Characters are objectively harder, even for Chinese

[Ian_Lee]

Most of the words in Chinese vocabulary consists of two characters. Try figure out how many combinations there can be within the 3,000-3,500 characters in dual form.

Actually I absolutely have no argument that Chinese is hard for learners who are born using a Romanized script (But I would say Japanese is even harder).

But I wonder how can they determine that Chinese is hard for Chinese speakers merely on the circumstantial evidence that some kids forgot how to write certain characters.

Just like I would not say that Japanese is hard for Japanese kids or English hard for American kids.

In fact, some Romanized languages are so hard that their grammar actually sounds silly to me.

For instance, the gender assigned in the German language -- Der, Die, Das -- is outright absurd.

Why is the door in German female and the table neutral and the chair male?

German also has lengthy complicated word without space or punctation, i.e. funfhundertzwanzigdrei in number.

But would I say German is hard for German or Austrian kids? Of course not.

[sm_sung]

Green Pea wrote:

Which is easier when learning to write: memorise 100 rules with a 98% predicative accuracy or 4,000 characters with only 5% predictive accuracy?

I know these figures were made up for demo purposes but I think they’re skewed a little too much. I'd estimate English spelling accuracy to be lower and Chinese "character formation accuracy” at 30-40% or more.

By the way, in Chinese one need only learn to write about 3000 characters. As Ian Lee pointed out, most Chinese words consist of two characters which would mean a Chinese user need only remember the right characters (morphemes) and put them together. Contrast this with English where one has to learn how to write down many more words (compound words are used much less frequently in English). In addition, English has many redundant words.

Example: Is there any appreciable difference between rotate, spin, turn, revolve?

What is interesting is why the number of commonly used Chinese characters tops out at 3-5,000. I would argue that beyond 5,000 the memorisation task is just too overwhelming…most can’t remember that many for practical usage. The system of writing Chinese just breaks down. English spelling rules do not breakdown.

I would argue that the average person NEEDS up to 5,000 characters in his daily life. I can't emphasize this enough: Chinese words are mostly compounds, 5000 characters means a lot of words. English spelling rules do break down. “breakdown” is a noun not a verb. Besides English loanwords (like “coup de grace”, “avant-garde” etc) often retain foreign (i.e. non-phonetic) spellings.

[ala]

English has many redundant words.

Example: Is there any appreciable difference between rotate' date=' spin, turn, revolve? [/quote']

There is an appreciable difference in English.

One doesn't rotate a screw voluntarily, one turns a screw (as it can be a motion that involves less than a complete cycle. Another example: I turned to face you). One certainly does not revolve a screw. The planets revolve around the sun, as it is a physical condition of orbiting AROUND a separate and relatively fixed object. One doesn't spin a screw as that involves taking a screw, putting it on the desk and spinning it like a toy (the act of rotating something very fast for a certain moment). To rotate is to refer to the STEADY motion of a body around its OWN center or axis, the earth rotates 360 degrees each day.

On the contrary, I find Chinese vocabulary to be extremely repetitive as the understanding of words differ so much between individuals, leading to high overlap between words, and general indecision in choosing words. Again this relates directly to the imprecision of Chinese words.

[Taibei]

to me a Chinese text written in pinyin is like a scribbled note written in English. Easy to write, hard to read. Of course, I don't read pinyin as often as characters, but I still feel that the high degree of homophony in Chinese would make me a slower reader if it weren't for the characters.

This is often asserted -- without any proof. Moreover, if we're talking about time here, how much more time does it take to become literate in Chinese characters than in pinyin? In this, we're talking about years. The notion that Chinese characters save people time has always struck me as absurd.

The so-called homophone problem is also largely a myth. When words are written as words, the problem largely disappears, especially when context is provided.

They don't fit Mandarin or any of the other Chinese languages particularly well, which is in part why they are so damn hard, to borrow Moser's phrase.

And why don't they? One character represents one morpheme and has one sound, most of the time.

And just what is that sound?

How do people know how to pronounce the phonetic compenent of a character? Only about one-quarter of characters supply accurate phonetic information (proper pronunciation, including tone). And even in those few cases it's not always clear which element is the phonetic one (because characters are not regular in this either).

Most characters fall into the category of kinda, sorta when it comes to phonetic usefulness, and that's putting it kindly. Many others are simply useless to modern speakers and must simply be memorized.

I wouldn't say they are perfect but don't see any obvious flaws either.

• A writing system that is an especially ill match for the sounds of the spoken language.
  
• A writing system that is so difficult it takes many more years of effort to learn than alphabetic systems -- and even then it is learned incompletely, even by its finest scholars.
  
• A writing system that is so difficult it has hindered the literacy of the largest single group of people on the planet.

And on and on and on....

And, please, everyone remember: The comparison that matters is not English spelling vs. Chinese characters but Chinese characters for Mandarin vs. romanization for Mandarin (which is enormously easier than English spelling).

[blob]

* A writing system that is so difficult it takes many more years of effort to learn than alphabetic systems -- and even then it is learned incompletely, even by its finest scholars.

Well yes, but I think that no one really learns this completely for any language. As was pointed out before, comparing a english alphabet to a chinese character isn't fair, comparing a character to an english word is more fair I would think. So, honestly, I don't think there are that many people in the world which knows all the words in the entire oxford dictionary, even by the finest scholars.

* A writing system that is so difficult it has hindered the literacy of the largest single group of people on the planet.

Would this be because of the fact that wenyanwen was used for such a long time before the revolution to baihua for written text? Perhaps the economy too had a role to play in this? (some other thread discussed this, I cannot remember which.... ). I think europe had quite a high illiteracy rate during the middle ages and a while after that. But they were using alphabets by then.

[Green Pea]

I would argue that the average person NEEDS up to 5,000 characters in his daily life. I can't emphasize this enough: Chinese words are mostly compounds, 5000 characters means a lot of words. English spelling rules do break down.

Let's try this again using multiplication as an example. What would you rather do to mulitply 2 numbers?

1. Memorise a mulitplication table of 200x100 giving you a finite set of only 20,000 answers?

2. Use a calculator that can multiply NxN for an infinite set of answers with a 95% accuracy?

I can move a mountain with a shovel, too, but I would rather use a bulldozer.

[nnt]

Let's try this again using multiplication as an example. What would you rather do

to mulitply 2 numbers?

1. Memorise a mulitplication table of 200x100 giving you a finite set of only

20,000 answers?

2. Use a calculator that can multiply NxN for an infinite set of answers with a

95% accuracy?

I don't think these examples are relevant for our comparision purpose.

You never memorize a multiplication table of 200x200. A human being would rather memorize 8 "elementary" multiplication tables for 2, 3, ...,9 (Multiplication tables for 1 and 10 are "obvious"). All other multiplications are based upon those "elementary" tables.

The difference between calculator and human calculus is speed, not their principles . A calculator does not memorize an "infinite set" of answers, it memorizes the rules which give the answers, just like a human being.

The main differences between a language like English and Chinese lie in the "unit of combinations" for each language and the rules you follow to make these combinations. (basic word components and grammar)

In English, it's not only the 26 alphabet letters, but also a number of prefixes, suffixes, roots, inflexions: you never learn 200000 "words" but a few thousands of basic words (as combinations of these prefixes, suffixes, roots, inflexions) from which you derive the comprehension of the others (at least those words not borrowed from other languages).

In Chinese, it's a few thousand of basic characters from which you derive the other words. ( For "non-basic" characters, you can often guess the pronuciation and sometimes the meaning from the components of the character).

The problem is just how hard it is to get the basics of a language (basic word components and grammar), and their combinations, and to keep this knowledge.

[sm_sung]

Ala wrote:

Other examples include: 投资 (to invest or to put in capital?)、计算机 (machine that computes or computer?)、精神病 (psychosis, psychological disorder, neurosis, folie, delirium, psychoneurosis, or mental disorder?). Although the English word computer derives from the verb to compute, the English term is far more capable of abstraction than the three in-your-face characters of 计算机. This muddling of zi and ci (and the lack of spacing as result) gives me great unsatisfaction in using Chinese to write anything highly technical (such as in pathological medicine) AND abstract (such as in the social sciences).

This phenomenon exists in other languages (like English) as well:

Television- Could it be something that extends one’s normal range of vision(like binoculars)?

Photocopier- Could it be something specifically designed to copy photos (photographs)?

Light bulb – Could it be a bulb that is light(not heavy)?

Generator – Why must it generate electricity?

Computer – Someone or something which computes?

Taibei wrote:

And just what is that sound?

That sound is one syllable out of a set of only 400 syllables.

This is often asserted -- without any proof. Moreover, if we're talking about time here, how much more time does it take to become literate in Chinese characters than in pinyin?

I don’t know. Do you have any prove that it takes less time to become literate in pinyin? And by literate I mean being able to read and understand pinyin at an equal level of proficiency to reading and understanding characters.

How do people know how to pronounce the phonetic compenent of a character? Only about one-quarter of characters supply accurate phonetic information (proper pronunciation, including tone). And even in those few cases it's not always clear which element is the phonetic one (because characters are not regular in this either).

This isn’t easy to explain but here goes. Once you know the meaning of a character a rhyme is all you need to remember the appropriate syllable. The syllable and the meaning of the character then allow you to figure out the tone quite easily, assuming you already know the character in its spoken form. Kinda like English. You don’t actually analyse every word to figure out its pronunciation when you read. You just know it. Also, remember that there are just about 3000 characters in common use. Reading the same character repeatedly reinforces your memory of how it’s pronounced.

(Could someone (Pazu! )help me explain this one in a clearer way?)

1-A writing system that is an especially ill match for the sounds of the spoken language.

2-A writing system that is so difficult it takes many more years of effort to learn than alphabetic systems -- and even then it is learned incompletely, even by its finest scholars.

3-A writing system that is so difficult it has hindered the literacy of the largest single group of people on the planet.

1-As mentioned above, there aren't that many characters, so it isn't that bad.

2-I don't quite agree with this. Aren't the number of schooling years almost the same in character-using and alphabet-using countries?

3-It probably has more to do with wenyanwen and poverty. Isn't it amusing that in the past 50 years it suddenly became much less of a hindrance to illiteracy(e.g HK and Taiwan)?

[pazu]

How do people know how to pronounce the phonetic compenent of a character? Only about one-quarter of characters supply accurate phonetic information

(proper pronunciation' date=' including tone). And even in those few cases it's not always clear which element is the phonetic one (because characters are not

regular in this either). [/quote']

This isn’t easy to explain but here goes. Once you know the meaning of a character a rhyme is all you need to remember the appropriate syllable. The

syllable and the meaning of the character then allow you to figure out the tone quite easily, assuming you already know the character in its spoken form.

Kinda like English. You don’t actually analyse every word to figure out its pronunciation when you read. You just know it. Also, remember that there are just

about 3000 characters in common use. Reading the same character repeatedly reinforces your memory of how it’s pronounced.

(Could someone (Pazu! )help me explain this one in a clearer way?)

I'll try my best.

Most Chinese characters are 形聲字 (well... how to translate this word? Something like "phonetic character", which means borrowing some existing

characters to form the phonetic part of a new character), it's reasonable to say that the sound of a character can be deduced in some way.

Let me give some more background of this. Chinese linguistics summarized the rules to make a new Chinese characters as Six Principles (六書, liùshū), I

just explained 形聲 here. For some reasons they have already made a character 嬰 to represent the sound "yīng" and it means "baby". Coincidentally the

sound matches "cherry", so a convenient way to make a new character for cherry is to use "嬰" to represent it. And because cherry is a plant, to avoid

confusion of cherry and baby, they put a 木 on the left hand side. So here's is cherry (櫻), the left part is an indication of the nature of the subject, and the

right part is for phonetic purpose only. Some other new characters formed in this ways are found in chemistry too: oxygen (氧: yăng) is a radical representing

"chemical" (气) plus a phonetic part (羊: yáng).

This is a very convenient and effective way to make new Chinese characters, and indeed most Chinese characters we use are 形聲字 (I can give you some

hard facts but I'm too lazy to check on the internet).

Okay, enough for the background. Now, because most of the characters are 形聲字, and it's quite easy to figure out the radical part, so the remaining is

assumed as the phonetic part.

Let's see some more examples:

羅 / 罗: luó

蘿 / 萝 : luó

The first two are common characters and I think most Chinese know them. Then here came another character: 桫欏雙樹 / 桫椤双树 (the twins trees where

Buddha attained Nirvana in Kushinagar).

Even though it's the first time you see "欏", most people will guess it as "luó" because of the obvious phonetic part.

And yes, it's pronounced as "luó".

Okay, this is a rule most Chinese rule to guess the pronunciation of a new character. In many cases, it works; but it failed in many situations too.

Use the same example again, 沙 is shā, but 桫 is suō, so the phonetic part is not accurate in this case, and the only way to check the pronunciation out is to

check your dictionary. But you see in Cantonese, 沙 is /sa1/ and 杪 is /so1/, so at least the consonant is the same, the little connection can indeed help

Chinese to remember the sounds of Chinese characters.

There're something more tricky than this. 扁 is biăn in Mandarin and it serves as the phonetic part of many characters, however the sound changed a little bit

through time, b to p, so 骗 is piàn rather than biàn.

Now imagine you see 翩 for the first time, you have to distinguish the radical and the phonetic part. Usually the radical is on the left hand side, but 扁 cannot

be the radical (oh why? Because in my impression I can't think of any other character with 扁 as the radical), I can deduce that 羽 is the radical instead. So

扁 is the phonetic part. Then I have to decide the sound, is it pian or bian? I think most Cantonese Chinese would choose "pian" because most other 形聲字

with 扁 in Cantonese Chinese starts with "p" rather than "b" (e.g. 騙, 編, 篇). This is however different in Mandarin Chinese (with a mix of b and p).

I agree with Taibei that the pronunciation of a Chinese character cannot be deduced ACCURATELY in many cases, but this is at least how a Chinese

connect a sound to a character, the connection may be blurred by time in history, or just totally confusing, but in my personal experience I found them helpful.

A writing system that is so difficult it has hindered the literacy of the largest single group of people on the planet.

And I want to tell you that, just by connecting the seemingly low literacy rate with Chinese characters is biased. Here's some facts:

Literacy rate in Hong Kong = 94% (CIA World Facts, 2003 est.)

male: 97.1%

female: 90.5%

Another data here:

Population uneducated and kindergarten kids in Hong Kong (2001 Census)= 8.4%

Most of the illiterates in Hong Kong should be those immigrants who came to Hong Kong after the War, the large difference between the literacy rate of male

(97.1%) and female (90.5%) was because of the unfair chances of education of different sexes at that time.

With 8.4% of uneducated population, still has a literacy rate of 94%, I think it's pretty obvious for the real reasons of illiteracy.

While it's obvious that Chinese characters are difficult to learn at later stage, I don't think you can count the kids in this category.

[Quest]

桫欏雙樹 / 桫椤双树

It reminds me of 沙加

[Green Pea]

You never memorize a multiplication table of 200x200. A human being would rather memorize 8 "elementary" multiplication tables for 2, 3, ...,

Yes, you're right. Silly notion isn't it? But that is precisely what we do with characters.

A calculator does not memorize an "infinite set" of answers, it memorizes the rules which give the answers, just like a human being

Bingo! English and other phonetic scripts are "calculators" with rules that generate probable outputs. This is why they are far more efficient and stay more efficient despite the number of inputs whereas Chinese writing begins to sink under its own weight. One only needs 5,000 characters? That's a nice coincidence. I wonder which came first...the need or the limitation of the system?

[nnt]

One only needs 5,000 characters? That's a nice coincidence. I wonder which came first...the need or the limitation of the system?

Just read Pazu's post above about the six ways to invent characters, and you'll see how to memorize a little less ...

Btw, Pazu:

The Nôm characters in your post (after the link to your photos) read:

俺伴过

em thấy bạn lạ quá

meaning: "you're so strange!"

all of them except "quá" are hình thanh tự 形聲字

You should have used "tôi" ( 碎 one variant among 36 others) instead of "em":

碎伴过

tôi thấy bạn lạ quá

If you want to keep "em", you should say:

em thấy anh/chị lạ quá

anh : 嬰 or 英 ("used as" characters)

chị : 姉 (Nôm pronunciation. Hán Việt is tỷ )

Old Vietnamese script is harder than Chinese!!!!

[pazu]

Nnt: indeed I wanted to use "tôi" but I don't know how to write the character. Did you check it in some online dictionary or a paper dictionary? I only know http://www.nomfoundation.org/nomdb/lookup.php

And a note to the character "Quá" (過 / 过, which means "very" in Vietnamese), I don't know the origin of this character but the part "咼" can also served as a phonetic part (聲符).

e.g. 禍 (huò, or wo3 in Cantonese), 媧 (wā), 鍋 (guō), 萵 (wō).

The "g" and "w" may be far-fetched to some people's ear, but I'm sure NNT should have been familiar with the connection already. Because in Vietnamese, "quá" is pronounced as either "G" in the North, or "W" in the South, though written as the same spelling.

[nnt]

tôi = 碎 sui4 (Hán Việt toái )"used as" tôi . It's a Chinese character used phonetically to transcribe a Vietnamese sound (like kun reading in Japanese)

The same for 過 / 过, guo4 (Hán Việt quá ) "used as" qua (through)in đi qua 多过 with 多 duo1 (Hán Việt đa ))"used as" đi = go , walk.

過 is "used as" quá in the sense of "very" after an adjective in Vietnamese.

Otherwise, in Hán Việt (Chinese characters Vietnamese pronunciation), quá 過 (pinyin guo4 )has the same usage as in Chinese: ex.

quá quan trảm tướng 过关斩将

[sm_sung]

Thanks for the explanation pazu! that was basically what I was trying to say.

Here are a couple of things I’d like to add to Pazu’s post:

形聲字 (well... how to translate this word?

I think it’s translated as semantic-phonetic compounds.

According to http://www.omniglot.com/writing/chinese_types.htm around 90% of all existing characters are 形聲字. Taibei and pazu are right in saying they don't always provide 100% accurate phonetic information but the clues they do give are nevertheless valuable.

Green Pea wrote:

Bingo! English and other phonetic scripts are "calculators" with rules that generate probable outputs.

English and other phonetic scripts are not “calculators”. The people who use them are. And as we all know, human calculation is far from perfect.

This is why they are far more efficient and stay more efficient despite the number of inputs whereas Chinese writing begins to sink under its own weight.

You’re making a very strong statement without providing any proof. If phonetic scripts are really “far more efficient” it should be very obvious that characters are inferior and there should be tonnes of evidence available to back up your claims. Where is that evidence?

One only needs 5,000 characters? That's a nice coincidence. I wonder which came first...the need or the limitation of the system?

Basically yes. But some people might need to know a few more. For example, if you are into chemistry you might need to know an additional 100 or so but these are rather easy to learn.

Here’s a slightly technical description:

Characters created to represent chemical elements incorporate both the normal state of the element (at RTP-room temperature and pressure) and also its pronunciation. These are represented by radicals. The 气radical is for gases, 钅 (金) for metals that are solids, and石 for non-metals and metalloids that are solids. The phonetic components, which are then added, are easily recognizable and often indicate the pronunciation to a high degree of accuracy.

Example:

Oxygen Yang3 (氧) -气radical plus 羊 phonetic

Copper Tong2 (铜) - 钅radical plus同 phonetic

There are very few exceptions (3 I think) for mercury, bromine and gold.

As you can see, these newer characters provide accurate semantic (meaning) and phonetic (sound) information. What more could you ask from a character?

[sm_sung]

Add-on to the above post:

Certain simple characters tend to be used as phonetics rather often, like 并，勾（冓），青etc. When you see them appear in more complex characters, you can quite safely bet that they are the phonetics.

[dmoser]

There is an odd quality to some of the posts in this thread. Some of the exchanges are almost Monty-Python-esque (I'm thinking of the famous "dead parrot" skit, in which a pet shop owner keeps coming up with increasingly strained and improbable rationales to cover up the fact that the parrot is, quite obviously, dead.)

It seems to me that Green Pea has already posted a very succinct and accurate characterization of the cognitive processing differences in the two scripts, and elsewhere several people have given very cogent, clear and reasonable appraisals of some of the problems of Chinese characters, and some advantages of alphabetic writing. Nothing presented in these posts has been new or even very controversial.

But it seems that, rather than just acknowledging these few basic problems (none of which are fatal, or entail eliminating characters altogether), people seem compelled to defend the characters against any criticism, no matter how obvious. Simply characterized, the replies sort of fall into the following types:

--the problems are really not so bad as you say

--alphabetic scripts have problems, too!

--the two systems are just different; you can't say which is better

--This is just another "West-good, East-bad" attack

--Well, *I* never had any problems with characters

--Let's change the subject to something else

None of these really accomplish anything, because they keep us from ever arriving at any consensus. And it causes us to go around in circles debating aspects that were never part of the original point.

No one is saying Chinese characters don't ever convey phonetic information.

No one is saying they aren't a useful tool for communication, or denying that they are successfully employed every day by users

No one is saying they should be thrown into the garbage heap [Well, maybe some people HAVE said such things, but the thoughtful posters to this site do not engage in such extremism)

No one is saying alphabetic systems are without flaws

No one is denying that Chinese characters may have some advantages over alphabetic scripts

And so on. It doesn't seem productive to keep going over this ground over and over again.

One obvious thing we criticizers are maintaining is that the cognitive burden of acquiring and maintaining mastery over the character system is harder, in human processing terms, than the more rule-based alphabetic principle. That's it, in a simple nutshell. Yet a considerable amount of cyber-ink is still being spilled over this. And a lot of hand-waving and side-stepping is going on to minimize the fact of this "dead parrot".

It seems to me the character-defender camp might be better off adopting positions like:

Yes, characters are indeed cognitively harder to process, and do result in many social, educational, and information-processing inconveniences;

BUT

They may be easier/faster to read and recognize

They are indisputably more beautiful

They are indeed a source of cultural unity, cohesion, and ethnic identity

The memorization requirement may not be a totally bad thing

They do provide a certain link to the past in a way that the alphabet never can

For better or worse, they are in some ways quite suited to the morphology of the language itself [though there is a chicken-or-egg issue here]

The terseness of expression they provide has indeed caused a greater schism between the written and spoken language than with other scripts, but this also has its wonderful aspects, and has enabled unique and subtle expressive qualities in the written language; Tang poetry wouldn't exist without it

We're used to the characters, and it would be too much of a hassle to switch

They are a noble and fascinating part of the human legacy

Chinese characters can be fun; the possibilities for wordplay are endless

They can take up less physical space on the page than alphabetic scripts

The difference in cognitive processing may provide benefits in children we haven't understood yet

They look great on T-shirts

And so on and so forth. There is much that can be said and debated, even having acknowledged the obvious flaws in the system. It just seems to me that the constant, almost reflexive defense of the characters, even to the extent of denying or ignoring the obvious, is not productive.

Just a suggestion.

David

[nnt]

I think dmoser has forgotten the main reason for defending Chinese characters:

because they are harder, even for Chinese!!!

In the good old past, when the scholars sect ruled, characters were a selection criterium, and the harder they were to learn, more select the scholar club was. It's just an "initiation process" in society : something hard to gain is something which will be defended all the more vigorously...

English has only one "selection criterium": the accent (British Queen English vs Whom-you-know texan accent). Too easy language, too democratic !

French seems more distinguished than English, because of its spellings, its genders, and its grammar (only foreigners know French grammar ), and every French Prime Minister attempt to "simplify" French has always failed.

And I think, as standards of living are rising, traditional characters could make a come back in Mainland China, because less people know them...

[pazu]

David, I think people are defensive of the Chinese characters because some other people here use their own learning experience (as an adult) to doubt the practicalities of Chinese characters, and blamed it for the major reason behind literacy rate, and they just ignored the acquiring abilities and cognitive processing of children. And this is why peope have to tell others that: the problems are really not so bad as others have said.

I agree with you that the points aren't controversial here, and Chinese characters will be used for the future, so people are defending some concepts (but not the future) of Chinese characters.

And I think, as standards of living are rising, traditional characters could make a come back in Mainland China, because less people know them...

Nnt: Can't agree with your point.

[nnt]

Pazu:

What I've written is not an opinion, just a Devil's advocate observation...