Hangŭl for Hokkien

Discussions on the Hokkien (Minnan) language.
Locked
Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Hangŭl for Hokkien

Post by Abun »

Hey folks,

after having finished my thesis, I decided I could begin to confirm the joke that only he who invented a new transcription system is a real sinologist :lol: That is, I tried to find a way to write Hokkien in Hangŭl. As everybody who knows both Hangŭl and Hokkien will know, Hangŭl should fit Hokkien quite well in theory since the syllable structure in Korean and Hokkien is kind of similar and quite a few of the consonants and vowels sound the same or at least very close to each other. However, there are also a few difficulties, the biggest ones being the absence of symbols for tones and nasalization in Hangŭl. Therefore, I quickly had to abandon the idea that the result would be writable by using the standard Unicode-encoded Hangŭl glyphs (well some syllables of course will be, but a lot of them won’t). That means that the product of my work can’t be used on any computer and likely never will be… but I did it anyways… just for the sake of it :lol:

Disclaimer: I will not use the Revised Romanization for Korean which most people seem to prefer these days but the McCune-Reischauer one, not only because I have a certain dislike for the eu’s and eo’s in RR but also because I feel it’s closer to POJ (especially the initials), thus allowing for a better comparison.

Initials
The initials worked pretty well with adopting existing symbols. The ones that sounded either identical in Korean and Hokkien or that I found phonetically close enough to adopt are the following (in Korean alphabetical order and with their POJ counterpart):
ㄱ k-
ㄴ n-
ㄷt-
ㄹl-
ㅁm-
ㅂp-
ㅅs-
ㅇø- (zero initial)
ㅈch-
ㅊchh-
ㅋkh-
ㅌth-
ㅍph-
ㅎh-
Which is all of the modern Korean basic consonants.

With these however, we still miss the voiced initials “b-” and “g-” and I don’t quite know what to do about them. I thought about using ㅃ and ㄲ but these sound significantly different in Korean (they are “tense” consonants (Koreans call them “toensori” 된소리, literally “hard/stiff sound”), which means in reality that they too are voiceless stops but differ from the normal voiceless ones in the fact that while releasing the stop, you also release a glottal stop). In effect I guess the Korean ㅃ and ㄲ would sound just the same as “p-” and “k-” to a Hokkien native, not like “b-” and “g-”, so I don’t want to use them, but I have no better idea at the moment, especially if I want to keep true to the featural nature of the chamo (자모=字母).

Apart from “b-” and “g-”, there are yet two other Hokkien initials which are still missing from the list: “ng-” and “j-”.
Modern Korean does not have “ng-” as a syllable initial, only as a final, so they use one single chamo for the final “-ng” as well as the zero initial (ㅇ) with the distinction between the two being made by their position. In Hokkien however, both the zero initial and the velar nasal “ng-” can appear in initial position. Luckily, the two were originally distinguished in Korean writing as well, with the nasal being written as ㆁ, so I chose this chamo for “ng-”. However, I’m not too happy with that choice because it looks too similar to the zero initial (only distinguished by a little hook at the top of the nasal, only slightly longer than the one that the zero initial sports in serif fonts).
For “j-” there is no chamo which really matches. For now, I decided to use the obsolete chamo ㅿ, which sounded like a „weak s“ (possibly voiced [z]). Still, I’m not that happy with that because of the difference in sound quality. I am also thinking if I could just parallel Bopomofo and go for ㄖ, the Bopomofo simplification of 日. This could also be explained as a picture of the mouth with the tongue in the middle, so it would at least to a certain extent get close to the other chamo in the way that it also depicts the position of the vocal organs when uttering the sound.

Vowels
With the vowels I had a little more difficulty than with the initials. I think we can safely adopt the following:
ㅏa
ㅑia
ㅓo͘ (the one with a dot)
ㅕio͘. This letter combination is impossible in POJ, but the sound does exist, although only in conjunction with final consonants (especially -k or -ng) or nasalization. In these cases it is spelt without a dot in POJ
ㅔe
ㅖie, I would tend to adopt this spelling for POJ ia in front of -t or -n
ㅜu
ㅠiu
ㅟui
ㅞoe
Furthermore, I will also adopt ㅐ(open e, like the one in “bread”, today effectively the same as ㅔ in Korean) for POJ ai, because this vowel chamo is originally a combination of ㅏa + ㅣi.
Although the Korean vowel is somewhat different, I would also adopt ㅗ(closed o, like the start of the vowel in “know”) for POJ “o” (the one without a dot). However, apart from the difference in vowel quality this also produces the problem that I can’t spell POJ “oa” and “oai” as ㅘ and ㅙ like the Korean people do because I feel the difference between POJ “o” and the medial “u” in these vowel combinations is too big. So I have to resort to combinations of ㅜand ㅏ orㅐ respectively, which are not allowed in original Hangŭl and therefore not encoded in Unicode. But since I will most definitely encounter that problem sooner or later anyways, I don’t feel too bad about that.

What really gave me a headache however are the vowel combinations “au” and “iau” in POJ because there is no vowel combination even remotely similar to this in Korean, so I guess I will have to come up with something new here. My best idea so far was born from a combination versions of ㅏ and ㅜ, which in handwriting often tend to become └ and ┐ respectively, so I came up with └┐ for “au” (written at the same position as ㅏ, to the right of the initial). For “iau”, my idea would be to split these up, writing ┐below └, thus getting two parallel strokes in the middle to imitate the pattern that normal Hangŭl has for iotizing vowels.

Then there are two vowels which I have difficulty to grasp since they are not used in Mainstream Taiwanese: the Choanchiu vowels that are spelt “ir” and “er” in Tailo (don’t know about POJ). The problem is, I’m not exactly sure what they sound like. I have a vague impression of “ir” being pretty much the same as Korean ㅡ(IPA [ɯ]), in that case we should use this vowel letter. But I have no impression of “er”, so I don’t know how best to render it in a Hangŭl transcription. What makes it worse, “er” also occurs in at least one diphthong: “ere” (in fact, I have only ever seen it in this combination, does it occur on its own at all?), so I would have to keep a way of rendering that one in mind as well. My only idea so far would be to use the obsolete vowel ㆍ (the so called “sub-a”(아래 아), which was written below the initial), which is thought to have sounded like IPA [ʌ], but I don’t know how close that is to “er”…

Apart from the “real” vowels, we also have vocalic “m” and “ng” in Hokkien. My current idea for these would be to simply write the ㅁ/ㆁ next to the initial. Kind of looks like eyes, especially zero initial + ㆁ (ㅇㆁ), but what the hell :lol:

I have not yet come up with a satisfying way of noting the nasalization. For now, the only idea I had was to adopt the way it’s done in Extended Bopomofo (i.e. adding a little circle to the vowel sign), but I’m not too happy with that. Unless the solution I come up with in the end would mean too many additional strokes, I would also very much like to indicate the nasalization even with nasal initials.

Finals and Tones
As for final “-m”, “-n” and “-ng”, the case is really easy, I will just go and use ㅁ, ㄴ and ㆁ.
As for the ji̍p-siaⁿ (入聲) syllables, I guess it would make sense to express both final and tone in a single letter, since we have several letters qualifying as symbols for “-p”, “-t” and “-k” respectively anyway. I would spell these three as ㅂ, ㄷ, ㄱin 4th tone and as ㅍ, ㅌ, ㅋin 8th tone. Since syllable finals are implosive in Korean too anyways (at least in most cases including an isolated syllable), they latter three won’t be misinterpreted as another sound than the former three. For the glottal stop however, I’m still not sure. My best idea so far is takingㆆ(an obsolete letter which probably originally marked glottal stop) for 4th tone and ㅎh for 8th tone.
There is one problem about this however: If I use ㆍ for POJ “er”, it could lead to confusion between syllables ending in -ereh (ㆍㅔㆆ) and -e̍h (ㅔㅎ) because the ㆍ could be mistaken for the dot on top of ㅎ (and the other way around), so I have to find a solution there.

As for the other 6 tones, I’m not quite sure. My idea would be to think of a symbol for each and put it in the patch’im 받침 (the bottom part of a Hangŭl syllable), to the right of the final “-m”, “-n” or “-ng”, should there be any. For economic reasons, it would make sense to only assign symbols to five of the six tones and leave one unmarked. Since it doesn’t really matter which tone that is and most of the better known transcriptions leave the 1st tone unmarked, I would tend to do the same.
For the other 5, I decided on the following symbols for now:
2(陰上): Z (or ㄹ or 2 as variants). Originally this one was born from playing with the character 二 but it also resembles a 2 or a handwritten ㄹ, which is the initial for 二 in some variants
3(陰去): ㅅ, from the word sam(三), which would be 삼
5(陽平): ~, from its similarities to the POJ circumflex
6(陽上): ㅊ because it kind of looks like 六
7(陽去): ┌┐, born from the tone contour in TWese as mid-level flat tone (although I know it’s not the same in every variant but I didn’t have many ideas that I liked)

However, it doesn’t really matter which symbols are used. Since there are plenty of Hangŭl chamo which cannot appear in the patch’im as a final (most notably ㄹ, ㅅ, ㅈ, ㅊ and the doubled characters ㄲ, ㄸ, ㅃ, ㅆ as well as a whole bunch of obsolete characters) I could take one of those (although I would hesitate to use the doubled characters since it would get pretty crouded down there if there’s a finalㅁ, ㄴ or ㆁas well). However, the only reason for that would be laziness, since this use of the letter would have nothing to do with its normal pronunciation. So I could just as well try and come up with something new.

Summary
So for now, I use the following letters (I will mark the ones that I’m particularly unhappy with in red):

Initials:
ㄱ k-
ㄲ g-
ㄴ n-
ㄷt-
ㄹl-
ㅁm-
ㅂp-
ㅃ b-
ㅅs-
ㅇø- (zero initial)
ㆁng-
ㅈch-
ㅊchh-
ㅿj-
ㅋkh-
ㅌth-
ㅍph-
ㅎh-

Vowels:
ㅏa
└┐au
└ over ┐ iau
ㅐai
ㅑia
ㅓo͘
ㅕio- before final -p, -t, -k, -m, -n, -ng or nasalization
ㅔe
ㅖia before final -t or -n
ㅗo
ㅛio
ㅜu
ㅜㅏoa
ㅜㅐoai
ㅞoe
ㅟui
ㅠiu
ㅡir
ㅣi
ㆍer
ㆍㅔere
ㅁm
ㆁng
Nasalisation will be expressed with a small circle at the top of the vowel chamo

Finals:
ㅁ-m
ㄴ-n
ㆁ-ng
ㅂ-p (4th Tone) ㅍ-p (8th Tone)
ㄷ-t (4th Tone) ㅌ-t (8th Tone)
ㄱ-k (4th Tone) ㅋ-k (8th Tone)
ㆆ-h (4th Tone) ㅎ-h (8th Tone)

Tones (non-ji̍p-siaⁿ):
1: unmarked
2: Z
3: ㅅ
5: ~
6: ㅊ
7: ┌┐

To give you an idea what that looks like, I wrote a few sentences of a random text (the beginning of Christmas story in 紅皮聖經: http://210.240.194.97/memory/tgb/thak.a ... 1&page=285) and wanted to upload an image, but for some reason the forum won't let me... I can mail it if you're interested :lol:

To Do
1. Sort out the things that I’m unhappy with.
2. I would also very much like to come up with a way of indicating whether a syllable is read in standing, running, neutral and possibly even “9th” tone (the one that the first syllable in tripled words such as âng-âng-âng (紅紅紅) takes). Since the highest percentage of syllables are in running tone, I would tend to leave that unmarked, but for the other 2-3 I would like to have a way of indication. This however would have to happen outside of the Hangŭl syllable so it can be applied to words written in characters as well. I’m thinking of using punctuation, but I somehow feel it would be too much of a break if I used for example a comma to indicate standing tone (especially if I think of words like āu--ji̍t(後日)).


I'm happy to hear your comments :mrgreen:

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

aBun,

Glad U've decided to tackle the age-old Hologul issue. Here is a summary of what others have done:
http://icanimoxen.pixnet.net/blog/post/29112660

I looked into the matter as well, once upon a time. The system I came up with was almost identical to Oddy's, which is reviewed in the blog article.

Initials b- vs m-, l- vs n-, and g- vs ng-, which are two series in classic POJ (inc. 台羅), should actually be one series. The nasal series only occurs in front of a nasalized vowel with no final nasal stop; the non-nasal series only occurs elsewhere. Mûi is really just bûiⁿ, ngē is just gēⁿ, and so on. The use of initial m-, n- and ng- in Banlamese is actually kind of "unscientific".

Using just b-, l-, and g-, but with the vowel consistently labelled for nasality, is the most "pure" solution in POJ -- or any phonetic writing system for the language.

B- vs m-, etc., might be in contrast, though, in Teochew.
I have a vague impression of “ir” being pretty much the same as Korean ㅡ(IPA [ɯ]), in that case we should use this vowel letter. But I have no impression of “er”, so I don’t know how best to render it in a Hangŭl transcription.
Yes on the first. "-er" sounds schwa-like, but a little higher and/or tenser, not relaxed like the "-o" in southern Taiwan or Amoy.

I like what U were trying to do here. I dig most of your underlying philosophy.

The elegant solutions sure are elusive as hell. I tend to feel that a hiragana-like system would be ideal, but it doesn't need to be based on actual Japanese hiragana.

_______

Back to the topic of tones, though. It's possible to mark out the Hoklo tone system using just three symbols. We need one symbol to mark 上, and another for 去. Then we need a symbol to mark 阳. Everything else falls into place. 入 doesn't need a marker b/c there's always a -ptkh ending there.

Imagine how elegant POJ could be using such a scheme. For example, one accent could be used for 上, and the other for 去. The underdot (as in VNese) could be used to mark 阳 -- fitting, b/c those are the low tones in Hoklo, going by the running contours; and in Sinitic languages in general.

Goá ánlenn káisoeh ú ̣ chengchó--bọ? Goá ánlenn siá lín tạkgẹ é ̣ bẹngpẹk--bé ̣?

(It's not displaying correctly on my Mac when I put an underdot and an accent on one letter. It worked well on my last machine, with Windows.)

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

Hey amhoanna,

first thanks for the feedback, even more so since it's encouraging :mrgreen:
amhoanna wrote:Glad U've decided to tackle the age-old Hologul issue. Here is a summary of what others have done:
http://icanimoxen.pixnet.net/blog/post/29112660
Interesting, and not much different from mine :mrgreen: Of the rest here is only one thing I might think about adopting though: using ㅏㄹ for POJ au.
I don't quite like the way they dealt with the rest of the problems. I feel we should stay as close as possible to Hangŭl sound values as possible. Therefore, if I wanted to use ㆆ as an initial, I would use it for ø- (glottal stop) and not for ng-. ㅏㅜ and ㅑㅜ look weird in my eyes, and not only because they don't exist in original Hangŭl, but also because Hangŭl rules say we should read from left to right if there's an upright vowel, but with ㅏㅜ and ㅑㅜ, I feel like I have to read in sort of a zigzag course. And while I agree that my way of writing vocalic m and ng is far from elegant, I still think it has advantages over the ones presented there. albyu and oddy use ㅡ as a vowel placeholder which I can't do because I already use ㅡ for ir. The "new" system wants to use free ㅁ/ㅇ, which I can only see working if there is no initial. I can't see how he would transcribe tńg or pn̄g. As for the tone indicators, I would be interested to see how they use them. albyu seems to write them to the left or top left of the syllable if I get it right, which would work great in vertical writing but in horizontal writing it would annoy me. With the other two, I don't quite understand, how they would employ their symbols if they do. (Oh and btw, I would really like to know what made them exchange the sound values for ㅗ and ㅓ).
The latter two systems also advocate normally leaving out the tone mark. This of course makes it possible for oddy's system (and only his because ㅏㅜ is not encoded) to be written on a computer, save for the nasalization marker. I guess it comes down to philosophy there. I am willing to sacrifice that aspect in favour of a more accurate sound system. Some people might also argue that tones are not very intuitive to write, which may be true. But I think, if people can rememer a Hàn-jī to spell a certain word with, they can also remember the tonal letter in my system.

EDIT: My bad, ᅷ (not ㅑㅜ however) actually did exist in original Hangŭl and so did ᅶ and ᅸ, but they were used only for transcribing Chinese pronunciations. Although my feelings about zigzag reading still remain, this makes me think again about using some of these (either ᅶ and ᅸ because they are both attested, or ᅷ and ㅑㅜ which would be more consistent using ㅜ instead of ㅗ, but on the other hand only one is actually attested). Although I actually like └┐ better aesthetically :lol:
amhoanna wrote:Initials b- vs m-, l- vs n-, and g- vs ng-, which are two series in classic POJ (inc. 台羅), should actually be one series. The nasal series only occurs in front of a nasalized vowel with no final nasal stop; the non-nasal series only occurs elsewhere. Mûi is really just bûiⁿ, ngē is just gēⁿ, and so on. The use of initial m-, n- and ng- in Banlamese is actually kind of "unscientific".
That's true of course, and I'm still thinking of simply spelling b- and g- as ㅁ and ㆁ + Nasalization. But my goal was not to create a phonetic script adhering to the "one phoneme one letter" rule but one which captures the sound differences and m- does sound different from b-, even disregarding the nasalization. Anyway, using ㅁ and ㆁ is something I do think about, but have so far restrained from doing not because I asserted phonetic differences but because I feel the realization is too different.
amhoanna wrote:
I have a vague impression of “ir” being pretty much the same as Korean ㅡ(IPA [ɯ]), in that case we should use this vowel letter. But I have no impression of “er”, so I don’t know how best to render it in a Hangŭl transcription.
Yes on the first. "-er" sounds schwa-like, but a little higher and/or tenser, not relaxed like the "-o" in southern Taiwan or Amoy.
Hm it's hard to grasp sounds via an explanation but what you describe seems very different from [ʌ]... The problem is, I haven't yet come up with an idea for another vowel symbol for "er", yet. Have to think about it.
amhoanna wrote:The elegant solutions sure are elusive as hell. I tend to feel that a hiragana-like system would be ideal, but it doesn't need to be based on actual Japanese hiragana.
I'm not quite sure what you mean. If you mean a syllabary would be ideal, my first reaction would be to disagree. In my eyes, real syllabaries only work for languages with a very very limited amount of possible syllables, otherwise the amount of characters would be very huge. That alone would be not too big a problem as we know from Chinese, but the more syllables there are, the shorter the morphemes tend to get until we almost have "one syllable one morpheme", and in that case a syllabary would be no different from a morphemic script as modern chinese languages are. Even Japanese which has a syllable inventory of at most a few hundred possible syllables doesn't get by with the syllabary and frequently has to violate the "one character one mora" rule (see しゃ and so on). That being so, I feel that the syllabic inventory of Hokkien is too big for a syllabary to be practical, which is precisely the reason why I see Hangŭl as the best way to be able to stick to the "one syllable one character" rule while at the same time getting by with a limited amount of characters by building the syllable blocks from smaller entities.
amhoanna wrote:Back to the topic of tones, though. It's possible to mark out the Hoklo tone system using just three symbols. We need one symbol to mark 上, and another for 去. Then we need a symbol to mark 阳. Everything else falls into place. 入 doesn't need a marker b/c there's always a -ptkh ending there.
This is a very interesting idea, I hadn't thought about that. I have to think about how I could implement it, though, because it would involve the use of two symbols in the case of 陽上 and 陽去, which in my system would make it possible to have three symbols in the 받침, which I don't think is ideal... But maybe there are workarounds, for example ligatures which combine the symbol for 陽 and those for 上 and 去 into a single symbol respectively... I'll give it some thought.
Last edited by Abun on Fri Aug 09, 2013 4:03 pm, edited 1 time in total.

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

U are probably right, a lot of drawbacks may be inherent in a "Hoklogana".

I feel that Japanese hiragana, at least, is much prettier and much more kanji-compatible (aesthetically) than hangul (or katakana). Circles, esp., are almost alien to kanji. I say almost, b/c there is "〇". But I don't agree with the people who say that hangul is hideous. I think it's all right. Frankly, I would be glad to see either a Hoklogana or a Hoklogul.

Oddy and I communicate exclusively in a mixed script of kanji and hangul. He is able to type his nasal mark at will. The lack of tone marking still kills it for me.

The website I linked to also has an article adapting the Bopomofo for use in a hybrid script. The blogger is well-versed in calligraphy and was able to remove the ugliness of Bopomofo. I recommend that article too. A Bopomofo system would meet with much less "political opposition" from the "public". For that I think it's worth considering as well.
my goal was not to create a phonetic script adhering to the "one phoneme one letter" rule but one which captures the sound differences and m- does sound different from b-, even disregarding the nasalization.
This is a slippery slope leading down to biases and habits that each person has acquired, mostly from other languages and scripts. For most Taiwanese speakers, "ts-" is palatalized in front of "-i-". Does the script need to reflect this as well? It does "sound different" from non-palatalized "ts-"...

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

Something interesting I just came across in Wikipedia:
Korean has a simple tone system often characterized by the poorly defined term "pitch accent". Hangul originally had two diacritics to represent this system, a single tick, as in 성〮, for high tone, and a double tick, as in 성〯, for a long vowel. When transcribing Chinese, these had been used for the 'going' (去聲) and 'rising' (上聲) tones, respectively. Although the pitch and length distinctions are still made in speech by many Koreans, the diacritics are obsolete.

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

amhoanna wrote:
my goal was not to create a phonetic script adhering to the "one phoneme one letter" rule but one which captures the sound differences and m- does sound different from b-, even disregarding the nasalization.
This is a slippery slope leading down to biases and habits that each person has acquired, mostly from other languages and scripts. For most Taiwanese speakers, "ts-" is palatalized in front of "-i-". Does the script need to reflect this as well? It does "sound different" from non-palatalized "ts-"...
You're right, I'm being inconsistent there... Although I admit I thought about distinguishing alveolar and palatal "ts-", "tsh-" and "s-", and only refrained from doing so because ㅅ can be both alveolar or palatal in Korean as well, and while I percieve ㅈ and ㅊ as always palatal (although I can say nothing about non-standard Korean), I view them as forming a series with ㅅ. Anyways, I'll think about it again. What kind of makes me refrain though, is that I would have to combine l- and n- as well, and I still can't bring myself to read ㄴ as l- or ㄹ as n- :roll: but that's probably just conservatism.
amhoanna wrote:
Korean has a simple tone system often characterized by the poorly defined term "pitch accent". Hangul originally had two diacritics to represent this system, a single tick, as in 성〮, for high tone, and a double tick, as in 성〯, for a long vowel. When transcribing Chinese, these had been used for the 'going' (去聲) and 'rising' (上聲) tones, respectively. Although the pitch and length distinctions are still made in speech by many Koreans, the diacritics are obsolete.
Yes, I know about the dots indicating pitch accent and considered using them. Ultimately I decided against it because it could only distinguish three different tones (that is no dot, one and two dots) and I wasn't aware of the possibility to distinguish Hokkien tones with only three symbols at that time yet. Now it seems more possible to me to use those in addition to a single 받침-symbol for 陽 tones (or, for that matter, 陰. Or is there another reason to mark 陽 instead of 陰 than because two of the four 陰 tones go unmarked in POJ?). In this case however, I would also add the possibility to write the dots above the syllable when writing horizontally.
(Oh and btw, I would be very interested to know how the dotted syllables were entered on computer)

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

In my mind, the crux of it is that the 阴 series are the default. This is reflected in the higher frequency of 阴 syllables in any dialog or text. The 阴 series also "comes first".

It's always good to question the status quo, though.

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

amhoanna wrote:In my mind, the crux of it is that the 阴 series are the default. This is reflected in the higher frequency of 阴 syllables in any dialog or text. The 阴 series also "comes first".
As far as I'm informed, the 陰陽 distinction has its origins in different onsets in Early Middle Chinese with 陰 syllables thought to have originally had voiceless onsets while 陽 syllables had a voiced one (with some variation throughout the variants of course, the sonorant-obstruent distinction sometimes making a difference as well). Thus, a higher frequency of 陰 syllables would mean that there might have been more syllables with voiceless initials in Early Middle Chinese than voiced ones, but in my eyes that doesn't equal "default". I would also speculate that part of the reason why 陰 tones come before 陽 tones in the ordering systems is because you say 陰陽 and not 陽陰 (and the reason for 陰 coming before 陽 here are simply euphonic rules. I know tones had probably not developed at the classical times, but apparently even then people favoured syllables that were to become 平聲 to come before those later becoming 仄聲 instead of the other way around).
Anyways, I don't believe either the higher frequency or the ordering to mean that 陰 tones are to be considered default. However, a higher frequency of 陰 in itself would be a reason to mark 陽 instead of 陰.

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

I agree with your main line of reasoning. If 阳 syllables occurred with greater frequency, say at a 5:4 ratio, I would favor marking 阴 instead.

When I said that the 阴 series is the default, I meant that it has been subjectively considered the default -- by everyone's who's gotten into this branch of learning -- historically, a subset of "Chinese" scholars.

Yes, the structure of the phrase 阴阳 may be related. The order of the phrase itself is actually pretty amazing. The phrases "capo͘ cabó͘ ", "lâmlí‧", "hiaⁿtī címōe", "ang-kah-bó͘", "huche", "pěebú" etc. are pretty well fixed in Hoklo and Lit. Chinese. Only in 阴阳 (and 雌雄 -- never used in Hoklo speech AFAIK) does the lady get to go first.

The concept of 阴 being the default, and 阳 being "like 阴 with cheese", is borne out in nature itself. Femmes have two X chromosomes. Throw in a Y chromosome in place of one of the Xs, and a dude is born.

Emptiness and passiveness are 阴. Throw something substantial in the mix, like "voicing", and U get 阳.

Now U're probably thinking, "Kàn-, such typical straight-guy things to say. So fu*cking unoriginal."

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

amhoanna wrote:Yes, the structure of the phrase 阴阳 may be related. The order of the phrase itself is actually pretty amazing. The phrases "capo͘ cabó͘ ", "lâmlí‧", "hiaⁿtī címōe", "ang-kah-bó͘", "huche", "pěebú" etc. are pretty well fixed in Hoklo and Lit. Chinese. Only in 阴阳 (and 雌雄 -- never used in Hoklo speech AFAIK) does the lady get to go first.
When I started learning Classical Chinese, one of the first things we learned was that in coordinations like this (i.e. AB=A and B), the order is often determined by a few rules:
1. 陰 before 陽
2. 平 before 仄
3. 上/去 before 入
Of course, at the time in question the tones did not exist yet, it was the things that would cause them to develop later which mattered (so Confucius wouldn't have thought that 陰 before 陽 sounds good, but that voiceless onset before voiced onset sounds good, or something like that), but since we have only vague ideas about how Ancient Chinese sounded, we help ourselves with Middle Chinese tones.
With these rules, most of these coordinations can be explained, including 陰陽 and 雌雄 (in both cases rule 1 applies). What's interesting are the instances where they are violated, such as 男女 (陽平 before 陰上) and 父母 (陽上 before 陰上) which according to the rules should have been 女男 and 母父. A lot of them can be explained with social order. 男 was supposed to be superior to 女 and 父 superior to 母, so they came first despite the rule. A similar example would be 老莊 (陽上 before 陰平). Cha-po͘-cha-bó͘ being a combination of two bisyllabic words wouldn't fall under the euphonic rules, the order is probably thus for the same reason as 男女, the same goes for ang-kah-bó͘ (although the rules here don't apply because of an inserted kah and not because the words are bisyllabic, and even if they did, they would give the same result). In the cases of 兄弟 and 姊妹 the social order is so nice as to coincide with the result the rules would have given anyways (btw, shouldn't 弟 be T6?). Which leaves us with hu-che, where none of the rules can be applied anyway because they're both 陰平.
So one could speculate now, why the preference for the male before the female was powerful enough to create 男女 and 父母, but not 陰陽 and 雌雄? My guess would be that it was because they were not percieved as human. 雌 and 雄 are only used for animals today (though I don't know about former times. After all, 雄 has other meanings definitely related to human males) and 陰 and 陽 were to be equal forces, none was supposed to be better than the other. This is just my theory of course.
amhoanna wrote:The concept of 阴 being the default, and 阳 being "like 阴 with cheese", is borne out in nature itself. Femmes have two X chromosomes. Throw in a Y chromosome in place of one of the Xs, and a dude is born.
[...] Now U're probably thinking, "Kàn-, such typical straight-guy things to say. So fu*cking unoriginal."
Haha bōe--lah, although I have heard the opposite statement as well: "dudes are just gals whith a quarter chromosome missing" :lol: Both equally groundless from the biological point of view I guess, but I'm no expert on that :mrgreen:

Anyway, my point was that if 陰 indeed is more frequent, it would be an argument for marking 陽. Does that statement of yours come from your own observations or do you have a source for it?

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

Re: the frequencies -- yes, by personal observation, and also from the informal observations of others. U can take any text or dialog and count the number of syllables with tones in each series. I guarantee U U'll find that the 阴 series carries a lot more syllables. This is true across languages, even into Vietnamese.
So one could speculate now, why the preference for the male before the female was powerful enough to create 男女 and 父母, but not 陰陽 and 雌雄? My guess would be that it was because they were not percieved as human.
I second this guess of yours.

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

I tried counting 陰 and 陽 syllables in my introduction text (http://www.chineselanguage.org/forums/v ... =6&t=58302 first post) for about the first third (up until 聯絡) and the result was rather close, 107/105. Of course, the result is probably slightly corrupted by the fact that I'm not a native speaker, and I might also have miscounted. I have to do a bit more testing there.

I also am testing a few other ideas for changes in the writing system:
First, I merged b-/m- into ㅁ, g-/ng- into ㆁ and l-/n- into ㄴ, thus making the system phonologically more consistent. However, I still have difficulties to convince myself to read those chamo as plosives (for some reason much more in the cases of ㆁ and ㄴ than with ㅁ), but that may just be conservatism.
I also tried a new system of indicating the tones using the idea I got from amhoanna (the one which uses only three indicators, one for 陽 register and one for 上 and 去 tones each). Since the dots next to the syllable were used to indicate 上 (one dot) and 去 (two dots) in Korean too, when they transcribed Chinese, I adopted this, leaving only the problem of indicating 陽 register open. Luckily, the merging of voiced plosives and nasals (see above) provided me with an idea, since it did away with the only two doubled consonants I was using. So my current idea would be to indicate 陽 register by doubling the initial (i.e. ㄲ instead of ㄱ, ᄔ instead of ㄴ and so on).
So:
tan tán tàn tat tân tãn tān ta̍t becomes
단 ·단 :단 닫 딴 ·딴 :딴 딷
Although this method does a way with the double character 받침 I had before, most of which are not included in any font, it requires a lot of doubled initial consonants which equally don't exist in fonts (i.e. 쌍ㆁ (not to be confused with 쌍이응, which exists in a few old Hangŭl fonts), 쌍ㅁ, 쌍ㅿ and the doubled aspiratae).

But apart from feedback, I would also ask for a bit of information about pronunciations across the variants:
  • 1. I still am not sure how to render Choanchiu -er. Is this a glide anyway, which always leads to an -e sound or is it a monotone vowel?
    2. I'm still having some questions as to variants, though. As far as I'm informed, there are four kinds of -e/-ue syllables:
    • a. Type 尾: Choanciu (Cn): -ere (or just -er?); Chiangchiu (C) and Tailam (Tl): -oe; Emng (E) and Taipak (Tp): -e
      b. Type 買: Cn, E and Tp: -oe, C and Tl: -e
      c. Type 馬: all -e
      d. Type 杯: all -oe
      (of course there are differences within the respective cities, especially now that mobility is much higher than it used to be, but you know what I mean)
    Anyway, a) can be transcribed as -ere (or -er) and everybody can read the way he speaks, and c) and d) don't make a problem anyway, but I'd like to find a way to transcribe b) in a distinct way but I don't know how... Does anybody know of a "middle way" between the two pronunciations as with Cn -er(e) with type a)? There probably used to be in former times...

amhoanna
Posts: 912
Joined: Sat Sep 18, 2010 12:43 pm

Re: Hangŭl for Hokkien

Post by amhoanna »

1. I still am not sure how to render Choanchiu -er. Is this a glide anyway, which always leads to an -e sound or is it a monotone vowel?
It can stand alone. There are some dialects in the hill country where it can be a glide. The sound is very similar to a schwa.
2. I'm still having some questions as to variants, though. As far as I'm informed, there are four kinds of -e/-ue syllables:
a. Type 尾: Choanciu (Cn): -ere (or just -er?); Chiangchiu (C) and Tailam (Tl): -oe; Emng (E) and Taipak (Tp): -e
b. Type 買: Cn, E and Tp: -oe, C and Tl: -e
c. Type 馬: all -e
d. Type 杯: all -oe
(of course there are differences within the respective cities, especially now that mobility is much higher than it used to be, but you know what I mean)
It's cool that your project is pan-dialectal.

I'm vaguely aware that the "four-way split" U just mentioned is actually a simplification. The reality of the hill dialects of Coanciu is more complex than almost anybody realizes.

As for counting syllables, U might want to try it with authentic material. Your self-intro was a gallant attempt at using the language -- as I said, worthy even of someone with Hoklophone heritage -- but it's not "representative" of the Hoklo language by a long shot. :P

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

amhoanna wrote:
1. I still am not sure how to render Choanchiu -er. Is this a glide anyway, which always leads to an -e sound or is it a monotone vowel?
It can stand alone. There are some dialects in the hill country where it can be a glide. The sound is very similar to a schwa.
Ok, in that case, I guess I think it would be sensible to use just plain arae-a for that rhyme instead of ㆍㅔ, which would be yet another vowel which is not encoded.
amhoanna wrote:I'm vaguely aware that the "four-way split" U just mentioned is actually a simplification. The reality of the hill dialects of Coanciu is more complex than almost anybody realizes.
Yes, of course the split is not that clean. For example, I have the feeling a lot of Taiwanese people, while usually having a tendency into the direction of either Chiangchiu or Choanchiu, still mix these rather freely. Especially 欲, 袂, 未 seem to be more popular in Choanchiu (i.e. beh, buē, bē), even with people who usually rather tend to have a Chiangchiu-like accent.
If there are more splits in smaller dialects, that makes the situation more difficult. I doubt that it's possible to take every single one of them into account though... Right now, I'd tend to make it my priority to render Chiangchiu and Choanchiu, which would enable one to write most of the bigger variants at least. Which is the reason why I wondered if there's an "intermediate" form for the 買 group which I could use to distinguish it from the other ones
amhoanna wrote:As for counting syllables, U might want to try it with authentic material. Your self-intro was a gallant attempt at using the language -- as I said, worthy even of someone with Hoklophone heritage -- but it's not "representative" of the Hoklo language by a long shot. :P
That is very true, I just took that one because I didn't have that many longer texts available (the only other one I could think of was the 紅皮聖典 and that's not ideal either because of the high frequency of non-Hokkien names). I was going to run a few more tests (in fact, that's why I haven't answered until now), but I don't find the time (and the nerves) for this rather dull work :lol: and I now doubt it's worth the time, too. I'll just mark 陽 for now and should it ever dawn on me that 陰 would be more economic, I can still change it.

Abun
Posts: 115
Joined: Fri Jun 21, 2013 4:15 pm

Re: Hangŭl for Hokkien

Post by Abun »

Small update:
I'm still unhappy with using both ㅇ and ㆁ for different phonemes (zero initial and [ŋ]) because I think they look too similar at least on the computer (in handwriting one could of course intentionally make the hook on top of ㆁ more prominent). However, I came to realize that if I consider it possible to distinguish them in initial position, the same should be true in final position. I therefor decided to use ㅇ as nazalization indicator, disposing of the little circles on the vowel chamo which I didn't like at all because I found it difficult to come up with a consistent rule where exactly to add them and still have all possible outcomes look acceptable.
So with this change, I would for example spell:
姓 as :싱 (or :셍 in Chiangchiu), as opposed to 四 :시 or 細 :세, or
行 as 꺙, as opposed to 奇 꺄

Locked