2A02:85F:EC32:2291:9828:EF55:23B2:7239

Regarding 自動詞 and 他動詞 being semantic terms rather than syntactic, that claim seems to be demonstrably wrong. Consider the following counterexamples:

戦う is a 自動詞, despite the fact that when you fight someone you're evidently exerting great physical influence over them. This is because it takes arguments with と, which are not considered to be objects.

望む and 待つ are 他動詞. This is because they do take objects (marked with を), even though most people would agree that when you desire something (Aを望む) you're not really doing anything to it or directly affecting it in any way (ditto for when you're waiting for something or someone, Aを待つ).

It seems the case to me that 自動詞 and 他動詞 are indeed syntactic terms, although, unlike "(in)transitive" in English, they're assigned on a *per-verb* basis, rather than a per-sentence one, and they denote *potentiality*, rather than in-context *actuality*. That is, 食べる is a 他動詞 because it *can* take an object, even not every instance of it does.

On the point of 道を歩く, 歩く is classified as intransitive because 道 there is not considered to be its object (i.e. を is not acting as an object marker). This を is considered to be an "area of traversal" marker that's used with verbs of movement/motion (～を行く、～を歩く、～を走る、～を曲がる、～を回る、～を飛ぶ、～を泳ぐ). There's yet another special, non-object-marking function を. That'd be the "point of exit/separation/detachment" を seen with 出る、降りる、離れる. Hence why all of those are intransitive too. This is supported by the JP article on the accusative case (対格):

> 現代日本語の対格を示す助詞「を」には、次のような用法がある。 > 直接目的語を表す。意味的には様々のものがある。「車を見かける」「車を壊す」「ご飯を作る」「小学生を教える」「英語を教える」 > ある種の移動の起点を表す。「家を出る」「大学を卒業する」 > 移動の経由地を表す。「高速道路を走る」「橋を渡る」「角を曲がる」 > 下二つの用法は、対格の典型的用法から区別する場合もある。

...which defines three functions of を: direct-object-を (accusative case clitic), detachment-を, and area-を.

Makino and Tsutsui's Dictionary of Basic Japanese Grammar almost agrees too (it actually defines a fourth "emotive" を, which marks the cause of human emotion, as in ～を悲しむ — call me arrogant, but I'm calling that part into question; 悲しむ is classified as transitive afaik, plus they reference semantic reasons for the distinction).

Further discussion.

Cross-linguistic differences in which verbs take objects and which don't have no bearing on the fundamental issue that the 自動詞・他動詞 grammatical categories in Japanese are orthogonal to syntax (a given word is classed as a 自動詞 or a 他動詞, regardless of the syntax of a given sentence), while the "transitive / intransitive" grammatical categories in English are dependent on syntax (a given word is classed as "transitive" or "intransitive" depending entirely upon the syntax of a given sentence).

Look again at the 他動詞#Usage_notes section, particularly from the sentence “For instance, in English, the verb ate in the simple sentence “I ate” would be considered an intransitive verb, because it is not followed by an object.”. Various 自動詞 can take objects, but that does not make them 他動詞. Any 他動詞 can be used without any stated object, but that does not make them 自動詞. Meanwhile, any verb in English that is used in a sentence without a stated object is being used intransitively. See also the following paragraph from the w:Transitive verb article:

Verbs that can be used in an intransitive or transitive way are called ambitransitive verbs. In English, an example is the verb to eat; the sentences You eat (with an intransitive form) and You eat apples (a transitive form that has apples as the object) are both grammatical.

Put another way, 自動詞 ≠ "intransitive", and 他動詞 ≠ "transitive", strictly speaking -- but for purposes of simplicity, much English-language material uses the term "intransitive" as a translation for 自動詞, and the term "transitive" as a translation for 他動詞. ‑‑ Eiríkr Útlendi │^{Tala við mig} 06:53, 28 June 2023 (UTC)

> Cross-linguistic differences in which verbs take objects and which don't have no bearing on the fundamental issue that the 自動詞・他動詞 grammatical categories in Japanese are orthogonal to syntax (a given word is classed as a 自動詞 or a 他動詞, regardless of the syntax of a given sentence), while the "transitive / intransitive" grammatical categories in English are dependent on syntax (a given word is classed as "transitive" or "intransitive" depending entirely upon the syntax of a given sentence).

Yes, I acknowledged this exact fact when I said that transitivity in Japanese is decided on a per-verb basis, while in English it's decided contextually on a per-sentence one.

> Various 自動詞 can take objects, but that does not make them 他動詞.

Care to provide examples? Because I literally know 0 of them. As far as I'm aware, no 自動詞 can take an object.

If you're gonna reply with a verb of motion, may I remind you that their を-marked arguments are not direct objects. Likewise for verbs denoting exit/separation/detachment.

> 自動詞 ≠ "intransitive", and 他動詞 ≠ "transitive", strictly speaking

Indeed, but that doesn't render 自動詞 and 他動詞 semantic terms. It's just that in Japanese they're a universal/inherent property of the word (i.e. a given JP verb either always is or isn't a 自動詞, end of story), while in English the transitivity of words depends on the context. 2A02:85F:EC32:2291:9828:EF55:23B2:7239 07:05, 28 June 2023 (UTC)

> Because I literally know 0 of them.

Welp, I already told you, but I just found one... probably. Need to check somehow that the を in 戦争を戦う is really an object marker (the one objecthood test I'm aware of is that object-を can be converted to が in the potential and ～たい [and ～てある I guess], whereas the other two をs will never do that). 2A02:85F:EC32:2291:9828:EF55:23B2:7239 07:44, 28 June 2023 (UTC)

See the Reddit thread for a fuller treatment: https://www.reddit.com/r/LearnJapanese/comments/14ktw3u/comment/jptvsu9/?context=3 ‑‑ Eiríkr Útlendi │^{Tala við mig} 08:21, 28 June 2023 (UTC)

One thing --

"Indeed, but that doesn't render 自動詞 and 他動詞 semantic terms. It's just that in Japanese they're a universal/inherent property of the word (i.e. a given JP verb either always is or isn't a 自動詞, end of story), while in English the transitivity of words depends on the context."

... um, that's exactly what I'm talking about with regard to "semantic". Whether a given Japanese verb is classed as 自動詞 or 他動詞 depends on Japanese speakers' conception of the meaning of the word, which is "a universal/inherent property of the word".

Hopefully this helps clarify? 😄 ‑‑ Eiríkr Útlendi │^{Tala við mig} 08:27, 28 June 2023 (UTC)

Meaning is an inherent/invariant property of words, but that doesn't mean that all such properties relate to meaning. I don't get how you make the logical leap from "syntax-independent" to "semantic" (i.e. meaning-dependent). Can't you just be both syntax-independent and meaning-independent? 2A02:85F:EC32:2291:9828:EF55:23B2:7239 08:39, 28 June 2023 (UTC)

It could be independent of both syntax and meaning.

However, Japanese sources describe 他動詞-ness and 自動詞-ness in terms of the meaning of the verb. See, for instance, the Britannica International Encyclopedia Japanese entry for 自動詞 here at Kotobank. The entry begins with the text (emphasis mine):

動詞を意味のうえで，その動作の直接の対象を必要とするものとしないものとに分けたとき...

So I don't think we can accurately describe 他動詞-ness and 自動詞-ness as independent of meaning. ‑‑ Eiríkr Útlendi │^{Tala við mig} 08:50, 28 June 2023 (UTC)

Yeah, good point. Thanks for the citation!

I guess my stubborn ass just wants to see Japanese people explain to me how my counterexamples aren't actually counterexamples to be fully convinced. And if they can't do that (i.e. they agree that, meaning-wise, 楽しむ should logically be a 自動詞, even though it's not), then that'd legitimately prove sources like Britannica wrong.

You can have more than one valid definition, by the way. Just because a semantic association does exist, doesn't mean you can't also (re)define 自動詞・他動詞 in terms of the (in)ability to take a direct object. Like, if it accurately describes/defines the terms... Well, there you go, it's a functional definition.

(although, again, I may already have found one counterexample to my definition [confirmation pending] and could certainly find many more yet in the future — feel free to chip in yourself) 2A02:85F:EC32:2291:9828:EF55:23B2:7239 09:04, 28 June 2023 (UTC)

Okay, personal theory officially disproven: 戦う is a 自動詞 that can legitimately take a direct object (far as I can tell). Still seems to be a really good rule of thumb regardless, that 自動詞 can't take objects (and that 他動詞 can is, to my current knowledge, always true). 2A02:85F:EC32:2291:9828:EF55:23B2:7239 09:22, 28 June 2023 (UTC)

Re: 楽しむ, 悲しむ, and that set of words, it seems like the [ADJ STEM] + む class of verbs derives from something like "to think about the object in a way that makes one feel the adjective". I've seen a couple Japanese definitions like 悲しむ = 悲しく思う, or 楽しむ = 楽しく思う.

ご参考までに。😄 ‑‑ Eiríkr Útlendi │^{Tala við mig} 21:33, 28 June 2023 (UTC)

あ、今まで辛抱強く付き合って、いろいろ教えてくれて、ありがとうございました。 2A02:85F:EC32:2291:9828:EF55:23B2:7239 23:09, 28 June 2023 (UTC)