What do people think about the idea that the Wubi input method could be helpful for remembering how to write characters by hand?

It seems to have the potential as it uses the structure of characters to type them, so whenever you type you'd be reinforcing your memory of how the characters are written.

Interesting, let us know how it works out for you. Santa just gave me a keyboard overlay so I can type in Chinese; it's not for Wubi, and it seems to be for one of the traditional methods. I'm very annoyed with the pinyin input methods and would love to try something new, but Wubi seems quite complex.

I used a similar input method (Zhengma) with that goal in mind, but found that it doesn't work quite the way you would hope. Because one key can represent several similar but different shapes, you find that there are some details (such as whether a character has a "dot" or not, stroke order, etc.) which using Wubi or Zhengma does not help reinforce. Moreover, for common characters you find that your fingers learn how to *type* them (muscle memory), but your brain doesn't necessarily remember what they look like!

In fact, my disappointment with Zhengma as an approach for learning characters is the reason that I'm thinking of signing up for Skritter. I won't be using a mouse, though -- for Christmas I bought myself a writing tablet.

James, mind sharing why you are annoyed with the pinyin input method?

Well, I can't answer for him, but what I don't like about pinyin input is that it's quite slow, with no chance of getting faster -- even if you have a lot of practice you still have to check which character you want, maybe even scroll down if it's a more uncommon one.
Lately I'm just writing them by hand with the Bamboo Scribe input program, it's faster (with chance of getting even faster with practice) and I practice my handwriting at the same time.

Let's see:

- you need to know the pinyin to type the character, no good if you see an unfamiliar character somewhere and want to look it up.
- there are often many matching characters, so you have to scroll through the possibilities to find the one you want. With the Microsoft IMEs on Windows XP the characters are really small and hard to tell apart, especially for traditional characters.
- rare and ancient characters show up at the end of the list, if at all, which means you have to waste time looking in the list for something that isn't there. This includes many of the radicals which aren't used as characters.

James

James, I agree with all your points about why pinyin-based input is annoying. Slightly disappointed by Todd's report that Wubi doesn't really help you remember characters. I study Chinese at university, so I was thinking that any kind of hint could be a life-saver in written exams. Being able to remember that the code for some character is these four letters might spark my memory.

Ideally I'd study Wubi via Anki, but I can't find a huge list of Wubi codes - is anyone aware of one?

I am studying Wubi with a book in French (I am French…) by Pascal Penotet. It works quite well and I really think it can fix some memory issues. But it is true that it has to be worked and practiced before you can use it fluently. It gets faster and faster, but still very slow for difficult words as far as I am concerned. It would be great to be able to check wubi codes on Skritter, for instance I have to try them in another software while I study on Skritter.

One small interjection, if I may. Have you ever been to a Chinese internet bar? Those young people can type in Chinese just as fast as a western young person can in English/Spanish/French. They are all using the pinyin method.

I am not criticizing everyone for wanting to use a different method. I feel people should find what makes them happy. The statement that pinyin is inherently slow is false, though. The pinyin method is so popular because it is so effective.

If you are having a lot of problems, I recommend trying a different pinyin program. Window's pinyin sucks. Google's is decent. You could also search different forums and the Internet to find a good program.

Sorry for the interjection, but I felt the need to put that in. Please go back to venting :)

@west316 That may be. I'll definitely give Google's IME a try. I don't see how any of the pinyin-based input systems could help you in the case where you're looking up an unfamiliar character, though.

@jww1066 - I agree entirely with that one. That is a serious irritation. I resort to keeping my electronic pocket dictionary beside me for cases like that.

The key to fast typing with a pinyin IME is to avoid entering single characters as much as possible. Most disyllabic words are fairly unambiguous, and with modern pinyin IMEs it actually helps to enter a whole *sentence* at a time. For single characters, I often find it fast to type a word containing that character, then delete the other character that I don't want, instead of searching through pages of homophones.

It's certainly possible to type fast with wubi or zhengma, but compared with pinyin input it takes much more practice to reach that speed.

I don't know for sure if using Wubi will help your memory for characters compared to Pinyin, but it isn't a very good tool if used just for memory.

I second what Todd said. You can't remember a code and then decode it back to a Chinese character because: 1) There are too many possible combinations of roots in a code; each key has about ten. 2) the last letter of a code may denote a root or the positioning of the previous roots. 3) roots are omitted to keep codes within 4 letters long.

Also a factor: 4) Chinese characters are broken down in very unintuitive ways, such as with overlapping, and roots are not always radicals (maybe half?).

I believe these systems are meant only for typing, i.e. distributing Chinese characters across the keyboard and a code space.

The most helpful website for learning Wubizixing 86 and 98 I found, showing an actual graphical breakdown of many characters into their codes, is http://www.chenjiliang.com/Common/WbCode.aspx . For massive lists you can get files from input method editor projects, like IBus at Google Code: the package ibus-table-chinese source code contains a text file with Wubi 86 including all abbreviations. People have made Wubi 98 tables for IBus and posted them elsewhere. Also, SCIM came with Wubi 98.