In which cases is Hanyu Pinyin not phonetic?

[Demonic_Duck]

I don't think you can prove that a system is phonetic by reference to another not-entirely-phonetic system.

[li3wei1]

I wouldn't argue that Pinyin is phonetic. I was just making a comment about the pronunciations of 剥 and 夺.

I'd be interested to hear about how zhuyinfuhao is not phonetic, however. I never really studied it formally, and have forgotten most of it, but it always seemed to me to be one sound=one symbol and vice versa. What's not phonetic about it?

[Demonic_Duck]

I'd say zhuyin fuhao is roughly as phonetic as pinyin (i.e. mostly phonetic, but not entirely).

 

A good example of non-phoneticity is ㄩㄥ (yong).

[BanZhiYun]

Alright, we just did like 60% of the initials today, and we're not up to the finals yet, but the teacher has said that finals are much more complicated than initials. So far, for initials, when you have to categorise them, there are 2 criteria: 任何声母的音值，都是由发音部位和发音方法决定的。因此，声母可以按发音部位或发音方法进行分类。按发音部位，you got 7 types, 按发音方法，you got 5 types. Then, you have voiced and voiceless consonants, but in Mandarin, only "m", "n", "l", "r" are voiced, all the other consonants are voiceless. Then you have the 送气／不送气。

About finals' categorisation, 按韵母的结构特点（韵头，韵腹，韵尾）和发音情况，一般把韵母分为三类：单韵母，复韵母，鼻韵母。In 剥夺, 剥 (bo) is composed only of 韵腹 (o), while 夺 (duo) is composed of 韵头(u) and 韵腹 (o)，therefore their sound is definitely not the same. But 夺 (duo, -uo, is still a final) it's just that it's 结构 is different from 剥's 结构。剥 (bo) is basically  单韵母, 夺 (duo) is a 复韵母。

Then about rhyming in Chinese, it's not only about it's finals, but also about the tones and it's very different with Ancient Chinese and Modern Chinese. Their pronunciations are completely different. Mandarin doesn't include the 入声, so reading 古代诗歌 in Mandarin, often doesn't rhyme, even if it's the same final.

Then again, you can not look @ vowels and consonants on the same level as syllables. Vowels/Consonants are only a composing part of a Chinese's syllable, when you analyse a syllable you should look at it's initials and finals. Vowels/Consonants are the main composing parts of initials/finals (respectively, for the most part, besides some exceptions).

Hope this makes sense, I put it in the simplest way I could.

[Altair]

So-called phonetic writing schemes actually tend to be more phonemic than phonetic, since phonetic rendering causes all sorts of problems in representing different dialects and capturing differences that are unimportant and even hardly detectable by native speakers.  There is also the problem of how narrowly or how broadly to represent the phonetic differences.  For instance, the three "i's" in "bit," "bid," and "fill" are all phonetically different in most English dialects I am aware of, but hardly merit different representation in English spelling, since they all represent the same phoneme.  (If you can't detect the differences, notice that the first "i" is short in length, the second is pronounced about twice as long, and the third is followed by an off glide that could have been spelled "fi-ull."

 

Another issues is that the different Latin-based Chinese writing and transcription systems were created based on slightly different dialects of Mandarin and at different times.  These pronunciations have all shifted even over the last 150 years or so.

 

Here is some interesting stuff from Wikipedia about Wade-Giles and its representation of /uo/.  Wade-Giles was developed in the 1860s.  Kudos to Roddy for giving me the idea to check that article.

 

Vowel o

Final o in Wade–Giles has two pronunciations in modern Mandarin: [u̯ɔ] and [ɤ]. What is pronounced today as a close-mid back unrounded vowel is written usually as ê as in pinyin, but sometimes as o, depending on historical pronunciation (at the time Wade–Giles was developed). Notably, after velar initials k-, k'- and h- (and a historical ng-, which has been dropped by the time Wade–Giles was developed), o is used for characters like "哥" (Wade–Giles ko, pinyin ge), though ê also exists after velars, like in "刻" (Wade–Giles k'ê, pinyin ke). Ê is used in other environments. By modern Mandarin, -o after velars (and what used to be ng-) have shifted to -ê, thus they are written as ge, ke, he and e in pinyin.

 

What is pronounced today as -uo is virtually always written as -o in Wade–Giles, except for characters like "說" (pronounced shuo even back at the time Wade–Giles was developed) and the three syllables of kuo, k'uo, and huo (as in 過, 霍, etc.), which contrasted with ko, k'o, and ho that correspond to pinyin ge, ke, and he. This is because characters like 羅, 多, etc. (Wade–Giles: lo, to; pinyin: luo, duo) did not originally carry the medial -u-. By modern Mandarin, the phonemic distinction between -o and -uo has been lost (except in interjections when used alone), and the medial -u- is added in front of -o, creating the modern -uo.

 

IPA pu̯ɔ pʰu̯ɔ mu̯ɔ fu̯ɔ tu̯ɔ tʰu̯ɔ nu̯ɔ lu̯ɔ kɤ kʰɤ xɤ ʈʂu̯ɔ ʈʂʰu̯ɔ ʐu̯ɔ ʦu̯ɔ ʦʰu̯ɔ su̯ɔ Wade–Giles po p'o mo fo to t'o no lo ko k'o ho cho ch'o jo tso ts'o so Bopomofo ㄅㄛ ㄆㄛ ㄇㄛ ㄈㄛ ㄉㄨㄛ ㄊㄨㄛ ㄋㄨㄛ ㄌㄨㄛ ㄍㄜ ㄎㄜ ㄏㄜ ㄓㄨㄛ ㄔㄨㄛ ㄖㄨㄛ ㄗㄨㄛ ㄘㄨㄛ ㄙㄨㄛ Pinyin bo po mo fo duo tuo nuo luo ge ke he zhuo chuo ruo zuo cuo suo

 

 

Based on the above and especially the part I have put in bold, it appears that the phoneme /ɔ/ has not only changed pronunciation over the years, but also at different rates in different syllables, thus suggesting different written representations.  The boundaries between /uɔ/ and /ɔ/ have also changed.

 

On top of all these complications is the fact that different linguists analyze Mandarin phonology differently.  For instance, Wikipedia states the following:

 

 

The glides can also occur in medial position, that is, after the initial consonant but before the main vowel. Here they are represented in pinyin as vowels: for example, the i in bie represents [j], and the u in duan represents [w]. There are some restrictions on the possible consonant-glide combinations: [w] does not occur after labials (except for some speakers in bo, po, mo, fo); [j] does not occur after retroflexes and velars (or after [f]); and [ɥ] occurs medially only in lüe and nüe and after palatals (for which see above.) A consonant-glide combination at the start of a syllable is articulated as a single sound – the glide is not in fact pronounced after the consonant, but is realized as palatalization [ʲ], labialization [ʷ], or both [ᶣ], of the consonant.^[9] (The same modifications of initial consonants occur in syllables where they are followed by a high vowel, although normally no glide is considered to be present there. Hence a consonant is generally palatalized [ʲ] when followed by /i/, labialized [ʷ] when followed by /u/, and both [ᶣ] when followed by /y/.)

 

I find the bolded material quite interesting for this discussion, and yet it illustrates the very difficulty of arriving at a definitive answer by not being internally consistent.  In one place, it says that the [w] glide exists after certain consonants, but not after labials (except for some speakers).  Then it says that no glide is actually separately pronounced, but simply results in a modification of the preceding consonant, such as a labialization in the case of [w].

 

A particular difficulty exists for native English speakers in analizing Mandarin glides, since English tends to treat glides such as [w] as semi-consonants, rather than as semi-vowels.  Where our speaking habits differentiate between and consonants and vowels, we treat them as consonants.  We say "an itch" and "thee itch," but "a ditch" and "thǝ ditch" and "a witch" and "thǝ witch."  Many other languages have a different situation.

 

In French, unlike English, such sounds are generally treated as semi-vowels and thus are classed with the behavior of vowels.  For example, in phrases such as "l'huile" and "les yeux" they trigger modifications in the pronunciation of the preceding word in the same way that vowels do.  In yet other languages, such as Russian and Irish, these glides are felt to be simple modifications in the pronunciation of the "preceding" consonant and not as separate sounds by themselves.  In both languages, most consonants generally have two pronunciations (depending on the historical glide).  In Russian, the two pronunciations are called "hard" and "soft."  In Irish, they are called "broad" and "slender."

 

In Mandarin, different dialects seem to have been on different places on a continuum between treating these glides as "semi-vowels" or as simple modifications of the preceding consonant.  We have this cross-dialectally and chronologically as represented in the two spellings: "Nanjing" and "Nanking."  The former represents the current palatalized pronunciation of the Beijing-base standard, and the latter apparently represents the Jianghuai Mandarin of the Ming Court, where the same degree of palatalization did not occur.  Similar, some speakers realize /bɔ/ as [pɔ], realizing the glide as a mere labialization that merges with the labial consonant; and other speakers realize /bɔ/ as [pu̯ɔ], realizing the glide as a semivowel that persists in all environments.

[Guest realmayo]

That's interesting, will drop Olle a line about that. Wonder if it's a Taiwan thing. Or just my ears 

 

 

Roddy if you're the same as me it might be your eyes, not your ears! My guess is you hear and produce these sounds the same, but when you stop and try to analyse it, you're misled by how pinyin is written. At least, that seems to be the case with me.

 

According to The Sounds of Chinese (Yen-Hwei Lin): the pinyin bo po mo fo are only written with an 'o' because of a spelling convention; the vowel is in reality 'uo', the same as duo guo zuo shuo.

 

When I read that it didn't seem right. But then I remembered someone I know, who can't speak any Chinese at all, stubbornly transcribing 莫 as Mua, based on how I spoke the pinyin mo. I couldn't work out where this person was hearing the 'u' glide from, but it was based only on what was coming out of my mouth. So clearly, I was pronouncing the pinyin mo with a 'u' glide, and basing that pronunciation on how I heard Chinese people speaking. But my 'knowledge' that pinyin mo has no 'u' was making me think that that was no glide, despite my producing it every time.

 

Meanwhile, my first reaction is that bo po mo fo are quite far from duo guo zuo shuo. But now I think about it, I don't produce a great big strong exaggerated 'w' sound when I say duo guo zuo shuo, and nor do the Chinese speakers that I know. It's not 多少钱 dwwwoorshao qian, it's far more subtle than that. In fact, just about as subtle as when I produce an bo po mo fo. But seeing the 'u' there in the pinyin has convinced me, when I'm sat down thinking about it, that there's a strong glide. But when I speak I only produce a mild one.

 

 

I'm tempted to say this is a downside of pinyin, that learners should be taught clearly that buo puo muo fuo have their 'u' deleted simply as a spelling convention ... but then again, why bother waste learners' time with this, after all I still say these sounds correctly, it's just that I don't analyse them correctly.

[roddy]

This is very interesting and I am now starting to doubt myself. Also found this, but don't have time to read it properly.

http://m.doc88.com/p-2915510991805.html

So Realmayo, if you listen to that link I posted earlier, or find a few relevant pronunciations in Pleco?

[Guest realmayo]

How about 薄膜 from Pleco's audio? Much clearer than I was expecting, I hear a glide. Or is it just me?

 

薄膜.mp3

 

 

 

薄膜.mp3

[Guest realmayo]

From Altair above:

 

characters like 羅, 多, etc. (Wade–Giles: lo, to; pinyin: luo, duo) did not originally carry the medial -u-. By modern Mandarin, the phonemic distinction between -o and -uo has been lost (except in interjections when used alone), and the medial -u- is added in front of -o, creating the modern -uo.

 

So am I right to hear 多 pronounced slightly differently in 多么好! and 多少钱?

[roddy]

You're right, but then 塑料薄膜 in the same entry seems closer to what I would expect. (Responding to two posts above)

[Olle Linge]

 

Spelt o, pronounced as uo

The syllables bo, po, mo, fo are actually pronounced buo, puo, muo, fuo (listen), which means that they rhyme with duo, tuo, nuo, luo and so on (listen and compare). Using IPA, this sound is written [woo]. For example, bo and duo are written as [pwoo] and [twoo] respectively. As we can see, these syllables rhyme.

 

This is an interesting question. I haven't done any empiric research into the difference between "-o" and "-uo", but it's safe to say that there are both regional and personal differences. To figure out which is which and to what extent would require more serious research. Personally, I think the sounds are more similar than they are different and I would transcribe them the same way. Naturally, there will always be small differences, but then there are small differences between all syllables.

 

Now, let's look at two clear examples of each case:

This is a clear case of where "-o" (in "bo") is pronounced as a single vowel sound.

This is a clear case of where "-o" is pronounced as a diphthong (click "introduction to Pinyin", then find "bo")

 

If we want to transcribe these specific sounds as spoken by these specific persons, then obviously we need to write them differently. However, I think the first sound here is not good and I would mark such a pronunciation as incorrect on an exam or pronunciation check if I decided to be picky. The Pleco clips linked to sounds good to me.

 

However, that doesn't end the discussion, because there are many variants in between. The "-uo" diphthong is pronounced by starting with tight lip rounding and then opening it up. Since it's also the case that all the initials that can be followed by "-o" are labial in some way and are thus produced as far forward as possible with the lips, you can't transition immediately from that to the final vowel sound; there is a transition.

 

So, either you treat the intervening lip-rounded sound as part of the initial and transcribe e.g. "bo" and "duo" differently, or you make the intervening sound explicit by writing it out. In that case, the transcription is the same or very similar. Which one you choose might depend on the speaker. I agree than the "u" part is longer and clearer in "duo" than it is in "b(u)o", but it's definitely there. If you want a more accurate answer than this, you need to find an empiric study or conduct one.

 

From a pedagogical standpoint, however, I always treat the sounds as the same. In my experience, doing anything else results in readings that are much closer to the first example I linked to above, which is not good. Naturally, this also ties into the much bigger question about how to treat transcription systems when teaching pronunciation, but I feel that's a much bigger issue and slightly off-topic.

[Guest realmayo]

Isn't uo a glide rather than a dipthong?

[Guest realmayo]

From The Phonology of Standard Chinese by San Duanmu:

 

For many SC speakers, the mid vowel is [oː] in an open syllable after the

labials [p, ph, m, f], where the labials become rounded, too. On the other

hand, some speakers use [ɤː] in this environment, where the labial is not

rounded. This is shown in (40).

 

(40) The mid vowel after a labial onset in SC

Variety A                Variety B

[pwoː]                    [pɤː]      ‘waves’

[pwhoː]                   [phɤː]     ‘slope’

[mwoː]                   [mɤː]     ‘to touch’

[fwoː]                     [fɤː]       ‘Buddha’

 

[jh17710]

There is no case to show that Pinyin is not phonetic. Pinyin is just not a good system for phonetic symbols because there are too many regulations/exceptional rules in Pinyin.

10/24/15 Friday said:

• In 我, “o” sounds more like “au” in “Australia”.

• In 周末, “o” in “末” sounds more like “o” in “okay”.

• In 萝卜, “o” in “卜” sounds like it has a very light “w” sound between the “b” and “o”, like “bwo”. I found transcripts using Zhuyin Fuhao which include the“w” sound in the spelling of this word.

The transcripts using Zhuyin Fuhao which include the “w” sound in the spelling of this word “卜” might be due to another meaning for the same character in traditional character system which means “foretell”. The "w" in front of 我 caused the difference naturally. For “末”, it does sound more like “o” in “okay” to me too. For the 萝卜, “o” in “卜”, the original design for bo,po,mo,fo in Pinyin and Zhuyin Fuhao were bwo,pwo,mwo,fwo with omitted "w". There are still some scholars ask teachers to teach bwo,pwo,mwo,fwo instead of the current practice in the real world.

Language is such a changing thing that when IPA defined a symbol for a language, the way that symbol is handled in the real world may change or even disappear within three generations, like 60 years. To me, the 37 IPA symbols for Chinese should reduce to 34 because the symbols for the j,q,x in Pinyin are actually z.c.s combined with the vowel "i". Just try it for yourself.

IPA is too huge a phonetic system to follow the changes of languages in the real world. I think people should focus on phoneme and adopt the phonemic spirit in between different languages like Chinese and English by ignoring the minor differences and merge the related IPA symbols whenever the mergence will not cause confusing in the meaning of words in both languages. By doing so, the numbers of symbols in IPA will reduce a lot!

10/27/15 Altair said:

For instance, the three "i's" in "bit," "bid," and "fill" are all phonetically different in most English dialects I am aware of, but hardly merit different representation in English spelling, since they all represent the same phoneme. (If you can't detect the differences, notice that the first "i" is short in length, the second is pronounced about twice as long, and the third is followed by an off glide that could have been spelled "fi-ull."

I think the "about twice as long" is a reasonable guessing, long enough for people to move the tongues from "t" position to "d" position. Regarding the "fill", I think "ill" sounds like a word of two-syllables so that the "ull" is more like the second syllable of "fill", to me.

[Olle Linge]

Isn't uo a glide rather than a dipthong?

 

Yes, sloppy of me. A glide plus a vowel, at least if we follow his way of transcribing it (which I agree with).

 

For many SC speakers, the mid vowel is [oː] in an open syllable after the

labials [p, ph, m, f], where the labials become rounded, too. On the other

hand, some speakers use [ɤː] in this environment, where the labial is not

rounded.

 

I don't think this is the same situation, though. [ɤː] isn't any of of the sounds we are talking about here. If you look at Duanmu's transcription of syllables at the end of the book , you will see that he has chosen to transcribe "-uo" and "-o" the same way, i.e. what I suggested in my article.

[Guest realmayo]

I don't think this is the same situation

 

I thought the quote was relevant because ... if the glide comes from the rounding of the lips ... and some speakers don't round their lips ... then they won't be making the glide ... and so the 'w' sound won't be produced. But just speculative.

[jh17710]

I agree with realmayo. I don't round my lips for "bo, po, mo, fo". After so many years of omitting "w" for "bwo, pwo, mwo, fwo" there are a lot of "bo, po, mo, fo" speakers in the real world.

As we all know, saying "bo" is quicker than saying "bwo". Why do we have to follow the old traditional way? Just let "bo, po, mo, fo" be added to the Chinese pronunciation, formally.

[Maijax]

拼音 has some confusing points. I had to use 注音符号 to clarify 拼音. After I did that, I never had a problem with 拼音

[jh17710]

There is no case to show that Pinyin is not phonetic. Pinyin is just not a good system for phonetic symbols because there are too many regulations/exceptional rules in Pinyin.

10/24/15 Friday said:

• In 我, “o” sounds more like “au” in “Australia”.

• In 周末, “o” in “末” sounds more like “o” in “okay”.

• In 萝卜, “o” in “卜” sounds like it has a very light “w” sound between the “b” and “o”, like “bwo”. I found transcripts using Zhuyin Fuhao which include the“w” sound in the spelling of this word. [end quote]

The transcripts using Zhuyin Fuhao which include the “w” sound in the spelling of this word “卜” might be due to another meaning for the same character in the traditional character system which means “foretell”. The "w" in front of 我 caused the difference in vowels naturally. For “末”, it does sound more like “o” in “okay” to me too. For the 萝卜, “o” in “卜”, the original design for bo,po,mo,fo in Pinyin and Zhuyin Fuhao were bwo,pwo,mwo,fwo with omitted "w". There are still some scholars ask teachers to teach bwo,pwo,mwo,fwo instead of the current practice in the real world.

Since language is such a changing thing that when IPA defined a symbol for a language, the way that symbol is handled in the real world may change or even disappear within three generations, like 60 years. To me, the 37 IPA symbols for Chinese should reduce to 34 in the 21st century because the symbols for the j,q,x in Pinyin are actually z.c.s combined with the vowel "i". Just try it for yourself.

IPA is way too huge a phonetic system to update the changes of languages in the real world. I think people should focus on phoneme and adopt the phonemic spirit in between different languages like Chinese and English by ignoring the minor differences and merge the related IPA symbols whenever the mergence will not cause confusing in the meaning of words in both languages. By doing so, the numbers of symbols in IPA will reduce a lot!

10/27/15 Altair said:

For instance, the three "i's" in "bit," "bid," and "fill" are all phonetically different in most English dialects I am aware of, but hardly merit different representation in English spelling, since they all represent the same phoneme. (If you can't detect the differences, notice that the first "i" is short in length, the second is pronounced about twice as long, and the third is followed by an off glide that could have been spelled "fi-ull." [end quote]

I think the "about twice as long" is a reasonable guessing, long enough for people to move the tongues from "t" position to "d" position. Regarding the "fill", I think "ill" sounds like a word of two-syllables so that the "ull" is more like the second syllable of "fill", to me.

[Olle Linge]

I thought the quote was relevant because ... if the glide comes from the rounding of the lips ... and some speakers don't round their lips ... then they won't be making the glide ... and so the 'w' sound won't be produced. But just speculative.

 

I didn't mean to say that it was irrelevant, just that it wasn't the same difference as the one that was highlighted earlier (such as in my post). However, I'm not sure if we're talking about the same thing here. Do we have a recording of this unrounded version? Is it something like the ㄅㄆㄇㄈ read by ordinary people in Taiwan, which is closer to a central, unrounded vowel? It seems we have three variants here.

 

I think the standardised sound is definitely rounded, which I've verified by checking with native teachers with certified standard pronunciation. That doesn't mean that everyone has to speak like that or that it's the only way or anything, but that's a different discussion. If we're after describing sounds as they are spoken in China, then we need several different transcriptions. If we're after teaching students, I'd stick with the lip-rounded one unless there is a very specific and clear reason not to.

I don't play the pedagogy card to kill the debate, but I think anything beyond this really needs empiric research. There might be papers or articles about this already, I haven't looked, but we have already settled that there is certain variation. It would be interesting to know more about how it varies, but we can't do that here, I think. It's a very interesting topic, though!