In the compound verb 手伝う, we see the following sound change:


Unexpectedly, the た from つたう turns into だ. 連濁, at least with the 手+伝う analysis, can't account for this, because it's not the first consonant of 伝う.

What happened here?






  • 和訳で「以外に」と書きましたが、お探しの漢字は「意外」でしょうか (^_^) あと、「意外なことに」という言い方が自然に聞こえるかもしれません (^^♪ しかし、和訳を付けた質問はあんまりありませんし、つけてありがとうございました!
    – Sjiveru
    Nov 27, 2015 at 21:28
  • 訂正ありがとうございます。修正しました。 Nov 27, 2015 at 21:39
  • Maybe it's a hybrid of rendaku and metathesis?
    – Earthliŋ
    Nov 28, 2015 at 6:53

2 Answers 2



There's two possible explanations here, and from a little bit of searching, it doesn't seem like anyone knows which is right (or that anyone's really thought much about it).

Option 1: Rendaku plus metathesis (as Earthling suggested). te + tsutau > tedzutau > tetsudau. There might be another example of this happening with 舌鼓 - shita + tsutsumi > shitadzutsumi > shitatsudzumi (though for this one the unmetathesised version seems to be valid in nonstandard dialects).

It's possible that this is a regular change restricted to words where the second half is at least tVtVCV. It might just be irregular.

Option 2: 手伝う is not actually 手 + 伝う, but something else entirely; with the kanji spelling the result of a folk etymology.

Those are the most likely possibilities.



一。 「不規則連濁」(連濁プラス音位転換、Earthlingさんの仮説)。て+つたう → てづたう → てつだう。「舌鼓」はもう一つの例なのかもしれません - した+つつみ → したづつみ → したつづみ (けど「したづつみ」という非音位転換した方も標準外の方言で可能みたいです)。


二。 「手伝う」という言葉は実は「手+伝う」ではなくて、別の由来のある言葉だということ。その場合は漢字の由来は民間語源のはずです。


  • 1
    Very interesting! But hm... 鼓 is つづみ, even without the 舌, right?
    – dainichi
    Nov 30, 2015 at 20:46
  • 手づかみ is a counterexample to the tVtVCV rule, but of course, I realize it might not be productive anymore.
    – dainichi
    Nov 30, 2015 at 20:49
  • @dainichi - that's tVkVCV, though. (I meant the second part.) You're right about 鼓, the site I was looking at must have mentioned it in error.
    – Sjiveru
    Nov 30, 2015 at 21:38

I think this is not due to a shifted rendaku, but rather due to a change in pitch accent that occurs in all compound verbs (つたう{LHH} → てつだう{LHHL}).

It’s difficult to say てつたう{LHHL} without the た sounding voiced, because the ‘a’ in た needs to be fully voiced to clearly pronounce the pitch drop, at which point it sounds like the ‘t’ is voiced when you speak quickly enough.

Another example of this pronunciation phenomenon would be 恋い慕う. If you pronounce it as こいしたう{LHHHL} it sounds like こいしだう, although the orthography hasn’t changed to reflect that in the case of this word.

  • Although I guess one would expect rendaku here: 天伝ふ, 百伝ふ are both づたう, so that puts a dent in this theory, though I do still like it (since it’s simply easier to pronounce with a だ there). Jun 26, 2020 at 21:10
  • I had a look at the 1603 Nippo Jisho. I'd read before that Middle Japanese allowed for different consonants in coda position, and sure enough, their entry for tetsudai appears instead as tetdai. Entry here, left-hand column. Meanwhile, tsutau is here, right-hand column, realized as Tçutai, tò, òta, showing that つ was realized as such at the start of the standalone verb. Perhaps the collapsed vowel explains the shift? Jun 27, 2020 at 0:13

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .