In a question asked today, a user asked about akayorosi, which is represented by five characters:


The second character appears to be a hentaigana simplified from 可, representing /ka/. The user chose to represent it in text with the kanji 可. In this particular case, it caused some confusion because the character in question looks similar to hiragana の.

Would it be more appropriate to refer to this as あかよろし, using modern kana instead?

According to Wikipedia, hentaigana are still not included in Unicode. I wouldn't be surprised if that information is out-of-date, but I imagine user fonts are unlikely to contain hentaigana glyphs, so I'm not sure we'd be able to represent them directly without embedding a font of some sort. (I'm not very knowledgeable about this subject, though.)

How should we represent hentaigana in questions?

  • In China, there is a standard convention to use “□”(虚缺号) to transcribe lost, indiscernible or hard-to-represent characters. It's sometimes followed by a footnote. I don't know if Japanese has a similar convention. – Yang Muye Mar 27 '14 at 20:29
  • 青空文庫 solves this (外字, 欠字) by inserting annotations. e.g. あ※[# カの変体仮名]よろし. – Yang Muye Mar 27 '14 at 20:56
  • I wonder if @Dono knows anything about the latest situation in Unicode. – Darius Jahandarie Mar 29 '14 at 14:11
  • 1
    @DariusJahandarie Hentaigana are not encoded yet, nor are they even in the upcoming pipeline. I am not aware of a even a working proposal at this time. If may come sometime, but for the time being this is not a priority for the Japanese national body, from which any serious proposal would need to come from. As far as I am aware, the status quo may be found in L2/11-229, which may be found here: unicode.org/L2/L2011/11229-n4091.pdf . If the Japanese national body ever does decide to sponsor such a proposal, the Unicode Consortium will work on standardizing it. – Dono Mar 31 '14 at 5:40
  • Relevant link meta.japanese.stackexchange.com/q/1672/1628 – Earthliŋ Dec 6 '16 at 22:05

Since hentaigana are usually written using kanji due to the restrictions of Unicode, it seems that the furigana engine provides a natural solution to this problem:


You must log in to answer this question.

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