Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> "We spent three years developing Pinyin. People made fun of > us, joking that it had taken us a long time to deal with just > 26 letters"

I actually think it is quite fast. Grasping an entire languages phonology is a huge accomplishment and condensing it to its bare minimum is also impressive.

With that said, I must admit I don't know how pinyin relates to other phonetic systems such as the Bopomofo[0] or Wade-Giles[1] and for all I know it might just be the exact same system with different letters.

[0] https://en.wikipedia.org/wiki/Bopomofo [1] https://en.wikipedia.org/wiki/Wade–Giles



Pinyin is a little more compact (I.e. fewer Roman characters per Chinese characters) than Wades-Giles, so more efficient storage. That makes it easier to input chinese characters when using an English keyboard (or keypad on a mobile phone).


It's also arguably more aligned with the English pronunciation of the letters.


Not a all. As a phonetic enthusiast I can assure you Pinyin is by far the worst representing Chinese sounds with English letters. It is made to represent Chinese phonetic, but its creators are seemingly completely ignorant toward basic phonetic rules, such as voicing and aspiration, and how words actually sound in English.

As an example, b in English is called an unaspirated voiced consonant, which means you vibrate your vocal folds, but send only a little breath when you sound it. You can compare it with p, its aspirated voiceless counterpart, for which you do not vibrate the vocal folds (voiceless), but sends a strong breath through your month (aspirated). Chinese does not have most voiced consonants present in English, so Pinyin naively uses b (also d and some others) to represent voiceless aspirated consonants, which does not sound like a b, but more like p in “spade”. This results in most of the foreigners learning Chinese with Pinyin bringing English pronunciation when they speak, and makes them sound way too stiff and thus awkward.

Wade-Giles, in comparison, is much more systematic, and does a much better job hinting speakers with European (especially Germanic language) background to sound more correctly. Bopomofo does not have this problem, as it basically invents a new set of scripts. Pinyin probably lets you pick up casual speaking Chinese most quickly, but beyond that, it’s a curse.

[Edit]: Voice “folds” not cords, sorry. Also fixed some typos.


Pinyin is very well-adapted for (Mandarin) Chinese. It isn't specifically designed for English learners, but if you understand phonetics and learn the sounds first, that shouldn't matter. The alphabet is originally from Latin and is used in a variety of ways in different languages, e.g. Turkish c sounds like English j, German j sounds like English y, so the foreign accent problem exists generally.

There are two bilabial stop consonants, called p and b, in many different languages, with different sounds, e.g.

                          Mandarin English French
    unaspirated voiced       -        b       b
    aspirated voiced         -        -       -
    unaspirated unvoiced     b        -       p
    aspirated unvoiced       p        p       -
though there is, as you point out, an unaspirated unvoiced allophone of p in "spade". Other stops (t, d, k, g) follow a similar pattern.

English speakers using English p and b in Mandarin will still be understood even if their b sounds foreign, but French speakers might not be, because French p sounds like a Mandarin b.

The only change I'd like is replacing Pinyin -ong with -ung.


I think it would also help to expand -iu to -iou and -ui to -uei, and of course u->ü where applicable. IMEs could still accept both forms.

Pinyin is pretty well-adapted for entering Chinese, but I think that goal is at odds with being easy to pronounce. I'm not sure that having one ISO standard for both purposes is a good idea, although if China ever starts to export more culture, maybe everyone will eventually learn how to pronounce Hanyu Pinyin.


Choosing aspirated unvoiced consonants as "typical" of English is very weird. Native English speakers do not consciously distinguish aspiration but they do distinguish voicedness so it seems it would be more accurate to call the aspirated form the allophone (aspiration only occurs at the beginning of words and stressed syllables, and never after "s" as you point out, so the unaspirated form probably occurs more frequently too).

Still, I agree with your main point. The Latin alphabet is not used like the IPA by most languages, even English has not preserved all the original Latin sounds as used by the Romans. And there's even precedent for some of the sound choices made in Pinyin that would seem completely alien to English speakers. For example "c" is used in all the Slavic Latin alphabets (e.g. Polish, Czech, etc) for the unaspirated version of the Pinyin "c" sound. German and Pinyin use the letter "z" for the same sound, etc.


And maybe replace yu with yü, etc., for consistency.


When I was taught Pinyin, the instructor explained that the phonetics were influenced by the Russian pronunciation of the roman alphabet. Not sure if this is true, or if it explains the issues you're describing. But it seemed plausible that China thought they were going to be closer to the USSR given the shared ideology in the 50's.

In any case, as a person who has a decent pronunciation in Chinese but lacking the ability to fully read/write, Pinyin has been absolutely great for being able to type Chinese informally. I can communicate almost anything I can say, and that's truly wonderful. Part of that is the smart prediction software. But Pinyin has a big part in it too: it's not difficult at all to sound out a word and figure out how to spell it in Pinyin. That's the essential feature that Chinese characters lack, and Pinyin nails that essential quality as far as I can tell.


I know very little Russian and can’t say for certain, but I think Russian is similar to English in these consonants.

I should make it clear I don’t think Pinyin is a bad system, only that it does not map sounds to European languages well. And that is fine, as least for its original purpose. Contrary to common belief, Pinyin was not developed to better educate illiterate people (this does not even make sense if you think carefully), but as the next step after letter simplification (what we know as Simplified Chinese today) towards Chinese Romanisation. The goal was not to represent sounds of Chinese characters, but to outright replace them. The actual reason behind using (for example) b and p instead of p and p' is exactly this—they are optimising for ease to write, not sound accuracy.

And in regard to making it easy for people not prolific with Chinese letters to read, be them foreigners or simply illiterate, a phonetic transcription system is indeed important. But although having a standard phonetic transcription is key to the jump in literacy in China, I would argue the same can be achieved with any decent system, and a few of them are already around by the time they started working on Pinyin.

All in all, while Pinyin is good for its original purpose, it achieved its current status not because of academical reasons, but political ones. The Chinese Communists government had always have a habit ignoring existing systems and inventing their own. :p


Definitely not. Russian has no aspirated consonants at all, nor would they pronounce them in the Latin alphabet. In fact a Russian would pronounce the roman "T" exactly the same way a Mandarin speaker would pronounce the pinyin "D" so I don't think there's much credence to your instructor's story.


There was Russian influence, but maybe not for all the letters. Pinyin 'q' is based on Russian 'ч' if I recall correctly.


I don't think pinyin was even meant to represent voicing and aspiration in a way that is similar to English or otehr Germanic languages. There are many other languages using Latin alphabet which are a little bit different with both consonants and vowels, and that doesn't stop from sharing the letters.

Pinyin does a good job in providing a fairly consistent way to present Chinese sounds using Latin alphabet. It's also the workable way to input Chinese characters on a computer keyboard. It works, it is consistent, and it is now ubiquitous in China.


Well, it's not that simple though. You have 2x2 possibilities:

1. unaspirated unvoiced (av)

2. unaspirated voiced (aV)

3. aspirated unvoiced (Av)

4. aspirated voiced (AV).

German and I think English have 2/aV (b, d, g) and 3/Av (p, t, k), the diagonal so to speak (for the beginning of a syllable).

Mandarin has 1/av and 3/Av. So, 3/Av is pronounced the same - and Pinyin renders those as (p, t, k), as in German/English, so that's actually nice. Wade Giles renders them as (p', t', k'), indicating the aspiration.

German and English do not have 1/av (at the beginning of a syllable). Pinyin then chose to render them as (b, d, g), correctly indicating the unaspirated character, but tempting German/English etc. speakers to voice them (wrongly). Wade-Giles render them as (p, t, k), tempting German/English speakers to aspirate them (wrongly).

No easy solution. Plus, the apostrophes are so often ignored (e.g. Pinyin Taibei = Wade-Giles T'aipei).


Technically "p" in English does not default to the aspirated (pot) or unaspirated variety (hop). English speakers don't distinguish different word meanings when just those sounds by themselves are different, so they're considered homophones. Perhaps you knew that already and I just misread your comment, but thought it may be helpful info if not.


"Hop" comes out strongly aspirated for me. Certainly "mop" and "mob" don't sound the same.


You seem to be confused about the difference between "aspirated" and "voiced". Both these words are unaspirated for native English speakers.


Not for initials (consonants):

I find Wade-Giles ch'in more evocative than the Pinyin qin, and similarly hsüan vs xuan, and even tzŭ vs zi and tzʻŭ vs ci. Wade-Giles you basically have to learn not to ignore the apostrophe, Pinyin you have to learn most of the initials. Though the Wade-Giles jih vs Pinyin ri is a bit of a toss-up.

For finals, I find Wade-Giles more consistent than Pinyin yu, but Wade-Giles to less descriptive than Pinyin duo. But then, it's shorter.

So, yeah, "arguably" :-)


You're right about Bopomofo - both it and pinyin are based on the same pronunciations so there's actually a one-to-one correspondence between the two systems:

https://en.wikipedia.org/wiki/Bopomofo#Pinyin


As there are only about 400 syllables (ignoring tones) in standard Mandarin, it's a fairly quick exercise to make a complete one-to-one mapping :-)


Oh, and here we go: All 400 syllables of standard Mandarin in Bopomofo/Zhuyin Fuhao, Wade-Giles, and Hanyu Pinyin:

http://pinyin.info/romanization/wadegiles/basic.html


>I actually think it is quite fast. Grasping an entire languages phonology is a huge accomplishment and condensing it to its bare minimum is also impressive.

Absolutely. But I don't see how this statement could apply to pinyin. The work done on German's last spelling reform was impressive in this regard, but faulty pinyin was not.


Care to expand on the faults of pinyin, and how they are more erroneous than say the German spelling reform?

My experience with pinyin is that it is very robust, and I don't remember ever encountering pinyin that was ambiguous about how it is read (unlike written English or romanized Japanese). I have also never heard of standard Chinese syllables that are lacking a clear mapping onto pinyin.

To me that is very impressive.


Compare "xi an" with "xian". One is 2 words and one is 1 word. But for some reason Chinese people don't like to put spaces in between words so "xi an shi" is often written as "xi'anshi" which is different from "xianshi".


Mandarin speakers are seem to be fairly sloppy with their own phonetic alphabets - you're supposed to have spaces in bopomofo and apostrophes in pinyin to handle cases like these, but they often don't bother.

It becomes clearer with tones I guess, "Xi1 An1" vs "Xian1". But again, proper pinyin with tones seems to be a foreigner thing... I don't recall seeing tone marks or numbers on street signs.


But that's what you have the apostrophe for in Pinyin. Works fine.


But pinyin isn't Chinese so your point just doesn't exist.


Well of course it's not ambiguous like English because they had no historical baggage. But properly romanized Japanese is also never ambiguous.


"ene" can be えね and えんえ.

> But properly romanized Japanese is also never ambiguous.

What is "properly romanised Japanese" anyway? There are several modern systems in use [1] often mixed together (Even the government can't stick to the government mandated Kunrei-shiki). Some system represent ず and づ the same, some represent ティ and チー the same, some represent じ and ぢ the same. ō represents both O+O and O+U

[1] https://en.wikipedia.org/wiki/Romanization_of_Japanese#Moder...


えんえ should be en'e according to the govenment system. ティ is used exclusively in loan words anyway as far as I'm aware. Wouldn't you just write it in its original language?


Also the other ones he mentioned are pronounced the exact same way.


(you probably know this but) zu du and zi di do make a difference in terms of which kanji are used. The government has been reforming kanji readings over time though E.g. 藤 ふぢ → ふじ or 図 ヅ (look at the character in the box!)→ ず


Leaving out the apostrophe for えんえ is not standard in any of the three systems. Mixing and matching the systems is also not what I meant by "proper."


I thought 花 and 鼻 were ambiguous in most romanizations - at least I would write it as "hana" in all romanizations I can think of, yet the pitch accent differs. No?


There are systems (basically what looks like a long sideways L) but you're right that romanization doesn't record pitch accent. But there are big regional variations in pitch accent and people won't have trouble understanding you if you use the wrong one, mostly (and you could level the same complaint at hiragana or katakana, after all).

Nevertheless, that's a fair point that I wasn't really thinking about.


I would think the parent is referring to unambiguous in terms of pronunciation. You're right in that you wouldn't know which character was originally used in isolation—or whether kanji was used at all.

As far as I know, there's little (any?) pitch accent that distinguishes between words in Japanese. What pitch accent are you referring to?


Every word in Japanese has a standard pitch accent and some otherwise homophonous ones differ in pitch accent. However, from one dialect to another the pitch accents on the same word can be totally different so I don't think they're usually a big hindrance to communication. Pitch accent differs from English stress accent.

https://en.wikipedia.org/wiki/Japanese_pitch_accent


To be clear, you're saying that pitch can change due to regional accent, as opposed to distinguish between homophones, which is which is what 'singhblom was implying, correct?

I thought 花 and 鼻 were ambiguous … yet the pitch accent differs. No?


It does both things. Pitch accent is the only thing distinguishing some homophones (hana, hasi, ame, kumo) but it's also got variation in dialects including a handful where the Tokyo and Osaka versions have the same distinction between two words except with the opposite meaning (sorry, I don't have examples off the top of my head).

Fortunately there aren't that many sentences where it's equally plausible you meant both spider and cloud.


Interesting. Admittedly not a native speaker, I never found pitch accent used to distinguish between homophones during my time in Japan, nor was it introduced in any of the courses I took on Japanese while I was there, including the examples you provide.


I majored in Japanese and it wasn't really seriously introduced until I studied in Japan and I'd already been studying Japanese for three years. But while not learning it won't hinder your ability to communicate too much it can help you sound more natural. NHK publishes a dictionary of pitch accent in standard Tokyo dialect: https://www.amazon.co.jp/NHK-%E6%97%A5%E6%9C%AC%E8%AA%9E%E7%...


Thanks for the reference! Could I trouble you to provide an example from that text that describes such a homophone (as opposed to regional dialect) distinction?


The examples I gave upthread all have pairs with different accents in hyozyungo.


Okay. I was hoping for quotes from the text itself.


> but faulty pinyin was not.

Is this a real opinion or just a political stand against PRC? How is pinyin faulty, specifically with respect to Mandarin Chinese?

What Romanization system is better and how so?


OP elaborates more about this below.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: