Need Chinese pronunciation in IPA

[shibo77]

Beijing Mandarin, in strict IPA.

Sorry to bother!

I have wondered about this for a long time, so now I want to see what others think...

I think trying to fit all the sounds of the world spoken by human into IPA is a bit ridiculous.

I think the IPA system is in need of modification.

I think any person with a little time to compare the IPA values given in the appendices of Chinese dictionaries would find some faults.

Traditionally,

(, is a hook)

b[p] p[p'] m[m] f[f]

d[t] t[t]' n[n] l[l]

g[k] k[k'] h[x]

j[tc,] q[tc,'] x[c,]

z[ts] c[ts'] s

zh[ts,] ch[ts,'] sh[s,] r[z,]

If this is so, I thought that baba was very different from papa.

Compare Chinese and English:

爸爸['papa]according to the dictionary

papa[pe'pa:] (flipped "e", rounded "a" without the hook at the top.)

American['pa:pa] (both rounded "a"s)

At first I thought it might be because of tones and the difference in vowels, but I just can't put English [p] in "papa" as the same as Chinese [p] in "爸爸". I thought it more close to , but not quite .

Compare Chinese and English:

爸爸['baba]closest IPA value according to me.

basket['ba:sk it]("i" without the dot at top, same [a] as Chinese, only longer [a:])

The same happens with the [d] or [t] in "道". [g] and [k]. They are said to be voiced and voiceless, that would be nice, but it doesn't seem quite right when I compare it to English. I would say the Hanyu Pinyins' b d g are more close to the IPAs' [d] [g]. Don't worry about the vowels, try saying "baba" "dow" "guh guh" with your heavy English accent, they would sound much closer to 爸爸, 道, 哥哥, than your "papa", "tow-el" "kuh kuh".

Also what are your opinions of Hanyu Pinyin's r traditionally [z,].

-Shibo

[Altair]

Compare Chinese and English:

爸爸['papa]according to the dictionary

I believe this correct.

papa[pe'pa:] (flipped "e", rounded "a" without the hook at the top.)

I believe this is not correct. I am not sure, but I think the English use an aspirated "p" for both consonants in this pronunciation. In IPA, this is spelled as [p] in broad usage or as [p'] in more specific transcrition. The aspiration makes it closer to Pinyin "p" than to Pinyin "b," although there is less aspiration.

American['pa:pa] (both rounded "a"s)

Here the first "p" is aspirated as IPA [p'], but I think the second one is just [p] and often identical to Pinyin "b."

I think that part of the problem is that most IPA transliteration are rather broad and so contain errors if you analyze them in detail. Most of the time, such detail is not very important for those with knowledge of the language. If, however, you try to follow the transcription exactly, problems emerge.

Compare Chinese and English:

爸爸['baba]closest IPA value according to me.

My understanding is that the IPA is voiced, but the Pinyin "b" is never voiced. English "b" is somewhat voiced at the beginning of an utterance, but not as much as in some other languages. For those who have trouble with pronouncing consonants with and without aspiration, should simply equate English "b" with Pinyin "b," but this is actually not correct.

One problem with the Chinese transcriptions is that I believe Pinyin "d" and "t" are alveolar (i.e., pronounced with the tip of the tongue against the gum ridge); whereas Pinyin "z" and "c" are dental (i.e., with the tip of the tongue against the upper teeth).

Also what are your opinions of Hanyu Pinyin's r traditionally [z,].

I have never seemed able to get a straight answer about this. Part of the problem for me is that the Mandarin retroflex consonants "sh" and "zh" seem slightly different from the counterparts in European languages. I am not sure if the difference is in the tongue position or in the spread of the lips, but I think I detect a definite difference. I would also find it odd, if Pinyin "r" is simply a voiced equivalent of Pinyin "sh." With voicing absent elsewhere, this would seem an oddly balanced phonetic system. In an imperfect world, I have opted for pronouncing "r" with what I think is a slightly grooved tongue, which seems appropriate for "r"-like consonants and would keep it within a more normally distributed consonant system.

What does everyone think?

[ala]

think any person with a little time to compare the IPA values given in the appendices of Chinese dictionaries would find some faults.

The same happens with the [d] or [t] in "道". [g] and [k]. They are said to be voiced and voiceless' date=' that would be nice, but it doesn't seem quite right when I compare it to English. I would say the Hanyu Pinyins' b d g are more close to the IPAs' [b'] [d] [g]. Don't worry about the vowels, try saying "baba" "dow" "guh guh" with your heavy English accent, they would sound much closer to 爸爸, 道, 哥哥, than your "papa", "tow-el" "kuh kuh".

I believe the two concepts you are looking for Shibo are lenis and fortis.

A fortis consonant is a “strong” consonant produced by increased tension in the vocal apparatus. These strong consonants tend to be long, voiceless, aspirated, and high.

A lenis consonant is a “weak” consonant produced by the lack of tension in the vocal apparatus. These weak consonants tend to be short, weakly voiced or voiceless, aspirated, low, and the following vowel tends to be lengthened.

Most western languages have b as voiced, p as voiceless (either aspirated or unaspirated). Designers of Hanyu Pinyin (and Icelandic script) decided to assign b as lenis (in Mandarin lenis is voiceless unaspirated) and p as fortis (voiceless aspirated). They have their point, most western languages' voiced consonants are lenis, and voiceless consonants are fortis. So in other words to you especially when comparing with a weakly voiced language such as English, you might think 哥哥 sounds like English guh-guh. But 哥哥 is very much VOICELESS, while English guh-guh is still voiced. Both are lenis though (that's probably where you think they are similar). The IPA isn't wrong still.

English: voiced = lenis, no voiceless unaspirated, voiceless aspirated = fortis

Mandarin and Cantonese: no voiced, voiceless unaspirated = lenis, voiceless aspirated = fortis

French: voiced = lenis, voiceless unaspirated = fortis, no voiceless aspirated

Shanghainese: voiced = lenis, voiceless unaspirated = fortis, voiceless aspirated = fortis

[shibo77]

Thank you ala and Altair!

The lenis/fortis really helped!

The IPA for "papa"'s Received Pronunciation was from the Shorter Oxford English Dictionary. The IPA for "papa"'s General American was from the American Heritage Dictionary, which I changed the AHD into IPA.

-----------------------------------------------------------------------------------

Could I say this:

b[b o] p[p'] m[m] f[f]

d[d o t[t'] n[n] l[l]

g[g o] k[k'] h[x ]

j[dz{ o] q[tc{] x[c{]

z[dz o] c[ts'] s

zh[dz, o] ch[ts,'] sh[s,]

? --glottal stop

, --hook

{ --little flourish

o --voiceless

I made changes to the voiced consonants, b, d, g, j, z, zh. Why not write b with a small circle "o" From what I understood from the previous posts, Hanyu Pinyin's "b, d, g, j, z, zh" are voiceless unaspirated (Mandarin lenis). Hanyu Pinyin's "p, t, k, q, c, ch" are voiceless aspirated (Mandarin's fortis). But would that not equate Hanyu Pinyin's "b, d, g, j, z, zh" with voiceless equivalents of IPA's , [d], [g], [dz{], [dz], [dz,], in other words, IPA , [d], [g], [dz{], [dz], [dz,] with little circles denoting the voiceless quality. Rather than merely unaspirated equivalents to IPA 's aspirated [p], [t], [k], [tc{], [ts], [ts,]?

VOICELESS UNASPIRATED equivalents of voiced consonants

b[b o], d[d o], g[g o], j[dz{ o], z[dz o], zh[dz, o]

VOICELESS UNASPIRATED equivalents of voiceless consonants

b[p] d[t] g[k] j[tc{ '] z[ts'] zh[ts, ']

VOICELESS ASPIRATED equivalents of voiceless consonants

p[p'], t[t'], k[k'], q[tc{ '], c[ts'], ch[ts, ']

I don't know, maybe it's because of my voice, but I usually pronounce Hanyu Pinyin's "b" as in "菠菜bo1 cai4" as almost voiced. Saying [p..] feels awkward.

---------------------------------------------------------------------------------

I am still not sure about r. But I didn't quite understand. At the moment, Hanyu Pinyin's sh and r are given the IPA values [s,] [z,]. Which are simply voiceless, and voiced equivalents of a fricative retroflex. For me there is a definite "r.." sound there, and sound more like a flapped retroflex r[r,] than a fricative retroflex r[z,]. What do you think? Trying saying 热re4 or 日ri4, which sound more [r,] than [z,].

---------------------------------------------------------------------------------

Another thing from the "How to pronounce 'wu'". I gave it some more thought after my last post on that thread. From a traditional Chinese phonological view, a啊 o喔 e鹅 i衣 u乌 ü迂, and the y- and w- are merely stylistic in purpose. This sounds nice, but traditional Chinese phonology isn't that good.

From an IPA view. Hanyu Pinyin's y[j] and w[w] are semi-vowels/semi-consonants. In IPA, would 衣yi1 be or [?i]?

Would 乌wu1 be or [?u]? [ji] and [wu] are ruled out as they are not pronounced at all. I think this might solve the problem of wheather there is a glottal stop or not. By the way, thanks for the description about the glottal stop!

Opinions! Thanks in advance!

-Shibo

[Altair]

I made changes to the voiced consonants, b, d, g, j, z, zh. Why not write b with a small circle "o"

I think such circles indicate that a consonant has been devoiced. Given that some of the equivalent Middle Chinese consonants may have been voiced, this is not an unreasonable choice; however, I think that p, t, k, etc. are better choices.

Rather than merely unaspirated equivalents to IPA 's aspirated [p], [t], [k], [tc{], [ts], [ts,]?

IPA indicates aspiration with an apostrophe. Broad IPA transcriptions often do not specify aspiration in languages like English, even when it is present; but in languages like Chinese, where aspiration is an important factor to consider, the simple symbols are assumed to be unaspirated. Where aspiration is present, it should be marked with an apostrophe. This explains why the Wade-Giles system has “Tao k’o tao” for Pinyin “Dao ke dao.” Wade-Giles is more faithful to the IPA origins.

I think you may be proceeding from the assumption that IPA consonants are based on English pronunciation; but I do not think this is true. I think also that in most languages “b, d, and g” differ from “p, t, and k” on a number of issues, but these differences vary from language to language and the importance of the various differences varies. In addition to fortis and lenis, there is length of voicing, pitch of voicing, explosive vs. implosive, glottalized, ejective, and many other possible features. At its simplist, simply connotes a voiced bilabial plosive consonant and [p] connotes its unvoiced equivalent. All the rest has to be specified if it cannot be assumed for the language in question.

Take Cantonese as an example. Cantonese makes the same distinctions as Mandarin between Pinyin b, d, and g and Pinyin p, t, and k; however, there is no contrast in sentence final position. In this position, the Cantonese consonants are unreleased and so one cannot tell whether they represent unreleased allophones of the first series or of the second. Linguists seem to have settled for writing these sounds as p, t, and k; but one must realize that these consonants actually differ audibly from how such sounds are normally pronounced in languages like French, where final consonants are normally released.

From an IPA view. Hanyu Pinyin's y[j] and w[w] are semi-vowels/semi-consonants. In IPA, would 衣yi1 be [ji] or or [?i]?

My current guess is that Pinyin’s “yi” is normally IPA [?i] when at the beginning of an utterance, but otherwise IPA . What is still unclear to me is how important the glottal stop is in Mandarin. Clearly it is not a basic part of the phonetic system, and so the question arises as to how optional it is or rather what are the conditions under which it surfaces.

Would 乌wu1 be [wu] or or [?u]? I think this might solve the problem of wheather there is a glottal stop or not.

I do not recall every hearing [?u]. To my ear, I hear more than [wu] at the beginning of an utterance. I would also find [wu] a little odd, since it would be a somewhat unusual for a language to keep a sequence of hard to articulate sounds that does not contrast with anything. If it derives from Middle Chinese “ng,” I would quicker expect the consonant to be dropped than to change into a “w.”

[shibo77]

Right, I looked into it a little bit more. There it is! There are 2 sets of IPA values being assigned. One is the familiar, p, p', m, f, and another (bo, p', m, f), which I think is used by people dealing with phonology.

Hanyu Pinyin's "r" is an approximant retroflex, not a flapped retroflex. "r" inverted and flipped, with a hook at the bottom.

' -tiny h at upper right, aspiration

, -hook, retroflex

{ -flourish, alveolo-palatal

o -circle, unvoiced

Set 1:

B[p] P[p'] M[m] F[f]

D[t] T[t'] N[n] L[l]

G[k] K[k'] H[x]

J[tc{] Q[tc{ '] X[c{]

ZH[ts,] CH[ts, '] SH[s,] R[z,]

Z[ts] C[ts'] S

Set 2: (I also found a website: http://www.angelfire.com/pop2/pkv/ipa.html )

B[bo] P[p'] M[m] F[f]

D[do] T[t'] N[n] L[l]

G[go] K[k'] H[x]

J[dz{ o] Q[tc{ '] X[c:]

ZH[dz, o] CH[ts, '] SH[s,] R[flipped/inverted r,]

Z[dzo C[ts'] S

It's not that one is the early values, and the other, the later values. The two sets are in different dictionaries at about the same time, and are still used today.

------------------------------------------------------------------------------------

From what I can find, there is no glottal stop in 官:北京Beijing Guan (Pekinese). But it does exist in several other dialects of 官Guan (Mandarin), and existed in Middle Chinese and probably the early part of 官Guan (Mandarin). The glottal stop is found more to the regions near the长江Changjiang(Yangtze river). It also exists in the dialectal groups 吴Wu (Shanghainese) and 闽Min (Fukkienese) . As for most of 官Guan (Mandarin), Hanyu Pinyin's "衣yi1" and "乌wu1" are simply 衣[i55] and 乌[u55].

-Shibo