Hanji pronunciations in Taiwanese

[FutureSpy]

On SimL's thread on "Names of Radicals", I was to ask that:

I read on Wikipedia that Quanzhou has a third pronunciation they call "vulgar pronunciation". Anyone has any clues if there's such a thing in Taiwanese too? And if not, how useful could it be learning them from Quanzhou sources?

But I was afraid of leading that thread out of its discussion scope (my question had nothing to do with it), so I moved it here.

Then I got all hanji on the 1st lesson of my textbook and added its respective literary and colloquial pronunciations (not sure about accuracy tho --sources are: 台語辭典(台日大辭典台語譯本)查詢, 臺灣閩南語常用詞辭典 and 台文/華文線頂辭典):

中 tèng/tìng(p), tiong/tiòng(l), chhian(?)
也 iā/iá/ā/á
京 kiaⁿ(p), keng/king(l)
人 lâng(p), jîn/lîn(l)
你 lí
來 lâi, liô(?)
北 pak(p), pok(l), phian(?)
問 mn̄g(p), būn(l)
國 kok
學 o̍h(p), ha̍k(l)
對 tùi/ùi
師 sai(p), su(l)
我 góa(p), ngó͘(l)
日 ji̍t/li̍t
是 sī, hiō(?)
時 sî
本 pńg(p), pún(l)
東 tang(p), tong(l)
港 káng
灣 oaiⁿ/oan
無 bô/bû
生 siⁿ/seⁿ/chheⁿ/chhiⁿ(p), seng(l)
當 tànng/tǹg/tǹg/tng(p), tòng/tong(l)
老 lāu(p), láu/ló/nó͘(l)
臺(台) tâi, thai(?)
請 chhiáⁿ(p), chhéng/chhíng(l)
香 hiuⁿ/hiaⁿ(p), hiong/hiang(l), phang(?)

Now I have a few questions:
1. Why sometimes 台語辭典(台日大辭典台語譯本)查詢 give weird pronunciations? (I marked those with a "?" and bold)
2. Where there's an e/i (or a/o) variation (tèng/tìng, chheⁿ/chhiⁿ, hiong/hiang, etc.) should I stick with only one?
3. I see a lot of tonal variations. Do I have to memorize which ones take one tone and which ones take the other tone? Any possible explanation to why that happens?

[SimL]

Hi FutureSpy,

It's so great to see how enthusiastic you are about learning Taiwanese/Hokkien!

What follows is just my personal opinion. The "*", "**", "***" after three points are footnotes.

Hokkien literary vs. colloquial pronunciations are quite a distinctive part of Hokkien (as you and other regular Forum members undoubtedly know). Apparently, other Sinitic languages have it too*, but it's particularly prevalent in Hokkien. That having been said, I wonder if it's that useful - for a learner - to proceed from a list of the colloquial AND literary pronunciations, for each character. As you probably realise, the colloquial pronunciation is (generally) the "commonly occurring" one, and the literary pronunciation will be encountered in more learned/intellectual vocabulary**. Perhaps it would be easier to just learn each pronunciation as you come across it in a compound.

To give you an example: 五 has the colloquial pronunciation "gO5", and the literary pronunciation "ngO2". The only compound I can think of - off the top of my head - which uses "ngO2" is 五香粉 "ngO2-hiong1-hun2", the famous "five-spice powder" used in cooking. So, it's easier for a learner just to learn "gO5" as the word for "five", with the character 五, and then, if they ever come across the term 五香粉, then they can learn that in that compound, the first character is pronounced "ngO2", and that it's the character's literary pronunciation***. Similarly, you can just learn that 學 is "oh" (colloquial), and then for a few compounds, like 學生 and 大學, it's "hak8-seng1" and "tai3-hak8".

With this approach, you save yourself from learning (one or more) literary pronunciations for a character because you might not - in the first 5 years - come across a compound where the literary pronunciation of that character is needed.

Like I said, this is just a personal opinion, and in no way am I attempting to say how you should approach learning Hokkien, with respect to this issue. If you have a very structured approach, then perhaps learning from lists where both colloquial and literary pronunciations are given is a good way for you. [I have my own way of learning Mandarin, which nobody else uses, but it works for me. It could even be argued that my way is very inefficient and non-productive, but it works for me.] Also, my opinion that "the colloquial pronunciation is (generally) the 'commonly occurring' one, and the literary pronunciation will be encountered in more learned/intellectual vocabulary" might be just an illusion caused by the fact that my Hokkien vocabulary is so limited. If this is the case, then your approach would of course make more sense.

PS. In Penang, one of the most common uses of literary pronunciation is in personal names (undoubtedly in other variants of Hokkien too). So, 月 and 龍 - two very common characters for the names of girls and boys (respectively) - will often be pronounced "guat8" and "liong5" in names (but can also occur as "gueh8" and "leng5" in names). For girls, 玉 with literary "giok8" instead of colloquial "gek8" is also very common, but here too, "gek8" is known to occur.

PPS. [And here, I'm completely outside my area of competence!] Isn't my suggestion above a bit like how one could learn Japanese with regards to the "kun"- and "on"-readings? Instead of getting a list of characters and learning all the different readings for each character, one could just learn the "kun" or "on" form in words as one comes across these words. This save a lot of additional memorization, because there might be many characters where one reading is very common, and the other(s) much less, or extremely rarely.

---

Additional information (marked with asterisks in the text above, but not included there in order not to break the flow of the main line of discussion):

*: Someone explained to me that "even" Mandarin has/had colloquial and literary pronunciations, but that: 1) They were much fewer than in Hokkien anyway. 2) In the standardization of Putonghua, they "got rid of" many of them (just by declaring them non-standard - that's the luxury one can afford oneself when creating a standard out of nowhere). Apparently, some of the characters which can be read in two different tones are a remnant of this: one tone is the literary pronunciation, and the other is the colloquial. [Perhaps the "dai4" in 大夫 "dai4fu0" (大 is normally pronounced "da4") is also a remnant of this. I don't know if any other Forum members can confirm or refute this...?]

**: This is a very broad generalization, particularly for Hokkien. It's generally true for other Sinitic varieties, but for Hokkien, literary pronunciations will be found in very normal everyday terms too. [And conversely, sometimes - as in the case of Penang Hokkien - colloquial pronunciations will be found in a few items of learned/intellectual vocabulary - even where other varieties of Hokkien already have a well-established literary pronunciation for the same vocabulary items. Penang Hokkien "tua3-oh8" is the most well-known example.]

***: In one of the earlier threads on this Forum (www.chineselanguage.org/forums/viewtopi ... 71&p=27644), I did a survey of the literary pronunciations of the digits 1-10. Here you'll get a flavour of what I mean about the literary pronunciations being restricted to more learned/intellectual vocabulary.

[amhoanna]

First off, Sim just gave a ton of very sound advice.

A few things to add on my part:

I read on Wikipedia that Quanzhou has a third pronunciation they call "vulgar pronunciation". Anyone has any clues if there's such a thing in Taiwanese too? And if not, how useful could it be learning them from Quanzhou sources?

No idea what they're talking about. Can't trust everything U read on Wikipedia, U know. Feel free to post a link to the article. And, whatever it is, I'm pretty sure that if it exists in Coanciu, it'll also exist in Taiwanese.

Coanciu sources could work great, as long as U don't confuse yourself. There are also two or three different types of Coanciu dialects, and they're not esp. similar even in comparison to Ciangciu, if I'm not mistaken. For a linguist, and it seems like U are one, it would be relatively easy to go from Coanciu to the rest of Hokkien, and relatively hard to go the other way. For one thing, all 8 tone classes have been kept intact in Coanciu, at least in the city of Coanciu. One more caveat, though, is that the dialect of Coanciu has been changing fast, so sources from different eras may contradict each other.

Very useful 字典, based on Coanciu Hoklo:
http://solution.cs.ucla.edu/~jinbo/dzl/

1. Why sometimes 台語辭典(台日大辭典台語譯本)查詢 give weird pronunciations? (I marked those with a "?" and bold)

Let me address 香 = "phang" first. Phang is the Hoklo word for FRAGRANT. The "etymologically true" kanji is 芳, AFAIK. Meanwhile, the word (by def., written word) for FRAGRANT in Classic Literary Chinese is 香. Meanwhile, 香 is also associated with the Hoklo etyma "hiuⁿ" and "hiong".

So, 香 = "hiong" would be analogous to 山 = "san" in the JPnese context, whereas 香 = "phang" would be analogous to 山 = "yama". The analogy isn't perfect, since "phang" is in itself Sinitic. The shared feature is that in both cases the kanji is not "etymologically correct".

As for the other examples, no idea. Hiō is the only other word here I even somewhat know. I suspect these are beyond the grasp of the avg under-60 TWnese Hoklophone. Then again, I'm not even that. My tentative advice (not fully informed, etc.) is that U should ignore those as being idiosyncracies in the way the 台日大 is set up. 台日大 is probably unsuitable for this purpose, BTW.

2. Where there's an e/i (or a/o) variation (tèng/tìng, chheⁿ/chhiⁿ, hiong/hiang, etc.) should I stick with only one?

Short answer yes, but no penalties for straying. Whatever is easiest for U.

Details: -iong/-iang is generally Coanciu vs Ciangciu... All -eⁿ endings go to -iⁿ in Coanciu and Amoy, but Ciangciu has -iⁿ endings as well, just not on 生, etc.

Hiuⁿ / hiaⁿ seems like it might be different layers. I don't know any words using 香 / hiaⁿ. C.f. niû and niâ 娘 though.

Chhéng / chhíng is an attempt to illustrate the quality of the vowel. Nobody does this consistently, so I'd ignore it. Standard POJ uses -eng, but U'll see -ieng spellings too. The TW MOE adopted -ing. Not phonologically significant.

3. I see a lot of tonal variations. Do I have to memorize which ones take one tone and which ones take the other tone?

Short answer: they are DIFFERENT WORDS that just happen to be etymologically related, and happen to be attached to a certain kanji as well.

The relationship between words and kanji in Hokkien is very similar to what goes on in Japanese.

Any possible explanation to why that happens?

Short answer: different eras. Lāu is old-school "original" Hoklo. Ló is a "Tang pronunciation", imported during the Tang or later. Láu is intermediate between the two. THESE THREE ETYMA HAVE DIFFERENT MEANINGS, CONNOTATIONS OR USES IN HOKLO.

The same thing happens in Vietnamese. Bốn, tư and tứ all mean FOUR, but they are different words with different uses. Tứ is the Tang pronunciation, what they call "Sino-Vietnamese". It's associated with 四. Tư is a Sino loan too, but an earlier one, and not considered "Sino-Vietnamese". Ba is the "native" Viet word. Bốn, and tư too I believe, are written with "exotic" Vietogenic kanji...

Another analogy: agua and the acua- in acuático are different etyma with different uses, but they are etymologically related and would be etymologically associated with the same kanji, if kanji were involved. Other, more breathtaking examples no doubt abound.

Hokkien literary vs. colloquial pronunciations are quite a distinctive part of Hokkien (as you and other regular Forum members undoubtedly know). Apparently, other Sinitic languages have it too*, but it's particularly prevalent in Hokkien.

True, within "Sinitic", they are particularly prevalent in the "Min" languages. They're also prevalent in Vietnamese (though analyzed differently, as just being different words) and in Japanese. The "Wu" languages also have this going on.

This all goes back to the layering of the Sinitic elements in these languages -- Hoklo, Hokciu, Hinghwa, Hainamese, Japanese, Vietnamese, etc.

These languages have layers of Sinitic that can be traced to different bygone eras. These eras usually represent either high-water marks of "Chinese" civilization or periods of chaos when people and scholars fled -- always south.

These areas (Hokkian, Japan, Vietnam) were so remote from the centers of Sino-culture, that they could be "Sinfluenced" from the center only when Sino-culture was at full flower, or through the movement of people, and in both cases this Sinfluence may've indirect and lagging. Meanwhile, most of what is today China Proper was closer to the centers of Sino-culture, and so Sinfluence took place on a continual basis ... so that layers did not form to the same extent that we see in VNmese, JPnese and Hoklo...

[amhoanna]

Forgot to add, I think I understand what U're trying to do at this stage. It's not a bad idea. I did it too, kind of as a prelude to learning Hoklo. But a much easier way to do it would be to simply get a rhyme chart, if that's what it's called. There are two types. They complement each other, actually. I'm attaching screen grabs of the two. Not sure where to get a set, but they're "everywhere" in academia and bet there are some online for TWnese. Actually, the UCLA Coanciu resource seems to be the digitized version of one such chart. I can help U find one for TWnese if neces.

The screen grabs are for Haihong Hoklo, spoken in 広東, and with this dialect the adjustment to Mainstream TWnese is not huge.

擷取 2.JPG (117.58 KiB) Viewed 40805 times

擷取.JPG (142.41 KiB) Viewed 40805 times

[SimL]

Hi amhoanna,

Let me address 香 = "phang" first. Phang is the Hoklo word for FRAGRANT. The "etymologically true" kanji is 芳, AFAIK. Meanwhile, the word (by def., written word) for FRAGRANT in Classic Literary Chinese is 香. Meanwhile, 香 is also associated with the Hoklo etyma "hiuⁿ" and "hiong".

Good that you point this out. I think many people interested in Hokkien and characters - but who haven't gone into it very deeply - believe that 香 = "phang1". I certainly did for years, in the same way as I thought that 吃 = "ciah8". We develop some gut feeling for what cognates are between Mandarin and Hokkien, and sometimes we're wrong - misled by the semantic connection, as you explain. It was only in the last 5 years that I realised that these should be 芳 and 食. [It's cases like these where I'm completely comfortable with - and indeed supportive of - the 本字-enthusiasts.]

In connection with 芳, it's perhaps worthwhile pointing out that this has an alternative pronunciation "hong1". I never knew this until I discovered the characters for my (Baba) great-grandfather's name (by co-incidence, on the very same "marriage scroll document" I spoke about in another reply today!). Since my earliest years, I've known that his name was "Tan Hong Lim", but (as with so many Babas), I'd never wondered what the characters were. It's only when analyzing the scroll about 5 years ago that I found out that the characters were 陈芳霖.

Even if he were alive today (impossible, as he was born in the 1880's), he (and his siblings and parents) would probably not have been able to provide me with this information, as they were unable to read and write Chinese. I believe that names were chosen in those days based on what the horoscope said, so the characters were probably picked by someone advising the family.

Anyway, the main reason I went into this in a bit more detail is that there's something which has puzzled me for a number of years now (ever since I started learning Mandarin properly, and could begin to read characters). Namely, how some personal names are chosen. [Topic drift, topic drift!] My dictionary gives 霖 as meaning "continued rain", "long spell of rain", "copious rain". Now, a long spell of continuous, fragrant rain seems to be a rather unusual name to give someone... [No disrepect intended to my great-grandfather, I'm only speaking about this from a socio-linguistic perspective.] The same goes for quite a number of other Chinese names I come across. With some of the major scholars (or of politicians / historical figures) of the late 19th and early 20th century in China whose names I come across in (English) Wikipedia (or of members of my Chinese-educated, Sin-Kheh maternal family, for that matter), I wonder to myself: "Hmmm, bit of a strange name", when I look at the meaning of the compound.

As for the fact that 芳 can have a "ph-" or an "f-" initial, this matches quite well a number of other well-known ones. There's "put8" and "hut8" for 佛, and "png1" and "hong1" for the surname 方 (in this case, my maternal Sin-Kheh family's surname). My guess is that the "initial-h" form is the colloquial reading and the "initial-p" form the literary (because of 發 and 法 so commonly having the "initial-h"-form), but I could be wrong. [Actually, thinking about it, there's 風 ("hong1"), 方 ("hong1"), 法 ("huat4"), 發 ("huat4"), 粉 ("hun2"), 佛 ("hut8") with initial-h, and 放 ("pang3"), 房 ("pang5"), 分 ("pun1") with initial-p, so not really an overwhelming number in favour of initial-h. Does anyone know?]

[FutureSpy]

Thanks SimL.

Well, actually I never stopped to think about all that you wrote. I mean, when I took Japanese classes back in High School, teaching method for kanji was kinda conservative. Some teachers even included some mnemonic elements (not etymological ones tho) to aid memorizing, but basically it was a few characters + on'yomi + kun'yomi + vocabs list using that kanji and a test on next class. So instinctively that's more or less what I was trying to do here. Besides, I was planning to add radicals and perhaps the simplified form for each character as well, so that by the time I revisit (or should I say, actually study them properly) the lessons I could study every character on it.

But now to be honest, that system never really worked to me. I learned how to read most kanji I know reading or with song lyrics. And for those I never came across in real life and only learned at school, I probably can only remember their kun'yomi. So your ideas totally make sense to me. I think that would end up happening to me with Hokkien too (memorizing colloquial readings and not being able to remember literary readings in many cases). Still, having some well-organized charts to follow every lesson would come handy at some later time (or perhaps I just feel insecure if I don't have such fancy charts even though they aren't especially useful). Anyway I think I'm just wasting time on stupid things to put off getting to real learning

And at this time, I feel a little confused listing pronunciations for a character, as I never know if one pronunciation is equivalent to another, or they do happen in different words. I might even think of spending some extra bucks on that Hanji book re-edited by Maryknoll, but I have no idea of the contents...

BTW, long thread. I'll take a look at it later. Thanks.

I have some hanzi books for Mandarin I bought a few years ago, but I -NEVER- leafed it through... I'll see what it has to offer later.



Thanks amhoanna.

Here's the Wikipedia article I was talking about: http://en.wikipedia.org/wiki/Literary_a ... characters

For a linguist, and it seems like U are one

I'm actually an undergrad student in Computer Science in his last year wondering if he'll ever get to graduate... It's no joke, now I only have a bunch Math subjects slowing me down before going to intership, and there's a subject I'll be taking for 4th time next semester...

The dict seems down, but I'll take a look at it later.

Standard POJ uses -eng, but U'll see -ieng spellings too.

My textbook recordings had it pronounced as /i/, but when I first heard a Taiwanese pronouncing it I wrote down it as ie so that I'd remember to pronounce it that way.

Thanks for the enlighting explanations on tonal differences. As for the rhyme charts, sounds interesting. To what extend is the first table supposed to cover characters pronunciations? Is the second one more comprehensive?

I could perfectly leave hanji for later, but I'm really missing the writing/visual factor to reenforce what I'm learning.

[amhoanna]

Here's the Wikipedia article I was talking about: http://en.wikipedia.org/wiki/Literary_a ... characters

OK, got it. What they refer to as "vulgar" readings are just the situations where the kanji and the (spoken) etymon aren't etymologically related. 香 / phang would be one example. These examples abound in Hoklo, and in Japanese they've been institutionalized, but they're not very common in Canto, Mand, etc.

There are some very high frequency "vulgarisms" in almost any Hoklo kanji text. Examples:
一 / tsi̍t
人 / lâng
欲 / beh

What does this say about the Siniticity of Hoklo? Hmm...

And, these are shared across Hoklo, inc. Teochew -- they're not unique to Coanciu.

I'm actually an undergrad student in Computer Science in his last year wondering if he'll ever get to graduate...

So U're a scientific linguist. U can't afford to spit even one line of bullshit, b/c the computers will call it right away and return an error. Much needed in Hoklology.

[amhoanna]

To what extend is the first table supposed to cover characters pronunciations? Is the second one more comprehensive?

Yes. The first one only seems to give one for each combo. The second seems to be comprehensive. Sorry, haven't had time to take a good look myself. Slow computer.

As for the fact that 芳 can have a "ph-" or an "f-" initial, this matches quite well a number of other well-known ones. There's "put8" and "hut8" for 佛, and "png1" and "hong1" for the surname 方 (in this case, my maternal Sin-Kheh family's surname). [Actually, thinking about it, there's 風 ("hong1"), 方 ("hong1"), 法 ("huat4"), 發 ("huat4"), 粉 ("hun2"), 佛 ("hut8") with initial-h, and 放 ("pang3"), 房 ("pang5"), 分 ("pun1") with initial-p, so not really an overwhelming number in favour of initial-h. Does anyone know?]

I think I've read that p- represents the oldest Sinitic layer in Hoklo. There are at least three, as repped by:

方 png (oldest)
方 hng (early Tong?)
方 hong (Classic Tong Literary)

Surnames are usually, if not always, in the oldest layer. This makes me wonder if 林 is really the punji for nâ.

[amhoanna]

Very useful 字典, based on Coanciu Hoklo:
http://solution.cs.ucla.edu/~jinbo/dzl/

has moved:
http://alt.reasoning.cs.ucla.edu/jinbo/dzl/lookup.php

[SimL]

I think I've read that p- represents the oldest Sinitic layer in Hoklo. There are at least three, as repped by:

方 png (oldest)
方 hng (early Tong?)
方 hong (Classic Tong Literary)

Surnames are usually, if not always, in the oldest layer. This makes me wonder if 林 is really the punji for nâ.

Thanks, amhoanna!

Last weekend I started making lists of Hokkien-Mandarin correspondences (a sort of Swadesh-list thing). I haven't got very far, and I'm planning on keeping it quite simple. It's going to take quite a while (a few weeks definitely). When it's ready I'll post here.

Perhaps this sort of tabulated information is easily available on Chinese (I mean character-based) sites, but AFAIK, there's very little information about it in English. Is my impression correct?