Within “families” of phono-semantic compositions, it’s not rare for two kanji to have somewhat different readings. 尺 and 択, 干 and 軒, 工 and 空, etc. A lot of these are rendaku or generally understandable, but there’s one pattern I can’t wrap my head around: at the end of some readings, い can change into つ and viceversa.
Anyone know what’s up with this stuff? It’s so unintuitive but it looks like a real thing. Like with these:
い → つ
つ → い
I remember reading somewhere (I can’t for the life of me remember where since it was years ago), that in the case of 2 mora on’yomi readings where the second mora is a consonant+‘u’ combination instead of a pure vowel, it’s reflective of an old Chinese dialect that was in use at the time the word was imported, in which kanji readings could end in pure consonants (which of course can’t exist in Japanese so the ‘u’ is added).
If what I’m remembering is correct, this relationship you noticed between readings might be reflective of readings imported in different times when different dialects of Chinese were dominant, in which the reading changed to a consonant ending one from a vowel ending one (or viceversa).
That said, the history of kanji in Japan is long and complicated, since kanji and vocab were imported from China in many waves, in different eras, over many centuries. During this time, in China, different dialects were predominant based on which dynasty was in power at the time, so this would also have been reflected in how words and kanji got imported into Japanese as well.
I’m no linguist, or scholar of ancient languages by any stretch of the imagination and this is just my uninformed guess, but I am definitely curious about the topic, if someone more knowledgeable about this subject has more information.
That definitely makes sense, especially when looking at gairaigo. I’ve read that loss of certain phonemes contributed to tonalization in Chinese, I’ll look into these final consonants.