Interesting Article on the Difficulty of Chinese

Language Learning Chinese
41

I have some years’ background in Korean, and now some years in Chinese. As far as spoken Korean is concerned, many never get past a low intermediate level: The different honorific forms (Japanese doesn’t compare) that brings a new meaning to fear as a beginning learner of language, the backward grammar (not that different from Japanese), and the mix of pure Korean and Chinese-rooted vocabulary (I guess somewhat like Japanese) are serious, serious obstacles. Mandarin is a far easier spoken language. But if you get past this low intermediate level in Korean, I would say spoken Korean and Mandarin are about equal. I’m having trouble in Mandarin being able to express complicated ideas that involve stringing together many different thoughts- my Mandarin lacks depth. And struggling through this process is not much different than what I went through with Korean.

As far as reading comprehension is concerned, though, Chinese takes the top spot by far! I was definitely farther along in Korean with the same investment in time and effort.

42

Just my five cents worth, but I don’t find Chinese as difficult as the article makes out. Sure it’s hard. Harder than French even, but to me a lot of the pronounciation sounds are similiar to Afrikaans or even other African languages. The tones can be a bit of a bugger, but I’ve found that people seem to understand you even though your tones aren’t perfect.

As far as the writing is concerned. Well, I’ve been studying by myself for about six months. The characters I’ve learnt I remember. And that is to say, as long as I read them regularly and continue to practice my writing.

I’ve found that in a way it’s a lot like your native language. In that when you start to learn to write, you write all the time in said language and it just becomes second nature. It’s very much the same with Chinese. I just try to write notes to myself in Chinese rather than English and it just gets easier. Sure you make frequent mistakes but the same thing happened when we all started to learn to write in grade one, right?

43

[quote]rectifying the script would be incredibly simple. for example, assign every “ma” the same phonetic half- for conversation’s sake let it be “horse”.

“linen” would use “horse” phonetic + “grass” determinant while
“mother” would stay exactly the same as it is. [/quote]

I’ve always thought that this would be a great solution compromise.

If you take one phonetic determinaant for each syllable (regardless of tone), you get by with just a little over 400 characters (assuming fo course that you knew the words in their spoken form). Children would be taught these first, and when writing words they didn’t know, could right the phonetic part only. Adding a tone indicator would signify that they were writing phonetically.

Brian

44

He’s right, basically, but didn’t present a balanced view.
He didn’t mention the one benefit of Chinese characters, though he gave a good example:

It is precisely when guessing unknown words that Chinese characters are useful - and more useful than an alphabet. Of course someone who doesn’t know what phlebitis is in English is not going to be able to guess its meaning from the form of the word. You could guess the meaning of the Chinese word for phlebitis from its form, though – or at least get close.
There are lots of more common examples: no student of English could guess the meaning of words like “tuna,” “chipmunk,” “cedar,” etc.; a student of Chinese will certainly be able to tell that the equivalent Chinese words are “some kind of fish,” “some kind of rat-like animal,” and “some kind of tree.” And in many cases, just knowing this is enough - just like I, a native speaker of English, recognize that “red snapper” is a kind of fish, but actually would not be able to recognize one on a dish, in a store, or in the wild. A deeper knowledge of the word has never been necessary for me.
The point of reading is understanding the meaning. To do this, a perfect phonetic representation of the sound of the words is not necessary. Characters help with guessing the meaning of unknown vocabulary. Words written with an alphabet do not.

45

Bingo. Property #2 (see above), it encodes rich semantic information in each character.

Of course you can do the same thing in an alphabetic language, such as by latin roots or by compound word construction. However, we don’t do this very often, why? Because it is not dense enough in speech. Latin-root words are generally longer and compound words can get very long (such as in German). To make each unit of meaning shorter would imply an increase of homophones, which in an alphabetic language cannot* be distinguished even when written down.

  • Of course there are homophones that are spelled differently (read, reed, Reid, etc.), but you have only so many degrees of freedom to play with. The only way to increase the degrees of freedom while constraining written length is to increase the alphabet size, which is effecively what Chinese has done.
    This also accounts for the basic problem in Korea and Japan in getting rid of characters: they’ve borrowed short homophonic Chinese word pronunciations along with characters with rich semantic information. Now, replacing characters with an alphabet and what do you get? Lots of ambiguity.
    One more thing it accounts for: the divergence of Chinese speech and writing styles. In speech, you implicitly have to get around homophones by redundancy (which is what “colloquial” really is), but in writing, characters are really powerful expressions and so you can write very concisely and expressively. Classical Chinese is an extreme case of this divergence.
46

Anuther problum in reeplaacing characters with an alfabet, iz that I hav ofen shoon pin-yin sentencez or paragrafs, with tone marks, to maanland Chiineez (hoo wer very profishent with pin-yin) and they wood need too “studee” the pees for a minut or too beefor they wer aabl too understand enuf meening. The problum uzulee that by not imeedeeatlee recognizing the meening of the soundz, they wood not noo the natuural intoonaashen and porzing and this wood slow them riit doun. (Hav yoo ever red sumthing and not been aabl too rekognize a word, and hav to ree-reed the sentens until yoo cood ges wich word it was? I ges it’s sumthing liik that.)

If we want to change the Chineez langwige, maybe we shood look at ading, let’s say, about another 1,000 silabl posibilitees, this wood maabee proviid enuf vaireeaashen in speling to alow us to rob them of thair culture and impooz an alfabet sistum on them.

47

Don’t write stupid like that, it hurts my brain. But I got the point.

48

GCZ, thanks for making vividly my point that phonetic elements of even an alphabetic language are not used when an adult is reading normally. It’s all just shape recognition. When the shapes changed, as in GCZ’s post, then you have to use the phonetic aspect and as you can see, you cannot possibly be doing that in normal reading due to the much higher speed at which you normally scan.

49

IMO, this doesn’t prove anything. One could easily conclude that the reason an educated native user of the language is slow at reading paragraph length discourse written in phonetic symbols is only because it is not his habit. He is only used to using phonetic symbols for words and syllables and characters for anything longer.

Because English intonation and stress vary more than that of a language like Chinese, MA TESOL/Applied Linguistics students often learn to read and write IPA for language samples longer than word length. Sure, some of us are better at it than others, but if I can read paragraph length discourse rendered in IPA in a language where phonemes change quite a lot depending on what came before and after the phoneme, then I don’t think the argument against writing Chinese in tone marked pinyin is especially strong. If it becomes people’s habit, then it wouldn’t be very hard at all in my opinion.

50

I was talking about something else, but okay.

51

I believe in evolution. How did written languages start? Cuniform (speelin?) and hieroglyphs, ie. symbolic. What are the vast majority today? Phonetic. Actually besides Chinese, I can’t think of another language which relies on a symbolic form for “everyday” use if you will. I’m sure there a a few more examples, but regardless the vast vasy majority are phonetic, and there are reasons why there was a move to phonetic alphabets.

  • Easier to learn
  • Easier to adapt words from other languages
  • Easier to create new words (If the previously given point in favour of symbolic’s ability to create new characters, why are so few created and the vast majority just borrowed for the phonetic component?)

I like characters just as much as the next guy, but I think a more Japanese approach, where one can communicate entirely in a phonetic language but can also use characters, is preferable. Now of course Japanese has a completely different structure than Mandarin, but I’m sure there’s a way.

52

Why not use both characters and an alphabet at the same time? That’s what they do with children’s books - bo po mo fo alphabet symbols on the side of the characters. Why can’t they do that with adult books, too? Too space consuming? It would be a great help, as it’s much easier to look up unfamiliar words in a dictionary using bo po mo fo than figuring out character components.

53

Apparently during the “Great Leap Forward” or the “Cultural Revolution” Mao wanted to abolish characters entirely and replace them with pinyin or a “suitable” romanised system. However, fortunately they managed to convince him that it was a bad idea. I think it would have been very sad if they had done that. In fact, I feel that the characters are part of the appeal of studying the language.

54

For an interesting review paper on how people read English, see

microsoft.com/typography/ctf … ition.aspx

55

Good article. One should note that there is no contradiction between word shape recognition vs. letter shape activation as general concepts. I do think “word shape recognition” as it is defined in the article is truly atrocious. It is all-or-nothing or outline-only, neither are what I consider to be “word shape.” If you saw a square, then you saw a square slightly bent out of shape, those two shapes are different, and you haven’t seen the exact “bent square” before, but you still recognize it much faster than some random lines or some random lines in a perfect square outline. One could make the point that it is precisely how it works with words, too, that conventional non-words like “mave” are so much closer in shape (including strokes) to a whole bunch of previously seen word shapes (“maze” “made” “save”, etc.) than unconventional non-words like “amve.”

There is a flaw here per what I said. Adams is assuming you are “unfamiliar” with the pseudoword shape because you have never seen the exact sequence of letters of the pseudoword. But you HAVE seen partial shapes of the pseudoword, so of course it is also going to be familiar.

I also somehow highly doubt the unit of primary recognition is letters and not something less well delimited. Why should qu- activate at the letter level? I don’t think so. Just compare quxxx, qaxxx, cuxxx, caxxx. For the sake of argument let’s say caxxx words are more common than quxxx words than cuxxx words than qaxxx words. The scanning speed from fast to slow will probably then be

caxxx
quxxx
cuxxx
qaxxx

where single letter recognition would predict a lot more cxxxx words than qxxxx words and perhaps more xaxxx words than xuxxx words, so caxxx should come out on top and quxxx should come out at the bottom. This isn’t going to be true, since “qu” is one unit shape. It is not read in units of letters.

On the other hand, the entire article does suggest everything is shape based and that phonetics do not play much of a role. The very fact that “mave” and “amve” are recognized at different speeds is testament to this since “mave” and “amve” not only have the same number of sound-expressing letters but the same number of “syllables” also.

In light of that, I would say that characters are quite nice. “Rally” may mean nothing to you, could mean as little as “mave” for the purpose of illustration, but “造势” as invented by the Taiwanese definitely means something. You get a lot more second chances by activating partial meanings from individual characters (partial words) you do recognize from other contexts.

56

For the most part once we have learned to recgnise words, we do not generally speaking use our knowledge of phonetics to read words. Thus when we come across misspelled words, we pick up the general shape or letter combination e.g. “He was walknig down the spreet.” A native speaker is not suddenly thrown into confusion by the alteration of letters, we rely on recognition of the group of letters, as well as our knowledge of grammar to know what these words should be. It would not think of “swimming pool” or “lasagne”.

Why do people someitmes, add, change, omit, words when they are reading? Because they are not paying attention to the details of every word, (or even every word) when they read, and so they sometimes miss, add, omit according to what they do see, or what the brain is anticipating will come next. Therefore I can write, " He clmbed up the _______." and your brain will supply a host of possibilities based on what it already knows.

When we see words we quickly identify the possibilities and then focus on the defining factors. e.g. “The problem was caused by an interpolation of ideas.” or “The problem was caused by an interpretation of ideas.” When you see “interpolation” and “interpretation” your brain very quickly recognises the “inter” combination, very quickly recognises the “tion” combination, but spends a fraction longer on the middle of the word to select the correct possibility.

Next we know that some letter combinations ore not possible in our native language. Do you know all the words in the English language? Would you bother looking up “sprundane” in a dictionary to find it’s meaning? Probably. Would you look up “stpundane” in a dictionary? Probably not. (Although “sprundane” is probably not a word, it could be, However, “stpundane” could not be an English word.)

The Chinese character system is not as random as some would have us believe. Chinese people know that certain combinations of elements are just not possible. And although the number of elements is overwhelming to a beginner, they are definitely fixed. They know what to look for to distinguish characters. Like us, they pick up on the familiar elements very quickly and then focus in on the parts of characters that will define a word. For example a Chinese person would very quickly see 青 in these following characters and then focus on what determines one character from the next character. 菁 清 蜻 請. In a very similar way in which we read.

The key is to become familiar with the smaller elements and how they combione together. Then you will start to see the patterns that reoccur.

This takes patience and time, as well as knowing where to look.
Reading and Writing Chinese, McNaughton, Wenlin and www.zhongwen.com are good places to start.

Seriously, How long does it take to teach a Chinese kid with no prior alphabet experience to master the alphabet? How much longer does it takes them to master the phonics? (which in reality is just teaching them to recognise the different possibilities of letter combinations - OK it’s a little more than that, but the principles are very similar to learning the combinations that make up characters)

57

One thing I noticed while studying a bit of Bahasa recently was that the words stuck in my head - I could remember them easily. Why? Because they are multisyllabic real words, not just the same three dozen monosyllables thrown together at random. Seriously, that is a major stumbling block with Chinese - the words all sound the same. Endless variations of “shi” and “men” and “mao” and etc., it’s hard to tell many of the bisyllabic (and 90% of Chinese words in the spoken language are bisyllabic) combinations of the same 3 dozen monosyllables apart. Different Chinese words are so phonetically similar to each other that they get confusing - “Oh, is it jingchang or jinchang or jincang or jingcang or jinzhang or jingzhang or…?” you get the picture. Chinese’s limited phonetic palette is a major stumbling block, and using tones is a poor substitute for phonetic variety.

58

That’s more a problem with the Taiwanese accent. Normally chang/cang/zhang are easily distinguishable.

59

Tones are a phonetic variety. They also mean something, you know, just as much as a syllable does. You just aren’t listening for them. You need to remember words with their (some say) “musical” elements. If you didn’t care about tones in real music, then every piece of music sounds the same, too. How can you tell them apart?

60

That’s more a problem with the Taiwanese accent. Normally chang/cang/zhang are easily distinguishable.[/quote]

It’s more than that. Words in English are more easily distinguishable from other English words than Chinese words are from other Chinese words, because of greater phonetic variety. Take a random English word - “beauty” - and imagine how confusing it would be if there were several dozen other English words extremely similar sounding, such as “beaufy” and “beausy” and “beauta” and “beauchy” etc. It is harder for Chinese words to stick into the memory because of limited phonetic palette. The Chinese language uses tones to try and rectify this problem, but most non-tonal language speakers have a difficult time distinguishing between tones.

Tones are cumbersome. I’ve noticed that Chinese speak more slowly in their native language than most other language speakers, in say Spanish (Speedy Gonzalez!) or English. I have a theory that the reason is that before Chinese speak a word, they have to consider what tone they should say it in, which takes a few microseconds of thought, and thus they have to measure their words out slowly in conversation. Sometimes when people are angry or excited they shout out Chinese rapid-fire but since they are shouting they aren’t caring if they butcher the tones.