Jianiyin: pseudo-phonetic romanization system

[chalimac]

假擬標音

JIANIYIN - PSEUDOPHONETIC ROMANIZATION

http://code.google.com/p/jianiyin/

Text Sample

I made this for my personal use, then I figured I might as well publish it in case it helps someone.

This tool automatically converts pinyin to jianiyin. The purpose of jianiyin is to overcome some of the shortcomings of pinyin, specially for learners of mandarin as a second language. These shortcomings are:

- Oversimplification of vocalic groups such as -iou, -uei, and -uen that are simplified as -iu, -ui, and -un in pinyin. (Oversimplification in reference to IPA as the perfect notation but the less typable).

- Oversimplification of syllables such as bo, po, mo, and fo.

- Addressing initials such as c- that have rationale in slavic languages, but are counterintuitive for romance or germanic learners of mandarin.

Jianiyin tries to approximate IPA (International Phonetic Alphabet ) while retaining virtues of pinyin such as syllable demarcation and typebility. As all romanizations it is a compromise, a hard balance. Jianiyin is open to suggestions, improvements and discussions. It is not made for polemic purposes. It just tries to be helpful realizing that the needs of second language learners and native speakers are different. No such thing as a perfect romanization exists.

The major changes that Jianyin introduces to pinyin are:

c > ts

q > tx *

-ui > -uei

-iu > -iou

-un > -uen (except jun, yun, qun, xun)

yi (standalone /i/) > i

yan > yen

-ian > -ien

bo > buo

po > puo

mo > muo

fo > fuo

* q > tx: The rationale lies in IPA. Since x- is [ɕ] and q [tɕh], it makes sense for q- to be tx-. Both share the point of articulation behind the lower front teeth. The "t" hints at the more forceful release of air in the case of q-. Another possible solution could have been tj-

UPDATE 1 (Thanks to imron and renzhe)

-un > -uen (except jun, yun, qun, xun)

- yuen restored to yuan

UPDATE 2

- Reverted tz back to z. I realized the resulting text had too many words beginning with t.

- New tool to reverse-convert from jianiyin to pinyin.

UPDATE 3

Tonal spelling version:

- Third tone is spelled as doubled vowels.

- Second and fourth tones are represented by accents.

- First tone and neutral tone are unmarked.

Tonal spelling version sample

INSTRUCTIONS

Unrar the file and:

1. Paste Pinyin text in pinyin.txt. The text must be in 1-4 number format.

2. Double-click Pinyin_to_Jianiyin.bat

3. The program will generate a file called jianiyin.txt with the transformed text in 1-4 number format.

4. This text can be converted to tone marks using tools designed for pinyin such as Wenlin or online sites such as http://www.foolsworkshop.com/ptou/

IMPORTANT: The program will overwrite the contents of jianyin.txt everytime it is executed.

Edited May 29, 2009 at 11:36 AM by chalimac

[Hofmann]

Good job. If you have yi to i, then why not change all y to i; and change all w to u? After you do that, you can change all [y] to y.

[imron]

What happens for example with 元 vs 云, which have different pronunciations but which based on the description above both appear to be romanized as yuen?

[chalimac]

What happens for example with 元 vs 云, which have different pronunciations but which based on the description above both appear to be romanized as yuen?

Fixed. Now yun is an exception to the -un -uen rule. Thanks for discovering this.

[imron]

Doesn't that start to make things confusing then when I see for example juen, kuen but then yuen (all examples listed in the sample text your provided), as -uen now has different pronunciation depending on the initial.

[Shadowdh]

c > ts

z > tz

q > tx *

-ui > -uei

-iu > -iou

-un > -uen (except yun)

yi (standalone /i/) > i

yan > yen

yuan > yuen

-ian > -ien

bo > buo

po > puo

mo > muo

fo > fuo

Some major problems I see is confusion with pronunciation... eg your z doesnt look right, if you want to write it as its pronounced then ds would be a better transliteration. Also the bo reads as b-oo-o... likewise the others that are similar... I agree with Imron re the others.

Also I feel that by changing the way pronunciation is read/learned then there is a huge risk of making the speaker sound really really wrong... however I may have misinterpreted your efforts and as I am having a bad day this is likely...

[renzhe]

To add to imron's comment, in your system, bo po mo fo (buo puo muo and fuo) have the same finals as huo tuo ruo nuo luo, etc, when they are obviously pronounced very differently.

[chalimac]

Thanks everyone for the comments.

If you have yi to i, then why not change all y to i; and change all w to u? After you do that, you can change all [y] to y.

I didn't do this because I wanted to preserve pinyin clear syllable demarcation. If pinyin fawen turned to fauen, it is not that manifest that the syllables are fa/wen, it could be fau/en. This is important being a monosyllabic language. Besides, a text with too many vowels looks ugly,

To add to imron's comment, in your system, bo po mo fo (buo puo muo and fuo) have the same finals as huo tuo ruo nuo luo, etc, when they are obviously pronounced very differently.

I am not sure they sound so much different, but then I am no native speaker. Of course, the point of articulation of the initial will somehow affect the final, and also the tones will create variances. I must rely on IPA transcriptions. In IPA final -o is o [u̯ɔ] and final -uo is also [u̯ɔ]. At least if wikipedia is not wrong:

http://en.wikipedia.org/wiki/Pinyin

[imron]

According to zhuyinfuhao they are different.

Compare for example:

bo - ㄅㄛ

po - ㄆㄛ

mo - ㄇㄛ

fo ㄈㄛ

With

duo - ㄉㄨㄛ

tuo - ㄊㄨㄛ

nuo - ㄋㄨㄛ

luo - ㄌㄨㄛ

etc.

[chalimac]

Even with IPA as a reference it is not that easy. There are several variations of IPA applied to Chinese. This paper lists 3 possible IPA transliterations:

http://www.sino-platonic.org/complete/spp052_chinese_ipa.pdf

I am still reading through the paper, although it seems to support some of my choices like the -ian > -ien or the yan > yen.

My system is not perfectly phonetic, it just tilts the balance a little closer to purely phonetic than pinyin.

[renzhe]

I am not sure they sound so much different, but then I am no native speaker. Of course, the point of articulation of the initial will somehow affect the final, and also the tones will create variances. I must rely on IPA transcriptions. In IPA final -o is o [u̯ɔ] and final -uo is also [u̯ɔ]. At least if wikipedia is not wrong:

They definitely sound different, and the IPA transcription is just an approximation too.

[roddy]

Ignore me, I'm sleepy (thanks Imron)

[roddy]

You're right, I wasn't thinking straight.

[chalimac]

All these replies are much appreciated and I will help me to finetune the system. Addressing the -o Vs -uo ending question, this article by Shibles gives ground to my solution. He proposes based on IPA:

mo /mwɔ/

duo /dwɔ/

tuo /twɔ/

fo /fuɔ/

luo /lu.ɔ/

Therefore, I think approximating these endings to -uo has some merit.

Another question was concerning pinyin z-:

your z doesnt look right, if you want to write it as its pronounced then ds would be a better transliteration.

z- is /dz/ in most IPA transliterations and pinyin c- is /ts/. Hence my choice of tz- and ts-. inline with MPS2 and Gwoyeu Romatzyh.

[imron]

I still think the yun issue has not been satisfactorily resolved.

Presented with the following syllables:

juen

kuen

yuen

yun

It is instinctive to think that the top three are pronounced the same and yun is pronounced differently - when in actual fact yuen is the odd one out.

[renzhe]

And what about txuan? Or is that written txuen?

The -uan (actually -üan) in quan, juan, xuan and yuan is pronounced the same.

It is different from -uan in kuan, duan, luan, etc.

It is also different than -un (actually -ün) in yun, jun, xun, etc.

In your system, yun = yun and yuen = yuan, but at the same time qun = txuen, while quan = txuan.

Whereas txuan and yuen have exactly the same final, pronounced exactly the same way, as does juan, etc.

[chalimac]

I have been looking into the -un ending issue some of you have raised.

I've found that there two groups of pronunciations of -un:

kun /-uən/

chun /-uen/

dun /-uən/

gun /-uən/

hun /-uən/

zhun /-uən/

lun /-uən/

run /-uən/

shun /-uən/

sun /-uən/

cun /-uən/

tun /-uən/

zun /-uən/

jun /-yn/

yun /-yn/

qun /-yn/

xun /-yn/

These is Duden 1990:99-102's notation, but other authors based on IPA broadly agree to this division.

What do you think of introducing the following exceptions in the -un > -uen rule?

jun, yun, qun, xun

These will be left as in pinyin since the e sound is not clear at all.

About yuan it is /-ɛn/, not the same sound than the others, being and open e, [correction] maybe it should be kept as yuan.

Edited May 20, 2009 at 01:50 PM by chalimac

[renzhe]

I've found that there two groups of pronunciations of -un:

This is not really true, as the second group is actually -ün. any finals starting with "u" cannot follow q, x, j and cannot stand alone.

When writing pinyin, the umlaut is dropped because there is no danger of confusion.

-ün is pronounced differently from -un.

But if you write jun, yun, qun, xun with -un (therefore substituting -un for -ün), then why not leave the others, like kun the same?

About yuan it is /-ɛn/, not the same sound than the others, being and open e, [correction] maybe it should be kept as yuan.

The same applies here. -uan in yuan is actually -üan, and pronounced exactly the same way as in quan, xuan and juan.

And differently from kuan, duan, luan, etc.

[chalimac]

renzhe,

Very informative. As per your explanation, it seems that those syllables with ü are not followed by the ə sound.

I will try to implement these changes based on your suggestions:

- Restore yuen to yuan.

- Add jun, yun, qun, xun as exceptions in the -un > -uen.

[renzhe]

Very informative. As per your explanation, it seems that those syllables with ü are not followed by the ə sound.

To my ears, they are. Exactly the same way that -un is pronounced, only the ü is softer. The difference is in the softness of "u", not in the -n ending.

- Restore yuen to yuan.

What about quan, xuan and juan? In any case, these are all actually -üan, and the "a" is different from the "a" in duan or guan.

If you make a change for yuan, you have to make a change for all those, because the final is exactly the same.