Variants in Traditional Chinese characters?

[Mark Yong]

I have been typing in Chinese for a few months using MS Windows XP (Traditional Chinese, using the Taiwan setting). I have been using the PMingLiu font regularly.

Recently, I decided to try out the SimSun font, also in Traditional Chinese. I notice that some of the characters appear differently. The examples I came up with are as follows:

給/給, 為/為, 曾/曾

(In case the difference do not show up here, copy and paste the word pairs into MSWord, and set the font for one word to PMingLiu and the other word to SimSun.)

Note again that both words in the pairings are Traditional Chinese, not Simplified Chinese.

For 給, I am aware that one form is the version that is used for printing (the PMingLiu version), and the other is the form used for writing (the SimSun version). This applies in general for all words using the radical 糸.

However, I am not so sure about 為 and 曾. For 曾, in the PMingLiu font, the top of the character is 八, whereas in the SimSun font, the top of the character is 丷. Note that in these two variants, the number of strokes remains the same (I personally find it easier tow write using the 丷 top, as I can never get the 八 well-proportioned!).

In the special case of 為, in the PMingLiu font, the top of the character is 爪, whereas in the SimSun font, it is techinically reduced to just 、, with the 丿 stroke merged with the rest of the character. In this instance, there is a difference of 3 stroke counts (the PMingLiu version has 12; the SimSun version has 9). Also, under the Kangxi radical classification, 為 falls under the爪 radical, which would not be obvious in the SimSun version.

Any comments? Are there two or more commonly-accepted standards for writing certain Traditional Chinese characters (perhaps differences between Taiwan, Hong Kong and old-China fonts)? Or are they just differences between the printed form (宋書) and the pattern form (隸書)?

[zhwj]

The image at the bottom of this page has a chart that appears in most mainland Chinese dictionaries detailing the differences between the old-style printed characters and the newer style ones that more closely resemble hand-written forms.

爲/為 and 著/着 are older variants, I believe, and not part of the systematic changes; some that aren't in most printed character sets include 扌 for 方 in words like 於.

[Mark Yong]

Hi, ZHWJ,

Thanks for the URL, it was really useful.

So, does it mean that for Kangxi radical-indexed dictionaries that use 為 instead of 爲, they will not classify 為 under the category of words using the 爪 radical (as is the case with all dictionaries that use the standard Kangxi radical classification)? Does the same situation apply to other variants of Traditional characters, i.e. they are classified differently?

Cheers.

[zhwj]

為 is indexed under 丶 (like 为)

爲 is indexed under 爪

Dictionaries will usually list all forms, and point the variants to the main entry. This is for ease of use - one shouldn't have to know the historical form of a character in order to look it up. Most recent dictionaries also have pointers for radicals with non-obvious stroke counts as well ( 扌 will say 见手 and so forth).

[malinuo]

I still don't get it. For 給 and 曾 it makes some sort of sense. However, for 為, unicode 70BA, why would PMingLiu represent it as a glyph which is identical to another perfectly valid unicode character 爲 (7232)?

Out of all the standard fonts, PMingLiu and MingLiu seem to be the only ones that do it. On my Mac, there is no single font that represents unicode 70BA as 爲.

The lookup radical varies from dictionary to dictionary. The Character Palette on MacOS uses 火 as radical for 為. (Not the first surprise with the Character Palette radical assignment.)

[zhwj]

The MingLi glyphs are actually not identical. 爲 has a 爪 with inward facing strokes, while 為 is represented by a 爪 that has outward spreading strokes.

[malinuo]

Still not clear. MS Mincho and MS Gothic both display 爲 7232 with outward spreading strokes, just like MingLiu's 為 70BA. It seems Japanese and Korean fonts mostly work that way. The fonts on my Mac confirm this.

So: If you want the character 爲 with outward spreading strokes you can eitheruse one of many Japanese and Korean fonts. Oryou can use 為 with MingLiu.

MingLiu and PMingLiu simply stand very alone against a wide range of other fonts, and probably also against the standards and intentions of unicode.

[zhwj]

為 is indexed under 火 in the Unihan database, so your character palette is not surprising this time.

The unicode standard does say "The shapes of the reference glyphs used in these code charts are not prescriptive. Considerable variation is to be expected." So there's probably some combination of the history of the Ming Light font families and a sense that the two glyphs aren't exactly identical that keeps the 為 shape from appearing in MingLiU fonts today. It does sort of wreck the radical indexing, though.

Another question: why does 直 have only one grapheme in the index, while 眞 and 真 are listed separately? Are there things going on in the other East Asian Languages that influence this (and perhaps 為 as well)?

[malinuo]

Another question: why does 直 have only one grapheme in the index, while 眞 and 真 are listed separately? Are there things going on in the other East Asian Languages that influence this (and perhaps 為 as well)?

I don't think there is any reason for that. They keep adding new codes all the time to solve that kind of situations. You also have pairs like 毎 vs 每 but only 海. I seem to remember that there only was one code for both 毎 and 每 in earlier versions of Unicode.

[Mark Yong]

Dictionaries will usually list all forms, and point the variants to the main entry. This is for ease of use - one shouldn't have to know the historical form of a character in order to look it up.

Unfortunately, the three dictionaries I own, i.e. 康熙字典, 辭源 and Mathews' Chinese-English Dictionary, list on 爲 (under the 爪 radical, as you correctly pointed out), but they do not provide the alternate form 為 - not even separately listed under the 火 radical or even 丶). Which dictionary did you see the form 為 in? I am just wondering, if the 爲/為 multi-radical classification exists, there could potentially be others, too.