Due to Han unification, the same code point represents the Chinese (T/S), Japanese and Korean variants of a given character. The rendered glyph is determined by locale, lang tag (web) or similar. If a variable like this is misconfigured, it can lead to confusion or irritation with natives and misunderstandings with learners.

One such character is 噌 like in 味噌{みそ}. Source Han Sans renders it as but MS Gothic on XP renders it as . This led me to wonder which one is actually correct.

There are characters for both variants of the right-hand side of 噌; 曾 and 曽. 曽 is more common according to Jim Breen's KANJIDIC.

1 Answer 1


It's off-topic, but upgrading OS is recommended as Windows XP support is already terminated.

The reason why the glyphs differ is because the rule changed in 2004, after the release of XP. The following chart (from WP) would be the summary.

enter image description here

The second from left of row 4 is the 噌. It becomes rendered in more traditional shape since 2004. The JIS standard decided to conform their representational glyphs to 表外漢字字体表 publicized by the Agency for Cultural Affairs. Before that, no official stipulation for those not in 常用漢字 so JIS committee chose glyphs on their own (see my another post to Why wasn't 欒 simplified to 栾, when 戀 was simplified to 恋?). All fonts shipped with succeeding OSes since Vista should match the new form.

The glyph change is only normative to printed characters. Both shapes are recognized in handwriting and people usually choose the simpler one.

The download page of JIS2004 conformative fonts for XP (in Japanese)

  • 1
    Thanks, this explained the matter perfectly. On off-topic, don't worry. I'm not using XP as my primary operating system. I run it in a virtual machine on linux to play Windows games.
    – siikamiika
    Dec 26, 2016 at 4:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .