Transitivity of Compound Verbs

My understanding is as such concerning transitivity of compound verbs V1V2:

1) If V1 and V2 are both transitive, then V1V2 is transitive; if V1 and V2 are both intransitive, then V1V2 is intransitive, e.g. 切り倒す, 持ち上げる, 踏み潰す; 飛び跳ねる, 泣き叫ぶ, 飛び歩く

2) If V1 is intransitive and V2 is transitive, then V1V2 is transitive, e.g. 跳ね返す, 泣き落とす, 乗り返す

3) If V1 is transitive and V2 is intransitive, then V1V2's transitivity is ambiguous, e.g. 巻き付く, 飲み歩く

Question 1: Are these propositions reliable, if not perfect, guide to determine the transitivity of V1V2 compounds?

Question 2: Concerning proposition 3, are there verbs that dominate other verbs so that if it exists in either V1 or V2 position it would automatically determine the transitivity of the compound?

Question 3: What's your experience in learning the transitivity of verbs? Do you memorize them? Or imagine the action and determine if an agent must be involved? Or other methods?

As always, thank you for time and input.

おはようございます。皆さんは元気ですか。私は元気です。

I can only answer your last question, as I do not know what you are talking about. :roflmao:

I just memorize them as whether they are transitive or not. I tried picturing whether a agent was needed as you said, but I found that would confuse and slow me down.

じゃ、また。

Alright, I'll go for the rest.
In the first place, I have to point out that what you asked is far more complicated probably than you thought. There are many types of compound verbs in Japanese. Some compound verbs just connect the two original meanings of each verb (e.g 走り去る, 焼き切る), and some have little to do with the original meanings and can be considered as a new verb (e.g. 打ち解ける, 引き立つ). Also, there are ones that only V1, which is called 前項動詞 "first compound verb" in Japanese grammar, gives the main meaning, and V2 (後項動詞 "second compound verb") only adds grammatical information such like aspect, direction of the movement or emphasis. In these cases, V2 can be considered a kind of auxiliary verb, and whether the compound verb is intransitive(Vi) or transitive(Vt) is decided by V1, as same as verbs attached to the -te form such like いる/ある/おく, いく/くる or やる/もらう/くれる. (Incidentally, these verbs are called 補助動詞, and the most common English translation of this is "auxiliary verb").
e.g.
始める (Vt, "the action starts")
降る Vi --> 降り始める Vi
書く Vt --> 書き始める Vt

過ぎる (Vi, "to do too much")
動く Vi --> 動き過ぎる Vi
動かす Vt --> 動かし過ぎる Vt

However, there are also verbs that can be considered as an auxiliary verb from the perspective of function, but also rules the intransitivity/transitivity of the compound verb.
e.g.
かかる (Vi, "direction of action towards~")
飛ぶ Vi --> 飛びかかる Vi
殴る Vt --> 殴りかかる Vi

And there also exists another pattern that V2 differs depending on V1.
e.g.
合う / 合わせる (Vi / Vt, adding a nuance of mutual)
混ざる Vi --> 混ざり合う Vi
混ぜる Vt --> 混ぜ合わせる Vt
(Note that 混ぜ合う (Vt) also exists. The meanings are "Two things mix", "The agent mixes two things" and "Two agents do the action 'mix' together", respectively.)

Furthermore, there are ones that V2 is the main and V1 works like a prefix (e.g. 差す(Vt); 差し迫る (Vi), 差し出す(Vt)), or Vi and Vt are the same form (e.g. 引く(Vt) 上げる(Vt); 敵軍が国境から引き上げた (Vi), 政府は軍を紛争地帯から引き上げた (Vt)).
Now you can see that exceptions exist even for 1) and 2), right?

Transitivity of syntactic compounds are certainly simpler to determine. Here I am just trying to figure out if people are cognitively constrained syntactically when trying to produce compounds that are not sums of their parts. I guess it would be difficult to analyze the reliability of these propositions without statistically analyze a sizeable corpus of lexical compounds. Do you know if there is a list of lexical compounds by transitivity? I could only find dictionaries of compound verbs. For example here. I guess I could extract all the VV verbs from this corpus, but I would still have to look up the transitivity of each, but there are over 1500 of them...

By the way, in the counterexample you use -- 敵軍が国境から引き上げた -- Isn't the direct object, e.g. 兵士たちを, implicit in the sentence so that 引き上げた is transitive?

飛べるハム太郎 said:

Here I am just trying to figure out if people are cognitively constrained syntactically when trying to produce compounds that are not sums of their parts.

Click to expand...

Do you mean you think people can make new compound verbs that have nothing or little to do with the original meanings of V1 and V2?

飛べるハム太郎 said:

By the way, in the counterexample you use -- 敵軍が国境から引き上げた -- Isn't the direct object, e.g. 兵士たちを, implicit in the sentence so that 引き上げた is transitive?

Click to expand...

There is no problem with その兵士は国境から引き上げた or 調査を終え、私は現場から引き上げた. Or, should I have given you really used ones, if you have doubts with examples I made?

他の者たちも各々自室へ引き上げた。
お鳥見女房 諸田玲子著

とにかく引き上げようと僕は思った。これ以上ここにいても仕方ない。
ダンス・ダンス・ダンス 下 村上春樹著

時間が七時を過ぎていたので、待ちかねて一人で役所へ帰られたものと思い、急いで引き上げた。
日本の原爆記録 11 家永三郎ほか編

先ほどまで一方の主役だった沢村前覇王は、ほんの十分ほど宴席にいただけで、自室に引き上げていた。
奥羽路七冠王の殺人 本岡類著

出されたものを引ったくり、床の上に叩きつけると、翌日は早々に江戸へ引き上げた。
維新暗殺秘録 平尾道雄著

結局、ジュドーは引き上げるしかなかった。
機動戦士ガンダムZZ 第2部 遠藤明吾著

関さんは意気揚々として、フェアウェイにいる私たちの方へ引き上げてきた。
サービスの天才たち 野地秩嘉著

「ちょっと調子に乗りすぎたみたいやな。さあ、そろそろ引き上げようか」
異物 玄月著

料理人は引き上げて誰もいなかった。
女性の自己表現と文化 水田宗子編

そこで彼は一たん自分の部屋へ引き上げることにして、何気なくちょいと観察室をのぞいてみた。
ザ・ベスト・オブ・ジョン・コリア ジョン・コリア著 中西秀男訳

お国は小座に引き上げて寝についた。
二本の銀杏 上 海音寺潮五郎著

彼が自分の寝室に引き上げた後、私は今あの部屋のベッドに横になっている彼の気持ちをいろいろ想像した。
カサンドラの城 ドディー・スミス著 石田英子訳

幸い、彼らは追撃してこなかったので、ツールまで無事に引き上げることができたが、
デルフィニア戦記 茅田砂胡著

そうかと云って、明智城を見捨てて、引き上げるわけにも行かなかった。
武田勝頼 1 新田次郎著

やがて皆が諦めて引き上げてしまうと、岩棚を下り始めた。
QED ベイカー街の問題 高田崇史著

へこへこと頭を下げて私は与えられた客間に引き上げた。
紙婚式 山本文緒著

「みんな出ろ。―俺だけ少し残る。他は引き上げていい」
死なないで 赤川次郎著

そろそろ引き上げようと腰を浮かしかけたとき、電話が鳴った。
危険のP スー・グラフトン著 嵯峨静江訳

Toritoribe said:

Do you mean you think people can make new compound verbs that have nothing or little to do with the original meanings of V1 and V2?

Click to expand...

For example, the meaning of 飲み歩く can not be immediately determined by the sums of its parts. A compound may be so, I think, because 1) they have a fossilized meaning as to become idiomatic 2) Either or both of the verbs are polysemous, so the compound becomes polysemous as well, among other reasons.

So this is as you say, more complicated than just three syntactical rules can describe. Originally I thought that compounding is a relatively productive process. Being productive means that there are only very few fairly rigid rules that one follows during the process. The three propositions are empirical observations of how these rules manifested. It seems those rules are not solely syntactical, but semantic as well. I'd not touch that.

I think this thread can be closed. Thank you for all your examples, Toritoribe-san.