I know おう sounds like a long o, but does おお sound the same, or should it be pronounced differently?

  • お sounds kind of like "oh." I think you are thinking of "oo" like "boo". that "oo" sound is closer to う. おお is like ohoh (you can blend them together or not) and おう sounds like oh-oo. my 2 cents.
    – yadokari
    Jun 21, 2013 at 14:46
  • 1
    I think that with a few exceptions, おう is /o:/ within a single morpheme, and /ou/ when it crosses morpheme boundaries. おお is always /o:/.
    – user1478
    Jun 21, 2013 at 15:10
  • 1
    We have a question with some discussion of おう: japanese.stackexchange.com/q/4718/1478
    – user1478
    Jun 21, 2013 at 15:15

1 Answer 1


I think that with a few exceptions, おう is /oː/ within a single morpheme, and /ou/ when it crosses morpheme boundaries. おお is always /oː/. The same thing is true of other kana pairs like こう・こお, そう・そお, and so on.

For example, look at the following pair of words:

  追う  おう  /ou/
  王   おう  /oː/

追う can be divided into the root ow and the verb ending morpheme u. The /w/ disappears before /u/, so we're left with /o/ and /u/ in separate morphemes, pronounced together as /ou/. The same thing happens with other verbs that end this way, so 思う is pronounced /omou/, not /omo:/. In contrast, contains おう within a single morpheme, so it's pronounced as a long vowel /oː/. You can make any number of similar comparisons:

  沿う  そう  /sou/
  そう  そう  /soː/

One tricky case is forms like 行こう. If you follow the traditional analysis and call this the 未然形+助動詞「う」, it looks like 行こ and are separate morphemes, but that doesn't fit with our rule so far--this form is always pronounced /oː/, not /ou/. In fact, separating between 行こ and is an artifact of analyzing Japanese in kana; romanizing makes things a bit clearer. Instead of iko.u, we can divide it into ik.ou, giving us the root ik plus the hortative suffix ou. Once we do so, we can see that ou doesn't cross morpheme boundaries, so it makes sense that it's pronounced /oː/.

The above rule mostly works, but we can come up with some exceptions like 今日 /kjoː/ or おはよう /ohajoː/. What do they have in common? They were historically subject to a class of sound changes called ウ音便. Let's look at these one at a time:

  • 今日 was originally /ke/ + /pu/. The consonant /p/ turned into /Φ/ and then was lost, leaving /keu/. This was subject to the sound change /eu/ to /joː/, giving /kjoː/. Modern kana usage respelled けふ as きょう to reflect this pronunciation, obscuring the fact that it began its life as two separate morphemes.

  • おはよう was originally the honorific prefix /o/ + the root /haja/ + the adjective ending /ku/. Again the consonant before the /u/ dropped out, giving /ohajau/, which was subject to the sound change /au/ to /oː/, giving /ohajoː/.

In both cases, the sound changes created long vowels that cross morpheme boundaries, and modern kana usage respelled them with an お段 kana plus . In cases like these, the pronunciation is /oː/ rather than /ou/.

  • Necropost, but 行こう comes from ウ音便 too. The correct analysis should be 未然形 /~a/ + う + ウ音便; ikau -> ikoo. For some reason traditional grammar uses the unworkable kludge of proposing a 未然形 ending in /o/ though...
    – ithisa
    Nov 26, 2013 at 6:22
  • @user54609 Well, that's not a synchronic process, it's history. It begins and ends life as /oː/ in modern Japanese. Shibatani addresses this problem by proposing a seventh 活用形 with /-o/ that う can attach to. I think the ending is a single morpheme, personally...
    – user1478
    Nov 26, 2013 at 6:40
  • Well, Japanese grammarians are still content to call あって 連用形+て...
    – ithisa
    Nov 26, 2013 at 6:54
  • @user54609 True, some Japanese grammarians are. And if you're content with that, then I see no reason not to describe ikou as underlyingly ikau, as you suggest!
    – user1478
    Nov 26, 2013 at 7:00
  • Well, personally I would prefer to add two 活用形s, a "音便連用形" and "音便未然形". The former would in particular be really useful without constantly invoking diachronic sound changes to explain て, たり, た, たら, etc.
    – ithisa
    Nov 26, 2013 at 7:02

You must log in to answer this question.

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