I don’t think I’ve seen a post on this specific topic yet, so I wanted to post this (my first post btw). I’ve noticed that many Japanese compound words often change their accents when they are put together. For example, 試験 has the accent on け, but in 日本語能力試験, the accent lies on し, according to OJAD. I’ve searched online for this but there aren’t many reliable resources that I know of. Even if they are reliable, they are part of some linguistic academic paper with confusing terminology. I was wondering if there were any general rules that helps explain this. Thanks.
It is very common in compound words (specifically jukugo) for the pitch to drop on the first mora of the second word. Your example is a bit odd in that it combines three words, but apparently it still follows this guideline.
Two examples off the top of my head are 物語 and 小学校.
I don’t have a source for this rule. I read somewhere that this rule is particularly strong when the two words being combined are both heiban independently, but I’ve noticed that it’s still incredibly common even when that’s not the case.
I remember watching Dogen’s video about this at some point and as far as I remember, the only ‘rule’ that is actually useful is that almost all long compound words (5 mora or more is my guess) are nakadaka. Fairly often the drop will land on the first mora of the second word (or third word etc). But even that is not consistent.
There are so many rules and exceptions to compound words that it’s better to just treat each compound word as its own vocabulary to learn, rather than trying to work out the pitch accent on the fly.