compound word dating in Japanese?

[LinguistCat]

How do you know (or have a best guess) for when a compound word was formed in Japanese, even just a general stage of the language's development? Is there a good way to do this besides looking up etymologies (which, until I find better references, limits me to what is on wiktionary)? Like, some would be relatively modern, if part of the compound came from a foreign language that didn't have contact with Japan before a certain time, or if it was a technology or concept that was only known to Japan after a certain stage. But what about compounds derived from onyomi of certain kanji, where it could have been derived earlier or later? Should I assume that compounds of native Japanese words probably date pretty far back (Old or even Pre-Old/Proto Japonic?) unless I have good reason to think otherwise, or is there a specific stage I should assume it's from unless there's evidence it showed up earlier OR later? Are there any general tips for "figuring out when a compound was made in any given language" that someone could offer even if they aren't familiar with the development of Japanese?

I'm trying to derive a relative to Japanese and the Ryukyuan languages for a modern fantasy story, so when compounds were formed or when words were borrowed is important for this. I might derive things differently if a word was borrowed/made after the split (it wasn't a clean language split until after the Heian period but there was some borrowing after that from Japanese), but I want to include most cognates from before the split including some things that fell out of standard modern Japanese.

[Soap]

There's definitely some diachronuc morphology that's engrained in the derivational morphology. E.g. rendaku comes from the genitive linking particle /no/, but from a time when it was pronounced /nə/. Exceptions like /amembo/ "lollipop " show that even newly coined words may use old tricks of word building.

Afaik the Chinese compounds are written as if Chinese, rega4dless of age, so e.g. when a word like /jitensha/ "bic6cle: is formed, it used the sounds of the Chinese characters in 700 ad, not 1900 ad.

[LinguistCat]

Thank you. I recently saw about no > rendaku. I was hoping there was some temporal component to how kanji readings were used but it does seem it's more about slight (or major!) differences in the meanings certain readings have. Ah, unfortunate, but still good to know.