Cursive/handwritten font?

#1
I've noticed that sometimes there's a pretty substantial difference between the way a character looks in a typewritten or onscreen font and the way it looks when someone casually writes it down by hand -- sort of like the difference between block letters and cursive in English, I suppose. With the exception of one software package (Easy Chinese Tutor) all the learning materials I've seen deal exclusively with perfectly-formed typeset characters. I very often find it next to impossible to read a handwritten character I'd recognize instantly in printed form.

If PlecoDict had a "cursive Chinese" font, that'd be a big help. Especially if I could tell it to use that font for particular flashcard lists (so I could track my progress with handwritten characters separately from my progress with printed ones.) Does such a font even exist?

Barring that, anyone have any good resources for learning to read handwritten Chinese characters aside from constantly asking people what they just wrote?
 
#2
You are echoing a long-time search for me. First of all, Yale has a learning to read cursive Chinese text. It's not new, but pretty useful. Second, I often use Huang Cai as my Chinese font (it is a cursive font). Third, there is a dictionary Web site that has each word in all forms. If I recall correctly, the developer's name is Richard. And thanks for telling us about Easy Chinese Tutor--I didn't know it taught cursive.

Sandra
 
#3
Where can I get that Huang Cai font? I Googled it and didn't come up with anything.

It would be a little bit of an exaggeration to say that Easy Chinese Tutor teaches cursive. What it does is present a few different forms of each character, one of which is handwritten. It's not a great piece of software but it has its uses.
 

mikelove

皇帝
Staff member
#4
Interesting idea. We actually tried to license a cursive font from the same company we licensed SimSun from for the Pocket PC version, and they didn't have one available. But there are enough font foundries in China now that I'm sure one of them would be willing to offer us a font for a reasonable price. Though it would be difficult to learn to read cursive from just one font - there are so many different Chinese handwriting styles that you'd likely need dozens of fonts to cover them all. If you really want to learn to read cursive the best approach is to read a lot of cursive and ask a Chinese friend about the confusing characters.
 
#5
Sorry about the really stupid mistake. I meant huang CAO not cai. Check out myfonts.com for that face. You will also find 154 (I think) other Chinese fonts, including three other cursives.

The book I was describing is Chinese Cursive Script by Wang Fang-yu and it is published by Yale. Unless you want to treat this as a text to a yearlong course in cursive (which, I believe, is what it is), this is a pretty inconvenient book for us plebes.

the dictionary Web site I mentioned is Richard Sears's Chinese Etymology http://www.internationalscientific.org/. I was supposed to help him add cursive to the site (required no knowledge on my part, thank heavens), but I ran out of time, darn it. And it looks like he didn't find anyone else to volunteer. So, he has only a few cursive entries, but if you need to identify oracle characters and seal characters, as well as some other written styles, this is the place. If you can, consider contributing some $$, since Richard has done an incredible amount of work on this database all by himself.

Mike--I have a few cursive faces and I find that if I print out a text (including one I want to handwrite in a letter) in both kaiti and one or more of the cursives, I learn a great deal.
 
#6
NOTE: this is an earlier and fuller reply--I thought I had lost it. On second thought, the later (but previous :) ) reply may be more thorough. Whatever.

Sandra

OK. I finally looked at the items I was describing. Huang CAO (not cai). I'm so sorry especially since it was such an obvious mistake. http://www.myfonts.com/fonts/founder/fz-huang-cao-s09/fzhuang-cao-s-09-gb-2312/ Also, a good resource for links to Chinese fonts is http://cgm.cs.mcgill.ca/~luc/china.html. And then there's a more angular cursive http://www.myfonts.com/fonts/founder/fz-ying-bi-xing-shu-s16/.

Second, the cursive book's title is Chinese Cursive Script by Fang-yu Wang. I know that a good one exists in either China or Taiwan, but it's only in Chinese. It's alwasy seemed to me that if you have enough Chinese to read that book, then you already don't need the darn thing.

And Mike--I've learned a lot from printing something out in both kai shu or soje other straight-up typeface and in Huang Cao. Wang's book is inconvenient to study from unless you want a full yearlong course in grass characters.
 
#7
Thanks for the pointers! That font sure looks familiar; I think I've seen it on lots and lots of Chinese-language web sites.

Mike, add my vote for a nice 96-pixel magnified version of that font for PlecoDict. I'm okay with paying extra for it as an add-on. What would be coolest would be if I could tap on a character and see the magnified versions in all the available magnified fonts at the same time. I have a gigabyte memory card, may as well use it!
 
#8
There's an impressive book on the subject, available only in French I'm afraid and possibly difficult to get hold of, but quite worth it, the 草子汇 or Dictionnaire des Formes Cursives des Caracteres Chinois published in 1909. It was recently re-published (1986) by the Ricci Institute in Taipei under Lexica Numero 38, which might be easier to find.
The book was compiled and hand-written by a French navy officer, Stanislas Millot, who gives a fascinating account of his skills in the introduction, explaining how he helped decypher a hand-written message sent to a Chinese admiral and prisoner of war, which even the Japanese couldn't understand.
The book is essentially a series of tables listing 7200 different entries, conceptually similar to a traditional dictionary with radicals, except the entry points are the basic cursive shapes he identified as initials or finals. There are a number of character indexes, and a detailed step-wise process for analysing characters. There weren't any of the PRC simplified characters around at the time, but it makes it obvious where many of these came from.
PS - ad hoc translation of the intro borrowed from http://abecedaria.blogspot.com/2005/11/la-plume-caporal.html
Sinologists are often stopped in their work when they meet cursive or antique characters and the least inscription on a collectible can discredit them in the minds of laypeople.
The 20th of June [1900] we were at Taku on board the cruiser Le Pascal in the middle of about 30 warships of different nationalities.
We lacked news of Seymour's column and the besieged at Tien-tsin, and anxiety was at its peak when we received a message in cursive characters from an imprisonned Chinese general. (The officers) hoped to find interesting information in it but even the Japonese in spite of the constant use that they make of Chinese characters declared that they understood none of it.
They were on the point of sending a cruiser to Chefoo to have the document translated when they thought of showing it to me. Thanks to the special study that I had made, by chance, of cursive writing, I could, not without difficulty, provide the desired interpretation.
 
#9
goulniky said:
There's an impressive book on the subject, available only in French I'm afraid and possibly difficult to get hold of, but quite worth it, the Dictionnaire des Formes Cursives des Caracteres Chinois published in 1909.
Wow, if somebody translated that into English I think there might be a much better market for it than there was in 1909. : )
 
#10
I'm sure there never was and never will be a huge market for such a scholarly piece of work.
But it's mainly character tables, so there's very little that needs translating other than the intro and a 10-steps analytical process. Mind you, the copy itself is handwritten, cursive too and very meticulous , I could give it a try. The real issue is to scan and organize the various caozi, which incidently also include the Japanese kana.
 
#11
That book is old enough that it should be in the public domain now, yes? Maybe someone somewhere has scanned it in. Seems like a perfect candidate for turning into a Web site. (Google didn't turn up anything, though.)
 
#12
I've asked the Ricci Institute whether it's in the public domain. I wonder how much space it would take to store 200 scanned pages at a reasonable resolution... Depending on Ricci's answer, I'll give it a try and store a few pages on my website.
BTW, according to EU law, copyright protection extends 70 years after the author's death. It used to be 50 years but was changed in 1993, gradually translated into member state law : so unless it was already in the public domain by 1995 (i.e. the author died prior to 1945) it would only be if Stanislas Millot died before 1935. Unlikely given he co-published another book in 1935
 
#14
I will be delighted to volunteer space to store the whole book and make it available to the public. My web server has plenty of unused space and bandwidth. If the book turns out to be public-domain, PM me and we can talk about how best to get the scanned pages to my machine.

As for how big it'll be, a 300DPI greyscale scan of a letter-size printed page in PNG format is around 4 megabytes. Obviously that'll vary depending on the contents of the page, but that should give you a rough order-of-magnitude idea. So that'd be on the order of 800MB for the entire book. If it were me, I'd want to keep a high-res lossless version around as a source, then use that to make some smaller screen-resolution versions for casual viewing (JPEG might be okay for those).
 

mikelove

皇帝
Staff member
#17
goulniky, thanks for posting these. One fun thing to do with them is put them into PlecoDict's handwriting recognizer - I just did some quick testing and it looks like it can recognize almost half of them. (mostly the simpler ones) Pretty impressive, considering that Chinese script has likely evolved a lot in the past hundred years.

We've recently licensed a very large database of Chinese handwriting data (mainly to use with a stroke order testing feature) and it's possible we might be able to mine that data to use for a cursive training feature - the data hasn't arrived yet, but once it does we should know for certain whether or not this is feasible.
 
#18
[21-feb-2006] I have now received permission from Benoît Vermander (Taipei Ricci Institute) and the Ricci Dictionary Association to make Stanislas Millot's dictionary available online.
I will start scanning and uploading it next week, I may also translate some of the explanatory notes - watch this space.
 
#19
Though it would be difficult to learn to read cursive from just one font - there are so many different Chinese handwriting styles that you'd likely need dozens of fonts to cover them all.
 
#20
Re:

goulniky said:
[21-feb-2006] I have now received permission from Benoît Vermander (Taipei Ricci Institute) and the Ricci Dictionary Association to make Stanislas Millot's dictionary available online.
I will start scanning and uploading it next week, I may also translate some of the explanatory notes - watch this space.
Is the space full yet? I would not mind perusing that book either...
 
Top