Why are words like 天、面、先、盐 romanized in pinyin as tian, mian, xian, and yan, even though the ending is pronounced [iɛn], and would be better represented as -ien? Likewise, why are words like 卷、全、选、元 romanized as juan, quan, xuan, and yuan, when the final is pronounced [yɛn] and would be better represented as -uen? (at least the way northern Chinese people pronounce it – it seems that southern Chinese people pronounce it as [yan])

Is it that those words actually were pronounced with a [ian] and [yan] final in the past? That doesn't seem like a complete explanation, as Wade-Giles uses -ien, and Wade-Giles is from 1892 (although it does use -üan and not -üen).

  • Simply different design, many writing system/language has more than one romanization/latin-based writing solution Commented Jul 28, 2019 at 3:10
  • 2
    This is not an answer, but please note that pinyin -en (本、恩、etc) corresponds to /-ən/ and not /-ɛn/. Basically, whatever combinations of letters you try to use for that rime, you will run into inconsistencies with some other established rime.
    – dROOOze
    Commented Jul 28, 2019 at 12:35
  • 1
    @droooze Pinyin already has a rule that -iê /iɛ/ is written -ie, so it wouldn't add another inconsistency to have -iên /iɛn/ → -ien.
    – mic
    Commented Jul 29, 2019 at 13:54
  • 1
    @Pedroski aɪ noʊ, ɪf ˈoʊnli wi kʊd ɹid ænd ɹaɪt ˈɛvɹiθɪŋ ɪn IPA
    – mic
    Commented Jul 29, 2019 at 14:04
  • 1
    Instead of having two inconsistencies (1. -ian is pronounced [iɛn] instead of [ian]. 2. -ie is pronounced [iɛ] instead of [iə]), we could have just one inconsistency (1. -e following -i is pronounced [ɛ] instead of [ə].)
    – mic
    Commented Sep 24, 2020 at 5:47

5 Answers 5


Although Wade–Giles uses -ien, Zhuyin (1910) uses ㄧㄢ -ian, Gwoyeu Romatzyh (1926) has -ian for the basic form of first tone, Latinxua Sin Wenz (1931) has -ian, and Yale romanization (1943) uses -yan. Likewise, pinyin's usage of -üan instead of -üen for [yɛn] has precedence in Wade–Giles (1892) -üan, Zhuyin ㄩㄢ -üan, Latinxua Sin Wenz yan, and Yale ywan.

Yale romanization in particular was supposed to be designed to be intuitive for English speakers to pronounce, which suggests that its yan and ywan were not pronounced [iɛn] and [yɛn] at the time, but rather [ian] and [yan].

Alternatively, these could be a holdover from Zhuyin. In Zhuyin, pinyin -ian is represented as ㄧㄢ -ian, and couldn't be represented as ㄧㄣ -ien, because that's used for pinyin -in.

Pinyin also draws inspiration from Zhuyin in other ways, such as the tone accent marks and the spelling of bopomofo (rather than *buopuomoufuo).

Perhaps Zhuyin could have created a dedicated -ên symbol, but overall, Zhuyin is somewhat minimalist in its inventory of finals. For example, it could have made sense to have dedicated letters for -in and -ing, rather than representing those sounds with -ien ㄧㄣ and -ieng ㄧㄥ. In fact, the character ㄣ -en is derived from 乚 yǐn, even though yin and en are different syllables.

Bopomofo letters

  • ㄩㄥ (same as -iong, but written as -üeng) is also a weird peculiarity of Zhuyin.
    – gnucchi
    Commented Sep 24, 2020 at 17:15

It seems /ian/ can be a valid broad transcription of (Pinyin) ian, but [iɛn] and [ian] can be allophones. Pinyin is best thought of as broad transcriptions.

  1. Some Chinese Pinying scholars indeed say 盐 deflect to yien.
  2. For most chinese, ju,qu,xu can not be pronouced out. Only xü,jü,qü can work. However, writing ü is more labored than u, so u replaces ü when confusion does not occur.
  • What does "盐 deflect to yien" mean?
    – mic
    Commented Jul 28, 2019 at 17:31
  • The sound of 盐 in Putonghua is indeed close to yien, rather than yan. Some scholars have already stated this in published books
    – cauckf
    Commented Jul 29, 2019 at 0:54

Putonghua is based on one typical Northern dialect. Maybe now southern and northern people read the same character differently, as in [jεn] vs [jan] [in IPA symbol], but the romanization is based on the northern pronunciation.


The the vowel in "an" is pronounced differently depending on the context it occurs in, but the realisations have been analysed differently by different linguists:

Like the mid vowel, the low vowel has several variants, and there is again some disagreement on their surface values. [...]

In (55) I compare four proposals (the underscore in the top row indicates the location of the low vowel; [a] is front in all transcriptions).


[ j_n] [ɥ _n] [ _i]/[ _n]
S. Xu ( 1980) [æ] [ɐ] [a]
M. Fu ( 1956) [ɛ] [ɛ] [a]
Chao ( 1968) [ɛ] [a] [a]
Y. Lin ( 1989) [ɛ] [a]([ɛ]) [a]
  • Phonology of Standard Chinese, 2.10.3. The low vowel

While Chao and Xu make a distinction between "-ian" and "-üan" (as represented in Wade-Giles), Fu treats them as the same (as in Zhuyin and Pinyin). The latter treatment is consistent with Chinese rhyme harmony:

Similarly, since [an], [uan], [yan], and [iɛn] rhyme, they ought to have the same surface vowel. It is not necessary to distinguish the low vowels in [an], [uan], [yɐn], and [iæn], as S. Xu does, nor the low vowels in [yan] and [iɛn], as Chao does.

  • Ibid., 3.5.2 Rhyming groups

Your Answer

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

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