Hacker News Reader

Japanese is great for this also, at least for signage written in Katakana or Hiragana. If you learn those alphabets you will always know how to pronounce words written in them.

Kanji is a very different story of course.

View on HN

There are exceptions, certainly in colloquial speech variants, 雰囲気 and 原因 come to mind. Both have an annoying ん ("n") in the middle that makes them a pain to pronounce even for natives, so the first gets changed from ふんいき to ふいんき and the latter from げんいん to げいいん. So, if written properly in hiragana you'd know how to pronounce them "properly" but not necessarily as they're actually used by a significant amount of people when spoken. Even the auto-kanji-conversion of my keyboard respects the variants.

I can accept ふいんき as an exception, but for 原因 that's just a consequence of the rules for how to pronounce ん, which depends on the surrounding sounds (it's not quite い either). I wouldn't really call it an exception. Consider 一千円, 禁煙 etc all of which are pronounced like that.

There are two distinct pronunciations, げんいん and げいいん, the first is the standard, the second is a variant. The only consequence of the ん in the second one is that it has been replaced because it's a pain to pronounce in that position. That there are other words with ん in a similar position (一千円 not being one of them) is by the by, 原因 and 雰囲気 are two of most frequently used words in the language (in the first 800), 禁煙 is used far less (in the first 5000) so under less pressure for adaptation.

Reply on news.ycombinator.com