Hacker News Reader

Comment by Sir_Cmpwn | Hacker News Reader

Recommended reading:

http://pinyin.info/readings/defrancis/chinese_writing_reform...

Summary: emperical evidence shows that students who were taught Pinyin (romanized, phonetic Chinese writing) were far better off than students who learned with Hanzi characters (the pictograms you're familiar with). Despite this, Chinese written language reform is unlikely to happen due to political pressures from the elites, whose family's positions are secured by the higher barrier to entry to becoming an intellectual. It takes more work to learn Chinese than almost any other language because they want it to.

For this reason, even setting aside the other moral quandries presented by China, I'm against teaching Mandarin in schools. Though, as an optional class for high-school-level children, it makes more sense.

Edit: probably worth clarifying that I study Hanzi characters and I am fluent in Japanese.

View on HN

>due to political pressures from the elites

As a Chinese, I can't think of ANY Chinese people would prefer to reform Hanzi to Pinyin and I'm surely not within "elites".

>whose family's positions are secured by the higher barrier to entry to becoming an intellectual

Holy shit, it's 2019 not 1950. China has a literacy rate of 96.4% [1]. What exact barrier are "elites" trying to set with that?

[1]: https://en.wikipedia.org/wiki/List_of_countries_by_literacy_...

> China has a literacy rate of 96.4%

So what? Literacy raises the average but doesn't prevent huge class gaps in populations. OP's point is valid in the same way that the 1% in the USA can pay for their kids to attend elite schools, be hired be elite firms regardless of ability, and continue the elite family tree often at the expense of the poor. China is no different.

So both China and the US should adopt newspeak?

Yes, it is ridiculous "more good" is spelt "better". I am unsure whether the exact form of "doubleplusgood" is optimal, but something like that makes perfect sense.

Given that China successfully adopted simplified Chinese characters, English-speaking world is perfectly capable of adopting Newspeak, and it will bring long term benefit, just as simplified Chinese characters did. But like many long term beneficial things it won't happen due to short term cost.

>Yes That's a myopic view of language. Changes like "better" --> "more good" don't last for long as irregularities will naturally form, making your "optimal" language yet another legacy constraint.

Besides, the vast majority of complexity of language exists because they carry semantic meaning. For example, the redundancy in the terms "Pig" and "Pork" might have initially been due to a need to signal sophistication, but they survive in modern usage because it's a concise way to differentiate the animal from the meat.

>Given that China successfully adopted simplified Chinese characters Simplified characters are more on par with spelling changes than newspeak. It only reduces the stroke count of certain radicals, it doesn't make any changes to the underlying vocabulary or grammar, whereas both newspeak and phonetic writing would.

Standardizing or simplifying the language adopted as lingua franca is a common practice. English is actually an exception in this respect. For example, see Bahasa Indonesia, which can be considered Malay newspeak, and very successful.

Yes, and that standardized lingua franca invariably splinters off and becomes less regular. Compare medieval Latin to the modern Romance languages.

> invariably

Well, we do have more TV and travel today than in prior millennia. Languages are dying out, while there are more people than ever. Let's see where the development goes.

>is unlikely to happen due to political pressures from the elites, whose family's positions are secured by the higher barrier to entry to becoming an intellectual.

I'm sorry but this is veering into the realm of conspiracy. If you actually talk to any Chinese people you'll realize that most are very opposed to the Latinization of the writing system. In fact the "elites" you speak of were themselves the ones most interested in simplifying the writing system and possibly adopting the Latin alphabet. Their former attempt, simplified Chinese, is often met with rancor in modern day Taiwan and Hongkong, whereas the latter attempt is often tauted in China as an example of CCP backed globalization gone too far.

Pinyin or Bopomofo [0] doesn't really matter what symbols are used.

[0] https://en.wikipedia.org/wiki/Bopomofo

The anger isn't towards Pinyin specifically, but towards attempts to switch the writing system to phonetic writing.

Pinyin is still taught in schools today starting from preschool. Most of the current generation of Chinese people are familiar with both and the common sentiment is that truly phonetic writing systems aren't practical, and that CCP attempts to use Pinyin as a "real" writing system were misguided and yet another example of the government trying to destroy Chinese culture.

> "If you actually talk to any Chinese people you'll realize that most are very opposed to the Latinization of the writing system."

Bopomofo isn't Latin characters...

But it's still phonetic.

I wonder if there's something about Chinese language that makes phonetic writing somehow less suitable for it. What about a possibility of a syllabic system?

Yes, that something is homophony. Mandarin has 1522 syllables when you include tone (413 syllables without). [0] English has 15,831. [1]

[0] https://chinese.stackexchange.com/questions/14596/how-many-s... [1] http://semarch.linguistics.fas.nyu.edu/barker/Syllables/inde... [via wayback: https://web.archive.org/web/20091115065840/http://semarch.li...]

The problem isn't Chinese in general, but the written register of Chinese specifically, and all the texts already using it. If you've ever looked at a verbatim transcript of a spontaneous conversation, you'll probably have noticed how much of an incoherent mess spoken language can be. Written speech is much tidier, because people have the opportunity to be very deliberate in what they write down.

For Chinese in particular, that means that people can be much more succinct in writing than orally, because the characters provide additional disambiguation beyond just indicating the pronunciation. Any purely phonetic writing system would require everyone to readjust their standard of what kind of sentences are comprehensible.

DeFrancis book The Chinese Language: Fact and Fantasy is excellent on that and many related questions [1].

First, Mandarin has no inflection (love/loves), which allows for writing in characters in the first place. Then, Mandarin has a fairly impoverished phonetic system, allowing for only about 400 different syllables ignoring tones, or 1200 or so when distinguishing tones (while English for example has some 7000 or so, without tones). This leads to many homophones, which can be distinguished by being written with different characters. Or so the traditional justification goes (I don't think much of that argument, because clearly people can speak to each other just fine, without having to clarify which word they mean - disambiguation happens by longer words (two or more syllables), which you could just write down. What happens is that in writing, you only need to write parts of a word: spoken out loud it would be ambiguous, but the chosen character disambiguates it, and thus people conclude that writing wouldn't work without the characters.)

However, arguably it's just a matter of habit (and resistance on the part of those that have already put in the time and effort to learn the characters). There are, in fact, Sinitic languages written alphabetically, eg Dungan [2].

The whole issue is very politicised, as you can see with the animated discussion here and the vigorous resistance to the initial comment (which was a bit unfortunately phrased, in my view, but fundamentally sound: Chinese is a very hard language to learn, mostly by virtue of its archaic writing system, and that is arguably somewhat deliberate.)

[1] https://en.wikipedia.org/wiki/The_Chinese_Language:_Fact_and...

[2] see the articles linked from here: http://languagelog.ldc.upenn.edu/nll/?p=40417

>just a matter of habit

That's a vast understatement. Adopting a writing system that would render everything written in the formal register incomprehensible isn't exactly a tenable position.

Europeans used to write in Latin, which is now incomprehensible to most, and yet, it has all worked out.

But granted, the transition took a while and was not just a matter of habit.

Having said that, why could one not publish important classic texts in a phonetic vernacular version, and leave the study of the originals to those with the inclination (just as most people in the West do not learn Hebrew, ancient Greek, and Latin; yet can find versions of the Bible, Euclid, Seneca, and Newton that they can read if they so desire.)

You're drawing a false analogy here, the problem isn't that phonetic writing isn't capable of rendering classical texts; the problem is that they can't competently render any text written in the formal register.

So, here we get to the fundamental issue then: Chinese seems to have a larger divergence between spoken and written register than most other languages (particularly alphabetically written ones).

It's unclear to me how far that extends. (This issue seems intertwined with the issue that many speakers of other Chinese "dialects"/"languages" also use Standard Written Chinese (ie Mandarin), btw.)

It is possible to write some register of Chinese in phonetic writing, though - that's shown by Dungan (written in Cyrillic), by the experiments in the 30s (quoting Mao again: "we have begun experimenting with Hsin Wen Tzu—Latinized Chinese. It is now used in our Party school, in the Red Academy, in the Red Army, and in a special section of the Red China Daily News."), etc.

Characters can be mapped to their phonetic equivalents no?

But they can't be mapped back.

Yeah, but not back (in other words, a mechanical mapping won't preserve sufficient information to disambiguate). It's definitely not trivial.

Chinese characters are syllabic.

Yes, and in between the former and latter, there was also a short-lived attempt to simplify characters further [0]. It failed, perhaps because it was completely contrived, whereas the first simplification adopted many existing cursive/shorthand conventions of the time. Unofficial conventions are easier to adopt, and Japan's independent Kanji simplification scheme actually ended up with some of the same exact (simplified) characters (e.g. 與→与, 學→学, 國→国).

0: https://en.wikipedia.org/wiki/Second_round_of_simplified_Chi...

It should be called Phoenicianization as you know.. That's where the concept of an alphabet and linear writing system comes from.

Writing has been invented several times independently, and AFAIK every culture that invented it made the jump to more phonetic writing (alphabetic/syllabic), at least partially, except China (well, maybe Indus valley).

How many cultures independently invented writing? By most accounts it’s either 3 or 4 times. English comes from a long family of alphabet writers, all the way back to Sumerian writing.

> Latinization

It's been always puzzling to me why borrowing Latin alphabet is usually considered today the only option. Many languages have phonetic alphabets of their own. Why is it that in the past people had no issue creating an original alphabet well suited for the language, and now the choice is between some bastardized form of Latin on the one hand and sticking to the inconvenient archaic writing system (however beautiful it looks) on the other.

It's English that has the bastardized Latin!

Things like "ough" for /a:f/, and vowel shifts whereby /a/ gets written "u".

Compared to that, some languages have sane Latinization.

Since English is so pervasive, people round the world can look at it and feel good about their own Latinization.

> It's been always puzzling to me why borrowing Latin alphabet is usually considered today the only option. Many languages have phonetic alphabets of their own.

The Latin alphabet is the most popular phonetic alphabet, and keyboards (and formerly typewriters) for it are widely available. Not so for many of the others.

I can see that being a major motivation for Latinization, especially for smaller languages.

The argument that Chinese writing has to be made phonetic because Chinese is too hard only holds sway to non-Chinese. Most Chinese people do not consider Chinese writing to be too hard. Therefore the most common argument for phonetic writing is compatibility with international standards. As of now the Latin alphabet is the international standard.

> only holds sway to non-Chinese.

Well, I'd say it holds sway to learners that a) know a better system, and b) learn as adults. Both factors allow them to form a better judgment than kids that don't know any better.

Except all Chinese speakers in the current generation are taught Pinyin in school.

And many speakers write characters (on their computer or devices) using Pinyin, and forget how to write rare characters, see

* https://en.wikipedia.org/wiki/Character_amnesia

* http://languagelog.ldc.upenn.edu/nll/?p=25291 (and the articles linked from it)

Yes and English speakers forget the spelling of rare words, especially if they were to spend extended periods of time using nothing but text-to-speech.

Those that do use Pinyin to write on their computers also know that the large number of homophones makes it an order of magnitude slower than using radical based input methods to directly type the characters such as Canjie, Wubi, Boshiamy etc.

Completely agree, see the mess that's been made in Kazakhstan: https://www.bbc.co.uk/news/blogs-news-from-elsewhere-4312681...

This ain't the 1800s. The vast majority of Chinese citizens who have finished elementary school are able to recognize at least 2000 characters. Knowing how to read and write Hanzi is no longer an indication of "intelligence" or "elite status". In Japan, maybe, but their language is completely different.

In case you've missed the last half-century, there has already been two Chinese written language reforms (though the second one didn't stick). It's called Simplified Chinese.

The rationale in your comment is absolutely bizarre and it reeks of arrogance with a side sprinkle of ignorance.

Not in Japan either really.

Turning character into pinyin introduces so much ambiguity and lose so much information at local level that we need to go back and forth multiple times to infer what that word really is according to the partially parsed part of the sentence. While written sentence in character we don’t even need to do linear character by character scan, usually a glimpse of it will give you most of what you need to understand it.

If you include an indication of the tone along with the pinyin, doesn't it end up being just about as informative as spoken word?

There are still way too many hompohones which can introduce a lot of amibiguity. Even Korean still includes Hanja to clarify technical terms in law/medicine.

Modern Standard Mandarin has way fewer syllables than Middle Chinese and Southern Chinese languages which makes homophones a lore more common. Vietnamese didn't really have this problem because the phonology is a lot more complex and it shows through the written word.

An example of homophones creating ambiguity: An Jung-geun was a Korean independence activist who assassinated a former Japanese Prime Minister in 1909. He is often called 義士 in Korean textbooks which means "hero, man of honor". However, 義士(의사) is pronounced in the same way as 醫師(의사) which means "doctor". The Hangul spelling is identical thus it has led to many younger Koreans to think An Jung-geun was a doctor and question why would a doctor intentionally kill someone. It's more of a tongue-in-cheek joke but it shows the problem of completely detaching a writing system from a language that does not sufficiently distinguish words without context. In Chinese this is not a problem because by people usually learn this type of stuff on paper first. The additional context helps reduce the ambiguity in spoken conversation. Being able to talk to each other without writing is not sufficient proof that a writing system can be abolished with minimal impact, because it ignores the fact that the writing system is what helped to dispel the ambiguity in the first place.

> There are still way too many hompohones which can introduce a lot of amibiguity.

Is this a problem when speaking? How many times a day do speakers of Chinese have to ask a collaborator which homophone they are using?

Enough times that replacing Chinese characters with a fully phonetic system will introduce more problems than it solves.

Did you read the example I wrote? Homophones usually aren't a problem because most people learn things in writing first, and that provides context in speech.

> In Chinese this is not a problem because by people usually learn this type of stuff on paper first. The additional context helps reduce the ambiguity in spoken conversation.

I read that, but I'm having trouble imagining how that works.

Do you have an example?

But I literally gave an example...

"Queue" and "cue" which are homophones in English. Now imagine if there is a language reform where IPA replaces the current writing system, and both words are replaced with /kjuː/ as the only written form.

"/kjuː/ the crowd" can be either "Queue the crowd" or "Cue the crowd". We are aware these two words are distinct from each other, since we learned these two words in their written form already, and in certain situations we won't have any ambiguity as to which action needs to be taken. If the reform lasts long enough that a new generation of people grows up only learning "/kjuː/" and not "queue/cue", they won't be so sure about which action is to be taken in "/kjuː/ the people" when it is presented as an isolated phrase without context.

Think of a hash table that resolves hash collisions via separate chaining. The collision is when words sound the same (have the same index). Abolishing the current writing system is like dumpking the bucket contents and only returning the index. It's not an immediate problem because we still know that the bucket used to contain chained entries ("cached" in our brain memories"). Those who have never accessed the hash table (e.g. younger generations receiving education after the change) won't know what was in there before. They just know the index but not that there used to be two different data entries stored at the same index. The downside of losing information should be evident.

So you're saying that, when conversing via speech, the problems created by ambiguities such as

"I heard a wail and I saw a whale" or "The knight was black"

are more tractable than the problems arising from

"When I lie, I lie on a bed" or "British left waffles on Falklands"

Is that right?

I'm not so sure. Given that you can only hear but not see, I feel they're probably equally problematic when there is not sufficient context to resolve the ambiguity.

The thing that changes if you move to a phonetics-based writing system, is that you're forced to include more context while writing. I don't think anything changes about oral communication. (But I'm happy to be refuted!)

>The thing that changes if you move to a phonetics-based writing system, is that you're forced to include more context while writing.

That would be a fundamental change for the language itself and I don't see why would it ever be necessary.

A person who has never heard of the first two sentences you mentioned will not be able to understand the meaning until you provide a long explaination of which word is which ("more context"), or just write down the sentence itself. Which one is faster and more concise? I don't see how does "British left waffles on Falklands" involve any homophones at all. Ambiguity of homophones can be resolved via context, but that's not one-and-the-same as all "out of context" sentences.

The beauty of the Chinese language is that it can be incredibly concise on paper despite having the same sentence sound like gibberish when read, which is not a problem as long as the current writing system still exists. The information density in East Asian languages are higher in general thanks to this.

"British left waffles on Falklands" is a moderately famous, ambiguous, historical newspaper headline from the UK [0]

If you have absolutely no context and know nothing about politics or human society, you'll probably take it to mean that the British went to the Falkland islands and then when they left the left behind some waffles.

In actuality it meant that someone was waffling on a decision to be made regarding the Falklands ("waffling" means being indecisive) and that someone was "the British Left" (the liberal faction of parliament).

[0] https://www.huffingtonpost.com/jonathan-haber/british-left-w...

It’s not uncommon for a chinese speaker to say something and for the other speaker to say something like “which X, is it X with Y radical?”, or “is it X as in <word containing X>”. This happens especially with names.

> especially with names.

That makes sense! Since names will always lack disambiguating context. And I've seen that in Chinese movies come to think of it.

It happens in English too: Is your name Kerry or Carrie?

Sure, but written Chinese is very different from spoken Chinese. That is the case in any language, but even more so in Chinese. Much more concise. That's the problem, really - the written language would have to change (beyond switching from characters to alphabetic writing), to be closer to spoken language.

That's interesting!

Do you have a good example of a common sentence/thought/idea that is expressed with much greater concision in written chinese than spoken?

Surely a basic sentence like "mingtian wo xuexi yihou qu ti zuqiu" can't really get much shorter?

Hopefully that's a real sentence. I speak mandarin at something like the level of a 3-year-old kid.

Great question. There might be better examples, but let's look at one of the most famous poems, "Thoughts in the Silent Night" [1] by Li Bai:

  床前明月光，
  疑是地上霜。
  举头望明月，
  低头思故乡。

"Translated" word for word, it's something like

  Bed before bright moon shine
  Think be ground on frost
  Raise head hope bright moon
  Lower head think home

Note that it doesn't even contain 我 ("I"). One can render it as:

  Beside my bed a pool of light—
  Is it hoarfrost on the ground?
  I lift my eyes and see the moon,
  I bend my head and think of home.

Or:

  Before my bed, the moon shines bright;
  Be it frost aground? I suppose it might.
  I lift my head, the moon to behold, then
  Lower it, musing: I'm homesick tonight.

Moser [2] offers these gems on Classical Chinese: "Forget it. Way too difficult. If you think that after three or four years of study you'll be breezing through Confucius and Mencius in the way third-year French students at a comparable level are reading Diderot and Voltaire, you're sadly mistaken. There are some westerners who can comfortably read classical Chinese, but most of them have a lot of gray hair or at least tenure." "Whereas modern Mandarin is merely perversely hard, classical Chinese is deliberately impossible. Here's a secret that sinologists won't tell you: A passage in classical Chinese can be understood only if you already know what the passage says in the first place. This is because classical Chinese really consists of several centuries of esoteric anecdotes and in-jokes written in a kind of terse, miserly code for dissemination among a small, elite group of intellectually-inbred bookworms who already knew the whole literature backwards and forwards, anyway."

Wikipedia says [3]:

"Classical Chinese is distinguished from written vernacular Chinese in its style, which appears extremely concise and compact to modern Chinese speakers, and to some extent in the use of different lexical items (vocabulary). An essay in Classical Chinese, for example, might use half as many Chinese characters as in vernacular Chinese to relate the same content." "Classical Chinese rarely uses words composed of two Chinese characters; nearly all words are of one syllable only. This stands directly in contrast with modern Northern Chinese varieties including Mandarin, in which two-syllable, three-syllable, and four-syllable words are extremely common." "However, even with knowledge of grammar and vocabulary, Classical Chinese can be difficult to understand by native speakers of modern Chinese, because of its heavy use of literary references and allusions as well as its extremely abbreviated style."

However, I don't have a neat example at hand... Anyone?

[1] https://en.wikipedia.org/wiki/Quiet_Night_Thought

[2] http://www.pinyin.info/readings/texts/moser.html

[3] https://en.wikipedia.org/wiki/Classical_Chinese#Grammar_and_...

静夜思 is taught in early elementary school and not exactly what most native speakers would describe as a difficult poem.

Not difficult, concise (and in particular more concise than the thought would be expressed in spoken Chinese).

Classical Chinese is not modern written Chinese.

True. My point is that written Chinese register tends to occupy a space between Classical Chinese and modern spoken Chinese.

That's a very nice poem (thanks!), but not a good example, mostly because it is a poem.

As a poem, it has no equivalent rendering in colloquial Chinese. Also as a poem, it deliberate sheds information, permitting (or even pursuing?) ambiguity in its quest for specific phonics, scansion, symbolism (in many cases - maybe not in this one), etc.

Probably the way to get an example is to start from something spoken in modern Chinese and see how it comes out different in written Chinese?

I don't know what your "fluent" means. But I recommend keep studying Japanese until you can read newspaper in native level speed and try to imagine the newspaper in 仮名 only 漢字無し, then in romanized latin letters.

Ane for fun: https://en.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_...

It is likely that this measure is being pushed by China in their bid to expand their global influence. Many such projects are being pursued right now in developed nations; one of them is probably pushing Mandarin as a second language.

In Kenya, that would be more like Mandarin as a fourth language. They commonly know their tribal language first, then Swahili and English, which are both official languages of Kenya. From fourth grade on English is the primary language used for teaching in school.

china has seemed to have recognized that chinese is too hard for most people to learn as a second language, and have been sponsoring efforts at establishing esperanto as an intl langauage

They’ve been sponsoring an Esperanto radio station for awhile (like half a century at least) but there haven’t been any increased pushes I’ve seen lately.

I have never heard about esperanto while in China, that was half of my lifetime, so I think there must be some exaggeration here. The efforts at establishing Confucius Institute branch schools all over the world to teach Mandarin I have clearly seen.

To be honest, I don't understand your reasons at all.

http://www.pinyin.info/readings/texts/moser.html

I can sum it up as "it's deliberately hard and teaching it to children has been shown to slow their learning"

Except both of those are wrong.

Remember, literacy was a rare and elite thing throughout the Western world as well, until the 19th century started seeing the first attempts at universal education. Widespread literacy is a 20th century thing. And China lagged behind the West, having spent the first half of the 20th century throwing off colonial oppression rather than developing a sophisticated, modern education system. So until recently, it cost China nothing to use pictograms rather than romanization. So no, not "deliberately hard".

Since then, China built a modern education system, and has literacy rates on par with western nations. So it's not preventing China from teaching children to read. Therefore, your claim that "teaching it to children has been shown to slow their learning" constitutes an extraordinary claim, and requires some extraordinary evidence.

Occam's Razor suggests that you think it's hard because it's unfamiliar, not because it's hard.

I'd think it's fairly uncontroversial to say that the Chinese writing system is harder to master (and it thus takes longer) than alphabetic ones.

It's an extraordinary claim that learning 6000 symbols is just as easy as learning 26 symbols (hey, make it 52, or, since we're on HN, 128 or even 256 :-)

It's not a claim I'm making, because it's an irrational comparison. My son knew the alphabet by the time he was three years old. Did it give him a 6000 word vocabulary? No. Standalone Roman letters are meaningless, unlike Chinese characters.

And there are other tradeoffs. Chinese writing is much more compact and easier to read, relative to alphabetic writing. A layer of translation has been removed.

I'm not convinced that learning to write Chinese with a "literate" vocabulary is substantially harder than achieving equivalent literacy in a language with alphabetic writing. It might look easier to you if you've never learned both. But proof is in the pudding - China's literacy rate is on par with Western literacy rates, for similar educational effort.

> China's literacy rate is on par with Western literacy rates, for similar educational effort.

Well, that's precisely the claim that is under dispute. For example, DeFrancis [1] aims to dispel what he calls

"The Successfulness Myth: Chinese characters are responsible for a high level of literacy in East Asian countries. (A weaker version of this myth is simply that despite the flaws of Chinese characters, East Asian countries still have a high level of literacy.)"

He makes the opposite claim that (from the Wikipedia summary)

"The Chinese script, with its huge number of characters, its complexity and its irregularities, is harmful to the literacy improvement efforts of the Chinese society, and needs to be replaced by a more efficient writing system if China is to achieve the benefits of modernization."

There is, in other words, serious scholarly dispute about the success of the Chinese script.

> Chinese writing is much more compact and easier to read, relative to alphabetic writing

Agreed on the first part, it's more compact (not surprising if you have ten thousands of choices per character, say 14 bits, instead of 26, or 5 bits). But I don't think it's easier to read (let alone write), and I'd want to see evidence for that.

[1] https://en.wikipedia.org/wiki/The_Chinese_Language%3A_Fact_a...

Couldn't the same thing be said for the horrendous spelling and grammar issues of English?

For what it's worth, my spouse has a Master's degree in Chinese pedagogy. I've not heard such arguments from her. A single book, weighed against observation, isn't proof.

> horrendous spelling and grammar issues of English?

Sure, English has pretty horrible orthography, arguably the worst among alphabetic languages (though it's not that easy to reform, either, see eg. [1]). But it's still more phonetic than Chinese. Spanish has a beautiful orthography, and pretty close correspondence between written and spoken word.

"Chinese characters are so difficult to learn that even the best system of rudimentary characters, or simplified teaching, does not equip the people with a really rich and efficient vocabulary. Sooner or later, we believe, we will have to abandon characters altogether if we are to create a new social culture in which the masses fully participate." "We believe Latinization is a good instrument with which to overcome illiteracy."

This was said by Mao Zedong (oh, dang, that's so ambiguous, I mean 毛泽东) in 1936, see [2].

"This has led many Chinese in the past to advocate the abolition of characters in favor of an alphabetic system, but such programs have met with little success." [3]

So, the idea that the Chinese writing system is too cumbersome is by no means new, or "Western", for what it's worth.

> A single book, weighed against observation, isn't proof

Well, no. But there is more than a single book, there are decades of discussion and scholarship on this.

It's conceivable to me that China could benefit from a move to alphabetic writing, and translations of Classical Chinese into vernacular (and anyone who wants can go study characters and Classical Chinese texts in the original, just like anyone in the West can go learn Latin and ancient Greek and Hebrew and read the classics in the original, or the Bible, for that matter).

[1] https://www.phon.ucl.ac.uk/home/wells/accents_spellingreform...

[2] http://www.sino-platonic.org/complete/spp171_chinese_writing...

[3] http://sites.asiasociety.org/education/VISIBLE_TRACES/curric...

> Sure, English has pretty horrible orthography, arguably the worst among alphabetic languages (though it's not that easy to reform, either, see eg. [1]).

English orthography, summed up, is: all the orthographies from all the languages from all the time periods cobbled together into a monster.

Nicely put.

I don't think quoting Mao is doing your argument any favors.

>It's conceivable to me that China could benefit from a move to alphabetic writing, and translations of Classical Chinese into vernacular

The issue is not that classical texts become illegible, the issue is that all written language becomes illegible. For example, in the other DeFrancis source you provide, he uses the term "shuangwenzhi", the meaning of which I simply could not figure out. I saw the English "digraphia" below, at which point I assume he may have meant "双文zhi" but I am still unclear what he meant by "zhi". As of right now only 66% of the phrase he used has any semantical meaning and I am still unclear as to what he meant to say by "shuangwenzhi". Without the English translation I don't think I would've deduced any of the semantic meaning of that term.

>So, the idea that the Chinese writing system is too cumbersome is by no means new,

You are correct, this argument isn't new, it's outdated.

Nicely put :-)

The emergence of smart ubiquitous always-connected computers (with Unicode) does change the evaluation maybe.

I think one of the fundamental arguments against the Chinese script (it takes about twice as long to learn than alphabetic writing) stands, while others are outdated, indeed (nobody struggles with building Chinese typewriters anymore).

However, there are many arguments put forward against switching to an alphabetic script that strike me as misguided (that the Chinese language is so special, that it has so many homonyms, that characters are uniquely suited to it and indeed indispensable, that it allows speakers of different "dialects" to communicate, etc.).

At any rate, good discussion.

>At any rate, good discussion.

Agreed.

Mao led the development of the simplified character system, iirc, so it's relevant. That was part of the push for universal literacy. It is much easier to learn than traditional characters.

The book you keep citing was published in 1984. A lot has happened in China, especially in regards to education reform between 1984 and 2018. Aside from that book, I see no other signs of "serious scholarly dispute about the success of the Chinese script.".

The Internet and Pinyin input methods has rendered these calls of reform largely irrelevant. Handwriting skills outside the education system are on the decline due to digitization, but the positive side is that any literate individual can utilize Chinese characters in online communication with trivial effort. For non-trivial ones there are always online dictionaries which can be accessed via online search in seconds.

An opinion on the Chinese language from 1984 is not applicable in 2018, pretending that it does is disingenuous.

Good point. Particularly with text-to-speech (and vice versa!) also being available now.

I'm still sorry for all those kids having to spend years of their childhood on characters, but maybe one can see it as a character forming exercise.

Danish children acquire language skills at slower pace than six other languages (including Swedish which is in the same language sub-family): http://cphpost.dk/life-in-denmark/the-danish-languages-irrit...

Is there also a conspiracy by the Danish "elites" to deliberately keep the language's phonological complexity so there's a "higher barrier of entry" to become an "intellectual"?

Without using profanities, I'd just say that summary is absolute NONSENSE.

This would be a better reason to discourage the rest of the world from learning English, a language whose native speakers often use familiarity with its laughably bad orthography as a proxy for assessing intelligence, and also a language which unlike Mandarin has seen few serious attempts at reform (not only because of lack of centralised bodies, but also because English linguistic prescriptivists have generally preferred inventing tendentious grammatical style rules to classify "good writing" over cutting the bits which might confuse people).

On the other hand, English is quite a useful language to be able to speak, much like Mandarin.

Agreed. DeFrancis (in his delightful The Chinese Language: Fact and Fantasy) wrote (quoting from memory) that Japanese has the worst writing system among the character-based ones, and English the worst among the alphabetic ones.

What a pleasure Spanish writing is.

https://en.wikipedia.org/wiki/The_Chinese_Language%3A_Fact_a...

The advantage of pictograms is you can infer meaning from component radicals when looking at more complicated pictograms. Pinyin is useful but limited, e.g. there is a number of different pictograms that are all pronouced 'shi4' so simply by looking at pinyin you won't be able to differentiate without additional context that might not be available in written scenario.

Homonyms exist in every language. The advantages do not outweigh the drawbacks.

There's order of magnitude more homonyms in Mandarin than English to the point where having a pictograms system starts making sense. That's why 80 percent of characters are multipart with one of the radicals carrying the pronunciation.

Homonyms don't exist in every language to the extant that you can write an entire essay with them.

https://en.wikipedia.org/wiki/Lion-Eating_Poet_in_the_Stone_...

> Homonyms don't exist in every language to the extant that you can write an entire essay with them.

Homonyms like that don't exist in any living language. Spoken Classical Chinese is a dead language, like Latin. You only get that effect when you use modern Mandarin pronunciations, they wouldn't have been homonyms if you didn't use anachronous pronunciations. It's sort of like writing a poem that means one thing in Latin and an entirely different thing in modern Italian.

Yet Classical Chinese is still a large part of Chinese literary canon, partially because the writing system allows for written Classical Chinese to comprehended by modern readers. There are no examples to "Lion-Eating Poet" in Italian and English precisely because phonetic writing systems cause most writing to be incomprehensible as pronounciations change.

Your example is also egregious also because the entire point of "Lion-Eating Poet" isn't that "it means different things in classical and modern Chinese" but that semantical meaning can still be transmitted by the writing system completely independent of pronounciation.

> Yet Classical Chinese is still a large part of Chinese literary canon, partially because the writing system allows for written Classical Chinese to comprehended by modern readers.

But only essentially though translation. It's different enough that a typical Chinese reader can't jump into those works with ease.

Latin works are still part of the Western literary canon, and for many centuries Europeans went though the trouble of learning Latin to read them. You can still find fossilized Latin phrases in modern languages like English, et cetera. That doesn't mean Latin is a truly living language.

> There are no examples to "Lion-Eating Poet" in Italian and English precisely because phonetic writing systems cause most writing to be incomprehensible as pronounciations change.

> Your example is also egregious also because the entire point of "Lion-Eating Poet" isn't that "it means different things in classical and modern Chinese" but that semantical meaning can still be transmitted by the writing system completely independent of pronounciation.

There can be no exact analogy because phonetic writing systems are more conserving of pronunciation while systems like Chinese are more conserving of semantics. That's why I said it's "it's sort of like." My point was the poem is more of a stunt than a real statement about a real language.

>But only essentially though translation.

Classical works are not translated into Modern Chinese.

I think that's the source of our disagreement. When the writing system is focused on semantics, then the threshold where something is considered translation is different than for a phonetic writing system.

>Latin works are still part of the Western literary canon.

Yes, but most modern westerners can not read Latin on it's own, because of the writing system.

>My point was the poem is more of a stunt than a real statement about a real language.

What is a "real language" here? The poem is comprehensible to modern Chinese, although the grammar follows Classical Conventions. What makes it not a "real language"?

> What is a "real language" here? The poem is comprehensible to modern Chinese, although the grammar follows Classical Conventions. What makes it not a "real language"?

I should have been more specific: I meant a real living language. The poem is a stunt that demonstrates how far modern Mandarin pronunciation has diverged from Classical Chinese (which is dead as a spoken language), but it doesn't make much of a statement about the practicality of replacing hanzi with a phonetic system for writing modern living Mandarin.

>doesn't make much of a statement about the practicality replacing hanzi with a phonetic writing system for everyday use.

What is everyday use? Surely a writing system should be capable of more than "Hello, my name is Knolax?". If a proposed new writing system was incapable of representing Shakespeare in any comprehensible way, would it be suitable for English?

I realize now that you consider Classical Chinese as a seperate language from Modern Chinese. I and most Chinese speakers would probably disagree, but I don't think that's a line that can be drawn without a lot of nitpicking from both of us.

> What is everyday use? Surely a writing system should be capable of more than "Hello, my name is Knolax?". If a proposed new writing system was incapable of representing Shakespeare in any comprehensible way, would it be suitable for English?

Sorry I edited on you, but what I meant was a phonetic system could be used in all ways contemporary spoken Mandarin is currently used without getting too tripped up by homophones.

There would obviously be difficulties when dealing with very old texts, but I think such difficulties are inevitable regardless of the language as the changes pile up. Something will be lost unless you spend the appropriate effort learning the state of language as it existed when the text originally written.

> I realize now that you consider Classical Chinese as a seperate language from Modern Chinese. I and most Chinese speakers would probably disagree, but I don't think that's a line that can be drawn without a lot of nitpicking from both of us.

Yeah.

> I meant was a phonetic system could be used in all ways contemporary spoken Mandarin is currently used without getting too tripped up by homophones.

Except phonetic writing systems are aren't even capable of representing Chinese names correctly. I think you're underestimating the extent to which homophones are prevalent in Chinese. It's not just classical texts, it's all written texts and most non-trivial spoken words, including names. The vast majority of Chinese morphemes are only a syllable long, and have up to hundreds of homophones. Unlike most European languages, Chinese names do not all come from a fixed list like "John. Mary, Jane, etc..." and are instead constructed directly from morphemes. The result is that whenever you encounter a Chinese name transliterated phonetically in any system, there is no deterministic way of converting it back into Chinese without ambiguity.

> I think you're underestimating the extent to which homophones are prevalent in Chinese. It's not just classical texts, it's all written texts and most non-trivial spoken words, including names. The vast majority of Chinese morphemes are only a syllable long, and have up to hundreds of homophones.

I guess my question is how much of a problem are these in spoken communication (say over a telephone where there's less shared situational context)? My intuition is that if the prevalence of homophones were so troublesome to make a phonetic writing system difficult/impractical to use, they'd also make spoken telephone conversations similarly difficult/impractical. Since Chinese people can obviously speak productively on the telephone, then there must be enough context in the phoneme-stream that it's sufficient to effectively communicate without the extra context provided by the characters.

That's not to say the characters don't provide an extra layer of depth and subtlety to written text over phonetic writing.

> an extra layer of depth

I think that's the best way to put it. Like you intuited, a phonetic system is perfectly capable of writing down spoken conversations, but without as much context as a spoken conversation it's going to get more confusing. For example, he she and it in Chinese are all pronounced "tā", but are written as 他，她， and 它 respectively in Chinese. In fact it's common to write "ta" as an equivalent to xhe/xer/xem.

Homonyms are words that are pronounced the same way. These words use different tones.

There are four tones. The poem has 33 unique words.

Good point.

Beyond that, FWIW, homophones are words that are pronounced the same way. Homonyms are pronounced and spelt the same way. I guess in the context of Chinese (where disambiguation by character is the big issue), one should speak of homophones?

On this issue, see also "The Homonym problem", in Wm. C. Hannas, Asia's Orthographic Dilemma:

http://pinyin.info/readings/texts/east_asian_languages.html#...

And yet people speak Chinese without displaying characters for help in comprehension.

Because spoken language usually has more context (as compared to a brief written sign or a headline).

In those contexts, I'd think you could usually choose a different phrasing that wouldn't be ambiguous.

>'shi4'

but isn't that when you use context around the word to know what the meaning is? The same happens in English...you don't always know without context.

学习中文很有趣！

Xuéxí zhōngwén hěn yǒuqù!

hanzi or pinyin reading either way I know what this is intended to say.

Except Mandarin has a lot more homophones than English does, likely under the long influence of the writing system.

For spoken Mandarin, you can probably infer enough from the conversational context, but for official documents, business and technical terminology, or any literature, full romanization renders the text much less incomprehensible.

This isn't helped by the fact that Classical Chinese, from which most complex vocabulary is derived, is even more homophonic than daily speech.

That's just spoken language. There are many words that are "literary", which means they are used primarily in writing and not spoken language. Without hanzi it would be very difficult to understand what is written.

Yup this can be true.

If you read something like duo(3) it could be a measure word for flowers (朵) or maybe mean “to hide or avoid” (躲)

totally agree with other two comments, tbh, nobody uses pinyin anymore after grade six, it is really just a stepping stone for kids to learn chinese before they can memorize all the common characters help them to read. As a Chinese i actually hate pinin as an adult because my name would appear in Pinin on my passport and it loses all its meaning.

Yeah I follow you. Pinyin helped me learn tones and reading. But now I can look at a character and find the radical. Then could the remaining strokes and look it up in a paper dictionary.

I have the books school children buy in china and yeah 7+ it’s just characters.

> The advantage of pictograms is you can infer meaning from component radicals when looking at more complicated pictograms. Pinyin is useful but limited, e.g. there is a number of different pictograms that are all pronouced 'shi4' so simply by looking at pinyin you won't be able to differentiate without additional context that might not be available in written scenario.

I kinda find it hard to believe that could be much of a real problem. Pictograms may allow homonyms to proliferate more than in an alphabetic writing system, but as long as the language still lives as a spoken language, there will be limits on them that prevent them from becoming too much of a problem. A spoken language with too many homonyms just isn't useful, so it will be adapted to eliminate the problem.

All bets are off when the spoken form is dead, though.

You have the same "advantage" with word components in alphabetic languages, such as "picto+gram", "sub+marine", and what have you.

Yes and in alphabetic languages those word components are much more ambiguous, especially in abbreviations. 加州 is a lot less ambiguous than CA, hence the greater prevalence of the former in daily use.

> alphabetic languages those word components are much more ambiguous

I don't think so. The components of a polysyllabic word can give (more or less) strong hints to its meaning, but typically the word and its meaning must still be learned. But the same holds for radicals in Chinese characters, and even polysyllabic words in Chinese (电视/電視 is electric + vision, and that's TV. Television is distant + vision, and that's also TV.）

That two characters with 14 bits each carry more information than two characters with 5 bits each is clear, but not a huge advantage of non-alphabetic writing, or a decisive one (compared to the disadvantages).

Yes but the point is that CA and 加州 are both two syllables in speech. In your example, "televison" is 4 syllables whereas "电视" is only two. As you've noted, the writing system has over time created a higher information to syllable ratio. A transition to a writing system with a lower information to syllable ratio would require a complete change in vocabulary and the translation of pretty much all written text.

> the translation of pretty much all written text

Yes, agreed. It would by no means be trivial. Question is whether it would be worth it (in terms of freeing up time for children to study something else, for example).

Better off how?

Reply on news.ycombinator.com