This is something I’ve often seen explained, but never explained WELL, and I’d hope to find someone able to write a really good explanation or point to one.
​
We all know about jidōshi and tadōshi pairs, and that many jidōshi can employ objects marked with the particle を making them look like tadōshi at first glance (道を歩く、橋を渡る、空を飛ぶ, etc.), as many sources train learners to assume objects marked with を to be direct objects by default. It almost looks as if a grammar could arguably split up jidōshi further into something like “true jidōshi” and “を-using jidōshi”, although I’ve never seen that being done.
​
I wonder if this is a recognized issue among (academic) Japanese linguistics, or if it never was an issue or has been solved for a long time and is now just something textbooks and grammars usually fail to explain well.
​
Furthermore, I wonder if classifying the object of a verb of movement through space and time like 歩く and 飛ぶ not as a direct object is based on truly fundamental differences between “real” direct objects (as in 「飯を食べる」) or if this is more born out of convention/pragmatism (“direct objects are for tadōshi, hence the を-marked objects of a jidōshi like 歩くor 飛ぶ aren’t direct objects by exclusion”).
​
Not being a native speaker, now I’m even sitting here wondering if the distinction between jidōshi and tadōshi is permeable and if in theory ～てある-constructions with a jidōshi like 「この空は飛ばれてある」 (this sky is being flown on) are grammatical or not.