Shape-based computer input?

[Joar]

Hello!

I have started to very casually learn Chinese during my free time. I have also decided to focus mainly on learning how to read and write, and not worry about learning how all the words are pronounced until later on. For those who care to know why, here is the reason:

I am basically learning Chinese just for fun, and so my main contact with Chinese is likely to be in the form of reading articles, documents and maybe books as well as participating in discussions the internet. All of those activities only require knowing written Chinese. In fact, the only time it would even be useful for me to know spoken Mandarin or Cantonese would be if I went to a Chinese-speaking country, which is unlikely to become a regular recurring thing. Thus, I could in principle get access to (most of) the platform of communication and information that is the Chinese-speaking world without actually knowing how Chinese words are read out loud.

 

In addition to this, I feel that I am naturally more able to memorize shapes and images than precise sounds. I have only begun to learn the characters and have so far only memorized ~100, but so far it feels very easy. If I focus on only (or at least mainly) the characters, I reckon that I should relatively quickly be able to get up to the level of at least basic conversational proficiency. To learn how to speak, however, I would have to get over the fairly steep learning-curve of Chinese pronunciation and the fact that a lot of words sound very similar.

 

So, in short, by focusing on learning written Chinese first, I would get to a level where it is useful significantly faster and at the same time get almost the same utility as if I had learnt both at the same time.

 

The problem is that I still need a way to input the characters to a computer, without necessarily knowing the pronunciation (as you need for pinyin). What I need is therefore a shape-based input method, preferably suited for simplified Chinese. Is there anything that you can recommend? I am interested in Cangjie, but it is difficult to find instructions in English for how to use it. Also, even though there is apparently a version of Cangjie for simplified Chinese, I have not been able to find where to get it or how to set it up.

Can anyone show me where to find Cangjie for simplified Chinese, and/or basic instructions for how to use it in English? Or alternatively, is there any other non-phonetic input method that you would recommend over Cangjie? If I can't find anything practical, I guess I will just have to input the characters using a tablet until I learn the pronunciation of everything.

[Shelley]

I have been learning chinese for pleasure for the last 30 years and never have or will ever go to china but i still learnt how to pronounce characters/words. I think you should learn the pronunciation even if you don't think you will use it for conversation it is very useful for more than this.

 

I can not understand how you can separate the sound from the character, it must be very complicated, much simpler to have a simple sound attached to each one, also some characters have more than one meaning and more than one sound or the same sound and different meanings.

 

I urge you to try learning the pronunciation.

 

You might like to try using a great app called HelloChinese - LearnChinese it will help you learn reading, writing, listening, and speaking. It is free, ad free and no in app purchases. 

 

You can find it in the app store or the Playstore. Have a look here http://www.hellochinese.cc/

 

P.S. There are a few shape based input methods https://en.wikipedia.org/wiki/Wubi_method

 

and there is a list here https://en.wikipedia.org/wiki/Chinese_input_methods_for_computers

[歐博思]

Because I'm learning Taiwanese in addition to Mandarin, both with very different pronunciations, I tried a few different shape-based IMEs. Among them, only Wubi appeared somewhat systematical and easy to learn. Nowadays, I'm back to using Sougou pinyin which is based on pronunciation in standard Mandarin—even sometimes while typing Taiwanese.

 

edit: I'm in agreement with Shelley about the importance of studying pronunciation, but initially I couldn't think of a practical reason until just now—other than it's just what I would do—with my reasoning being that it's not rare to mis-type a character as another character with the same pinyin, and if you're aware of this fact and know the pronunciations, then you can decipher the meaning via context. 四不四优点倒立？

 

Plus, playing with similar sounds in Chinese can be kinda fun... oh the chuckles we have in class when the vocab word is 鼻毛

Edited March 26, 2016 at 04:42 PM by 歐博思

[stapler]

My 2 cents: given that Chinese characters are phonetic (though not completely or to the same extent as alphabets) part of learning them is learning the sounds. That is to say, it makes remembering that much easier if you're aware of the built in phonetics. Example: 愧 魁 餽 - they all have 鬼 as a phonetic marker but their meanings are completely unrelated. Having a phonetic marker makes remembering half the character (and then often the complete character) easier. "Oh it's kui so it has the 鬼)

What you propose sounds the same as if someone told me: I'm going to learn to read and write English without ever hearing it or knowing what sounds the symbols represent. This would turn remembering words into a hugely laborious task without being able to fall back on "spelling them out" to help you write them.

[Joar]

Thank you for your input.

 

Ok, so first let me explain my unorthodox take on learning a bit further:

 

……………………………………………………….

 

Chinese is special (unique?) in that you (normally) cannot infer the pronunciation of a Chinese word from its character, nor can you infer the look of its characters from its pronunciation. No other language that I know about works like this. Knowing that “fish” is written 鱼 doesn’t tell you how “fish” is pronounced, and knowing that it is pronounced “yú” doesn’t tell you how it is written. This is further exemplified by the fact that Cantonese, Mandarin Chinese and the other varieties of Chinese are able to share the same writing system while still sounding entirely different to the point of not being mutually intelligible. As I have understood it, some other countries (such as, I believe, Japan and Korea) that have separate languages and their own systems of writing also sometimes substitute words for Chinese characters (to save space, I presume). All of this works, because there is mostly no necessary connection between the look of the characters and their pronunciation. Does 日本 read the same in Mandarin, Cantonese, Japanese and Korean? In a way, written and spoken Chinese could almost be seen as being like two different languages.

 

So how am I able to learn the characters without any pronunciation? By simply seeing the characters as images (which is kind of what they are anyway). 日 is, sort of, a picture of a sun. 鱼 is, if you have a bit of imagination, a picture of a fish. Likewise, it is possible to learn more complicated characters as a combination of meaning. Putting together “sun” thrice into a single character, 晶, gives you “sparkly”. Kind of like the game doodle god, if you are familiar with it. Maybe this sounds complicated, but so far this approach has allowed me to learn new characters (excluding stroke order) in a matter of seconds. Then how to you read 晶? Rì-rì-rì? No, apparently it’s “jīng”. To me, jīng is just a sound. Unlike the characters, I cannot attach it to any wider context or associate it to anything else. To learn it, I would have to repeat over and over again that “jīng is sparkly” until by brain accepts it as truth. Forcing myself to do this for every new character that I encounter, instead of just making a mental note of the character and moving on, would slow me down substantially. This is why I have decided to not really worry about learning the pronunciation of every character I know before learning even more characters. It’s really mostly not even about not learning the pronunciation, it’s just that the pronunciation is going to lag behind. I currently know some 100+ Chinese characters, but only the pronunciation of maybe 10-20 of them.

 

……………………………………………………….

 

So, back to the keyboard layouts.

 

I have already studied the Wikipedia articles linked by Shelley quite extensively and am aware of the various options that exist. The problem is that I can’t find any instructions in English for how to use any of them, with the exception of a really, really old document about Wubi on Yale’s website that unfortunately seems to have been taken down now. Canjie codes are often available on Wiktionary, but that is hardy sufficient guidance. I have also not been able to find how to set up Canjie with simplified characters (which is says on Wikipedia should be possible). If anyone could help me with either of these things, that would be very helpful.

 

It is possible to connect a phone or tablet computer to a desktop computer and input Chinese characters to the desktop computer by writing them on the touch screen of the phone or tablet. If I cannot find documentation of Wubi or Canjie, it is leaning towards that being what I would have to do.

 

I’m going to look into HelloChinese – LearnChinese.

[LiMo]

I'm afraid I don't know any shape inputs for PC.

 

On a side note, I'm going to have to agree with everyone else. I'm doing some research on Chinese internet fora and there's a lot of slang and pinyin that's used for fun and to avoid censorship. If you want to participate in internet discussions then knowing the pronunciation really helps because a lot of the slang uses homophones. I suppose the key is to remember to pay attention to the pinyin even if you don't put any effort into actually learning to pronounce them, that should work fine.

 

Good luck!  

[Luxi]

Not to mention Chinese radio, audio books, podcasts,  TV and films. At some point you may become aware that there is a fantastic amount of audio and visual Chinese stuff freely available out there, and at least some of it is really interesting, at least for anybody who may already have an interest in learning to read Chinese. Most of it doesn't have English etc. subtitles, and the Chinese subtitles are hard to read in streaming TV. 

 

Then what? Are you going to go back and re-learn the pronunciation of the several hundred of characters you may have already learnt to recognise?

[Kelvin 十六]

to: Joar.

 

I appreciate very much your interest to learn and type chinese characters by bypassing learning pinyin.

And very true you need to find a shape based chinese input method.

 

** 1 Wubi (五笔字型输入法) Input Method **

If you learn mostly Simplifed Chinese, you need wubi input method IME - that mostly available for ms Windows platform

and also available for Android Mobil phone.

 

Wubi input method is not very easy to learn. But learning it will make you in the end capable to inputting chinese very fast.

** 2. FourCorner (四角号码) Input Method **

And also agree with you that pronunciation is not a must to learn Chinese Character (ChCar). With computer technology advance, we can type 'horse' in our computer and go '马' which is the ChCar of horse. Even we can associated each ChCar with its english-word counterparts.

 

Other method to input chinese character using its shape characteristic is by using FourCorner method. This method sees each ChCar as composed of 4 corners - topleft topright bottomleft and bottomright corner. From shape of its corner you can code the character as number from 0000 to 9999.

 

to Other:

I agree also that learn how to pronounce Chinese Characters is fundamental in learning Chinese as a foreign language.

That for one who want to bypassing pinyin and go directly to learn to write Chinese Characters by using its graphical elements, now computer programs provide means to do that. I hope some day there is one-to-one table associating one chinese characters with one unique english word. And pinyin which is assuming its using is english speaking learner something is difficult for those who pronounce word with more order like german, dutch, latin, malay, and indonesia. For example : '八‘=pinyin ba=english ba=BUT in german and malay it is 'pa'.

 

chinese

大家

pinyin (english)

da jia

german

ta jya

[889]

Yes, the OP's plan is not a good one. In addition to the problems mentioned above, he's going to train himself to see characters, not words, and I don't see that forming the basis for any sort of fluency in Chinese, even if limited to reading only. Characters, like letters, are a useful tool for recording the language, but focus on them exclusively and you'll miss the critical point that the basic unit of the language is the word, not the character. That is, it's mingbai not ming and bai, just as it's not under and stand in English.

Further, we all know what's going to happen with his no-Pinyin approach: his brain will identify characters by some sort of English shorthand. And when he reads, he won't be processing Chinese, but some sort of Pidgin English written in Chinese characters. His brain will never go into language-acquisition gear. Hopeless.

Nonetheless, perhaps someone has developed a Chinese IME in which English shorthands can be used to bring up characters: ON SEA, NRTH CPTL, MID NTN, WAIT WAIT.

[Joar]

Ok, so since my posts have to be approved by moderators for some reason, they show up several hours after I make them. This means that new replies might be made between my posting and the appearance of my post, causing something of a discontinuity in the discussion. Keep that in mind. This reply was posted on Mars 27, ~21:30 GMT.

 

First of all, the purpose of this thread was really not intended to be to discuss my current learning plan. I am well aware of my own preferred learning styles and read quite extensively on neuroscience, in fact especially on the theories of how brains learn, store and retrieve information. I really am perfectly able to set up strategies regarding these things myself. Besides, people who learn how to speak Chinese first and how to write it later are hardly unheard of. Doing it the other way around isn’t all that different.

 

The comment on phonetic slang is interesting, though. Mistypings are also a good point. But again, it's not the case that I won't be learning any pronunciation, it's that I wont be learning all of it straight away. If this type of slang is mostly limited to certain words, I suppose I could prioritize learning the pronunciation of said words early. Is there any place where I can find some sort of catalogue of this form of slang?

 

But back to the main topic: input methods. Isn’t here anyone with more knowledge on this? Does Canjie for simplified characters use different codes, or does it just output a different character?

 

It someone has a saved copy of the (now unavailable) document on Wubi that used to be hosted on Yale’s website, I would be very interested.

[imron]

If you are using simplified then you should be using Wubi and not Cangjie.  Yes it's possible to do simplified with Cangjie and Traditional with Wubi, but they were not designed for using the other character sets and so you'll just find yourself hitting all sorts of edge cases and annoyances.

 

The 'Yale' document about learning Wubi can now be found online here (not sure if this is currently the canonical source, but it's the first I could find).  It might be old, but Wubi was developed in 1985 and the main rules haven't changed since then - there was a failed attempt at introducing some different root shapes on different keys, but it never took hold and so all later versions still use the 1985 keys.

 

There are also plenty of threads on this forum discussing Wubi (including for example why it's not so good for traditional - and similar arguments will apply to cangjie and simplified), so it's probably worth doing a quick search there as well.

 

For doing actual typing with Wubi, then for Windows I would recommend using 极点五笔, and for OS X I recommend WBIM.

 

Note, if you do not know in the region of 1,500-2,000 characters, then learning Wubi will be quite difficult and slow.   You can download a typing tutor for Windows here (not sure if there is an equivalent for OS X), note however that most (all?) typing tutor software is aimed at native Chinese speakers, so if you can't read Chinese then using them will also be difficult.

[imron]

his brain will identify characters by some sort of English shorthand. And when he reads, he won't be processing Chinese, but some sort of Pidgin English written in Chinese characters. His brain will never go into language-acquisition gear.

This is actually going to be a big problem, the OP just doesn't realise it yet.

[stapler]

Chinese is special (unique?) in that you (normally) cannot infer the pronunciation of a Chinese word from its character, nor can you infer the look of its characters from its pronunciation. No other language that I know about works like this. Knowing that “fish” is written 鱼 doesn’t tell you how “fish” is pronounced, and knowing that it is pronounced “yú” doesn’t tell you how it is written

 

I disagree. I agree with DeFrancis that Chinese can be thought of as an inaccurate/complex syllabary. This is to say, you can infer the pronunciation of Chinese characters from their shape (just you won't be as successful as you would with the English alphabet - which you can also be lead astray (river Thames anyone?)) Knowing that fish is written as 鱼 does tell you it's pronounced yu2 in the exact same way the two letters and single number "yu2" also tell you how to say it. The difference is the pinyin script is 100% accurate where as the sinogram can be quite a bit more vague or provide no hint at all. Even if I didn't know how to say 蘇 or 魯 I might be able to guess they sound like "yu2" (which they do in a vague way) - though at other times it will be no use at all (as with 鮮)

 

 

 

This is further exemplified by the fact that Cantonese, Mandarin Chinese and the other varieties of Chinese are able to share the same writing system while still sounding entirely different to the point of not being mutually intelligible. As I have understood it, some other countries (such as, I believe, Japan and Korea) that have separate languages and their own systems of writing also sometimes substitute words for Chinese characters (to save space, I presume). All of this works, because there is mostly no necessary connection between the look of the characters and their pronunciation.

 

This just shows different languages can use the same script to mark different sounds. I could learn how to write English using Chinese characters if I wanted. But why would want to learn a script that is functionally useless (other English speakers wouldn't understand my marks)? In fact this sounds like what you're proposing to do. Or in short: if you accept my contention that Chinese characters and the Latin alphabet are both just systems to mark speech sounds - why would you want to learn a system of markings that isn't linked to any language (just learning shapes arbitrarily) or one no other users of that language recognise (learning Chinese characters with English meanings)?

[Luxi]

It is possible to connect a phone or tablet computer to a desktop computer and input Chinese characters to the desktop computer by writing them on the touch screen of the phone or tablet. If I cannot find documentation of Wubi or Canjie, it is leaning towards that being what I would have to do.

 

 

@Joar, that sounds terribly complicated and quite unnecessary. If you have Windows 7 or later, you can input Chinese characters using hand writing with the mouse. I imagine the same is possible with Macs IOS and Linux, though I never tried. You could also use some of the online Chinese dictionaries that allow searches by handwriting. Google Translate has a very good handwriting input option.

 

In the Windows environment, it is easier with Windows 10 than earlier, as it is easier to switch between different languages and between different Chinese input methods. 

[889]

It's awkward to write with a mouse on a PC. Just get a stylus and drawing tablet, instead.

[imron]

I imagine the same is possible with Macs IOS

With Macs, if you have a laptop you can do Chinese handwriting input on the trackpad.  On iOS, there is a built in Chinese handwriting support that you can just enable in your iPhone/iPad language settings.

[Kelvin 十六]

instead of using jidian wubi from http://www.freewb.org/ , I use sogou wubi from http://wubi.sogou.com/ .

it can [also] input wubi and pinyin simultaneously.

for me inputting by wubi together with pinyin is more easy, since my wubi skill is still developing.

 

[imron]

Kelvin, I've removed a link to the site you linked to.  If you want to promote your own products, please make sure you disclose your connection to them.  Also read our guidelines for commercial posters.

 

For Chinese IMEs, Sogou is obnoxious.  I have no wish to see popup ads on my desktop.  Last time I tried it, its Wubi dictionary wasn't as good as 极点五笔's either.   极点五笔 also supports mixed Wubi and Pinyin.

[Joar]

It is possible to connect a phone or tablet computer to a desktop computer and input Chinese characters to the desktop computer by writing them on the touch screen of the phone or tablet. If I cannot find documentation of Wubi or Canjie, it is leaning towards that being what I would have to do.

 

That sounds terribly complicated and quite unnecessary. If you have Windows 7 or later, you can input Chinese characters using hand writing with the mouse.

 

That might be easier to set up, but it also sounds very unergonomic.

 

I agree with DeFrancis that Chinese can be thought of as an inaccurate/complex syllabary.

This just shows different languages can use the same script to mark different sounds.

Or in short: if you accept my contention that Chinese characters and the Latin alphabet are both just systems to mark speech sounds - why would you want to learn a system of markings that isn't linked to any language?

 

So you mean that "鱼" means the sound "yú", which in turn means "fish"? This sounds a bit counter-intuitive to me, considering that (and do correct me if I am wrong) 鱼 means "fish" in all languages that use Chinese symbols for writing, even though it would sound different when read out loud. What I have done is instead to connect "鱼" directly to the meaning "fish" (by which I mean the concept of "fish", ie the mental images and associations, rather than the English word).

 

 

If you are using simplified then you should be using Wubi and not Cangjie.  Yes it's possible to do simplified with Cangjie and Traditional with Wubi, but they were not designed for using the other character sets and so you'll just find yourself hitting all sorts of edge cases and annoyances.

 

The 'Yale' document about learning Wubi can now be found online here (not sure if this is currently the canonical source, but it's the first I could find).  It might be old, but Wubi was developed in 1985 and the main rules haven't changed since then - there was a failed attempt at introducing some different root shapes on different keys, but it never took hold and so all later versions still use the 1985 keys.

 

There are also plenty of threads on this forum discussing Wubi (including for example why it's not so good for traditional - and similar arguments will apply to cangjie and simplified), so it's probably worth doing a quick search there as well.

 

For doing actual typing with Wubi, then for Windows I would recommend using 极点五笔, and for OS X I recommend WBIM.

 

Note, if you do not know in the region of 1,500-2,000 characters, then learning Wubi will be quite difficult and slow.   You can download a typing tutor for Windows here (not sure if there is an equivalent for OS X), note however that most (all?) typing tutor software is aimed at native Chinese speakers, so if you can't read Chinese then using them will also be difficult.

 

Thank you very much, I will have a look at this!

[LiMo]

I don't want to take the thread off topic but it's been going for a while so I'll just throw caution to the winds.

 

As per examples given by Stapler, the phonetic nature of Chinese is undeniable and is, if you want to learn it well, crucial. It may seem that Chinese is non-phonetic and essentially random now, but once you get to a high enough level patterns begin to emerge in both phonetic and semantic radicals which are a key to having "good Chinese." (I suppose you could get used to recognising these on your own terms)

 

As far as I know 鱼 is fish in all these languages (but it doesn't matter if it isn't!), the issue isn't the semantic content, it's the phonetic content. Let's forget about all those other languages for a moment. As far as Chinese is concerned phonetic radicals are important as well, they give clues to pronunciation and are far from rare. 鱼 itself can be broken down into further components, but taken as a whole it has the phonetic value of yu2. When it makes up a part of other characters it can suggest that that character has a similar pronunciation. It may not. But when you take the Chinese writing system as a whole it's something we'd be foolish to ignore.

 

Not to mention the fact that your personal concept of "fish" probably includes the word "fish," you read 鱼 and chances are you don't get an olfactory or tactile impression or some platonic form of "fish", you most likely think "fish" (in English).

 

我觉得你们都很可爱

 

In your system this sentence probably reads:

 

"I think you're all cute"

 

I doubt it reads

 

"[me/I/mental image of me in the mirror]...[some kind of concept of feeling a certain way about something]..." etc.

 

The fact is you'd have to strongly resist the habit of reading in English as well because you'll be learning it through the medium of English and I think your natural instinct will be to "read" it in English.

 

 

Basically, we are questioning this method because you will probably end up teaching yourself "bad habits" for quick gains in the short term, but once you go back and try to "do it properly" you'll have a much harder time unlearning those bad habits. You can dissociate the script from the spoken language, but the question is should you? What do you gain from it in the long run except the ability to say that once upon a time you were "literate in 3 months." If you're actually planning to ever take this further than online chat (don't forget my comment above) then I doubt the time saved will be worth the hassle later on.

 

Edit: One of the ways I improved my reading was by watching TV shows, seems counter-intuitive but the subs make regular reading practice with minimum effort. Of course this will take time to work up to, the vocab is a challenge, but it's something to keep in mind. It will also lay brilliant groundwork for when you actually come to grapple with the speaking and listening part, you'll be able to casually/unconsciously associate sounds and characters without "actively" learning them. This might counteract any bad habits you have of associating the characters with any non-Chinese sounds that are floating round your head.

 

Edit: I'm actually undecided. I guess maybe we're being conservative old farts, it could be worth a try, just know that it's kind of a risk. You can be the pioneer who risked all to discover new methods of language learning. 加油