally accepted; my guess it's more of a rhotic consonant really.
    – flow
    Commented Oct 30, 2015 at 12:09
  • For many even from northern China, the actual realisation of pinyin "h" depends on the subsequent vowel. The "h" in "ha" is likely to be [h] but in "he" (where the subsequent vowel is [ɤ]) is almost automatically [x]. For those from southern China, if they get the vowel right (such as [ɤ]), the "h" would be automatically [x] as well. But it is always [x] in the standard accent used by state television newscasters.
    – user58955
    Commented Oct 30, 2015 at 13:06

Your Answer

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