There are quite a few examples were tsu actually means no. まつ毛・睫毛・睫 matsuge actually comes from 目の毛 me no ke (ma is another pronunciation for 目 me) 沖つ宮 okitsumiya(aslo written 沖津宮, but, as Meikyou tells us, 沖津 stands for 沖つ which in turn stands for 沖の) 中つ国 nakatsukuni, Middle-earth (like Tolkien's middle-earth)

I'm thinking:

  1. maybe みっつ、よっつ、むっつ、よっつ and so on are just later modifications in the pronunciation of みつ、よつ、むつ、よつ etc., and
  2. you can find these original pronunciations (みつ、よつ...) only as prefixes (as far as I know, but actually this is not that relevant), e.g. 三つ巴、三ツ星、三つ指、四つ切...

...so what I'm saying is it would make sense if つ (the generic classifier) actually meant の like in the words at the beginning of this post, meaning it was a particle with very similar use. That would obviously imply numerals had been used without a counter-word (but with a particle1), but that's not such a surprise, because there's the same pattern in a few other words, like 二の腕、一の宮、三の糸、二の丸... Also, looking at the pronounces of numerals, is it possible tsu always followed kun'yomi, while no followed on'yomi?

Have anyone here ever tackled this problem (or hopefully know or can find the answer)?


1 I add this because using a numeral without a counter-word and without a particle is even less of a surprise, since, due to lack of counter-words(?), you get expressions like 三次元 or 3説

  • 1
    三つ… is "three …" while 三の… is "the third …". – user4092 Mar 7 '18 at 3:16
  • The short answer is yes. It is known that in ancient Japanese, つ could actually mean the same meaning as の. (cf. 学研全訳古語辞典 ) – Faily Feely Mar 7 '18 at 15:39
  • @user4092 Still, in the new 新字源 you can find 七つ時 meaning the seventh hour (I dont't remember the right animal) under the entry for 七. Also I don't mean they are exactly the same thing... Just wondering if the つ we use as a 助数詞 nowadays came from the つ we see in 中つ国 which is so similar to の – Kokoroatari Mar 10 '18 at 22:25

Did つ (助数詞) actually mean の?


Did つ (not as a 助数詞) actually mean の?

Yes, in some ways, sometimes.

Breaking things down

Japanese has several different つ elements that can come at the end of things.


Indicates completion of verb action, also known as the perfective aspect. This was much more productive in older stages of the language. It persists in modern Japanese as the て conjunctive ending, in certain set construction patterns like 行き​[つ]{●}​戻り​[つ]{●}​, and in the つつ continuative ending as in 変わり​[つ]{●}​​[つ]{●}​ある.

The modern past-tense ending た evolved from classical たり, which in turn was a contraction of the 連用形{れんようけい} or continuative form て of this same つ, + あり: て + あり → たり → た.


Affixed to terms describing amounts or portions, and indicating that each amount or portion is of equal size. Equivalent to, and possibly derived from, ずつ. That said, ずつ is from older spelling づつ which is listed in some dictionaries as a reduplication of the counter つ (see below).


A counter, suffixed to native-Japanese numbers as a nominalizer (i.e. it forms nouns). In modern Japanese, you generally need a の between [NUMBER]つ and a noun, but in Old Japanese, this could be immediately followed by a noun.

My dead-tree copy of Shogakukan's 国語{こくご}大辞典{だいじてん}, 大辞泉{だいじせん} at Kotobank, and 大辞林{だいじりん} at Kotobank, all list this as the 訓読み for the counter suffix 個 or 箇, more commonly read as ko or ka.


This is the one you're thinking of.

Described in Daijirin (third つ entry down) as:

Attaches to a nominal or a something acting as a nominal, and forms an attributive modifier.

This is like modern uses of の when one noun, or noun-like term or phrase, is used to modify another. 青い車の人 "a blue-car person → a person of the blue car", 犬飼いの人 "a dog-keeping person → a person who is keeping a dog", etc.

It's a fine distinction, but it appears that Old Japanese つ in this particle usage did not indicate possession (one thing belonging to another): that was expressed by either が or の. According to Shogakukan, the distinction between the two was that が indicated a closer relationship than の.


If you're interested in the question of where the heck did this all come from, a fellow named Bjarke Frellesvig has been doing some interesting research into the origins of the Japanese language. If memory serves, he has a theory that the ancient precursor to the Japanese language may have had two kinds of copula (like English "is" or "to be"), one that started with n and gave rise to the particles に and の and the perfective ending ぬ, and one that started with t and gave rise to the particles て and と and the perfective ending つ. Many moons ago, snailplane recommended Frellesvig's A History of the Japanese Language, which I have yet to dive into. I suspect this would be a good deep-dive book for the Japanese learner who is also a hardcore word nerd.


The final つ in みっつ is the noun-forming counter. According to Shogakukan, the modern 促音{そくおん} (geminate) readings like みっつ、よっつ are just sound shifts from earlier みつ、よつ, and are not evidence of みつ + つ.

We know that these number words are nouns as they can be followed by particles such as は and に and を, in constructions that require nouns or noun phrases. This holds true even back to the Man'yōshū, where we can find instances of ひとつ immediately followed by these particles.

Notably, Japanese does not always require a の between one noun and another, when the first noun is modifying the second one. In the modern language, there are examples of this all over the place -- but these are more commonly thought of as compounds, words like 日本猿{にほんざる} (Japanese macaque) or 雲形{くもがた} (cloud shape) or 兎小屋{うさぎごや} (rabbit hutch). Similarly, 七{なな}つ (seven, the seventh of something) + 時{とき} (hour) = 七{なな}つ時{どき} (the seventh hour in the old timekeeping system).

  • It was a very nice answer and it certainly will be of most interest to a lot of readers... and I'm so glad I found another Old Japanese fan ...but I actually knew more or less anything you said, and I think you din't actually answer my question (aside from that "No."). When you say "you generally need a の between [NUM]つ and a noun, but in Old Japanese, this could be immediately followed by a noun." ...it actually means what I was saying: if you look at つ like it was a (slightly different) の, it would make sense. Btw as I said above I just read a 七つ時 which seems to fit the pattern (cfr. 二の腕) – Kokoroatari Mar 10 '18 at 22:39
  • Thx for let me know about A Histoy of th Japanese Language, I'll be sure to give it a try – Kokoroatari Mar 10 '18 at 22:42
  • @Kokoroatari -- I've updated with an addendum to try to directly address your comment. – Eiríkr Útlendi Mar 11 '18 at 22:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.