OT: I wonder if something similar exists for mandarin (hanzi+pinyin)

These kinds of fonts have existed for a long time, including some of which are free-to-use. The Wang Han-tsung font collection includes eight fonts that provide reading aids: wp{0..3}10-05.ttf and wp{0..3}10-08.ttf. This comprises two typefaces and three variants for the three (most-common?) alternate pronunciations. (Obviously, this suggests that some pronunciations may be unrepresentable—e.g., 「和」 apparently has six possible pronunciations—and I don’t know what one is supposed to do.)

This kind of functionality is very readily found, but I think it’s most commonly provided as a feature that does not use special fonts: https://support.microsoft.com/zh-tw/office/languages/add-phonetic-guides-to-east-asian-text In LibreOffice, there is a built-in formatting option that allows you to add arbitrary additional phonetic guidance to a text span (but the user is responsibility for determining what that should be.)

You’d have to ask a Chinese speaker, but I think that this is of similarly narrow use as another comment suggests it might be for Japanese. As I understand, the use of reading aids drops off fairly early in elementary school—e.g., 「國民小學一年級至三年級之教科用書及其習作，應全部加注音符號。」“In instructional materials from first to third grade, provide full Chu-yin phonetic notation”

For Chinese learners, well, since the drop-off for native speakers is so early, I suspect there is a lot of risk is becoming reliant on aides.

I suspect the above holds more-or-less for Japanese speakers, though I’m not sure about Japanese learners. So these fonts are probably pretty useful if you are writing introductory language learning materials or children’s books, but I imagine that’s about it. For anything else where pronunciation guidance is needed (e.g., proper nouns or Mandarinised, potentially non-character based, Chinese languages,) you probably want to match the typeface of the rest of the text. Additionally, I suspect in Chinese that this guidance might be useful primarily in very obscure pronunciations, which are unlikely to be covered.

What might be best for learning purposes would be to have your learning system determine (via spaced repetition?) the words in a given text that require additional reading guidance, then using the LibreOffice functionality to provide that only for those characters. Your LLM is probably pretty good at figuring out even moderately obscure Mandarin pronunciations that would appear in a learning text, so deriving and applying these should be easy.

These fonts are nice, and using built-in font features to select variants (rather than forcing the user to select these manually) is slick. I’m glad they exist, but I’m just not quite sure whether they’re altogether that necessary…

score 1 · canonical