this is my first question here so please bare with me.

Until the other day I was aware of two different ways use counters

As an adverb.

  • ペンを3本買いました。

As a noun.

  • 3本のペンを買いました。

The other day I found out that you could say

  • 友達3人と出かけました。
  • 犬3匹と遊びました。

But you can't say

  • ペン3本を買いました。X

At first I thought this was something to do with the と particle, but I found out these are also correct.

  • 鍋に水1リットルを入れる。
  • 鍋に水を1リットル入れる。

My question is, what are the rules for placing the quantifier between the noun and the particle? I apologize in advance if this question has been asked before, I was unable to find anything. Thanks.

  • ペン3本を買った is fine.
    – user4092
    Feb 1, 2015 at 7:55

1 Answer 1


Generally there are three choices for the kind of quantifiers addressed here. Although the OP's question uses numeral quantifiers, the same would be true for others, such as すべて, etc.

Attributive position

That is the case where the quantifier, marked by の, precedes the noun, e.g. the OP's third example.


Composition is when the quantifier follows the noun, but not the case marker. See the third to sixth of the OP's examples. OP claims that the fifth example is bad, but I don't think it is (as bad as is claimed).

Quantifier floating construction

Quantifier floating is when the quantifier follows not only the noun, but also the case marker.

 (1) 3本のペンを買いました。 

There the quantifier is subordinate to its host noun. From that position, the quantifier can float beyond its host noun. It then resides in a structurally higher position than in (1). The result is the OP's first sentence, i.e.

 (2) ペンを3本買いました。

Japanese quantifier floating has captured the attention of linguists because it helps identify a subject-object asymmetry. Compare the next two sentences:

 (3)   ペンを学生が3本買いました。
       A student bought three pens.
 (4) * 学生がペンを3人買いました。
       Three students bought pens.

In (3), the quantifier floats, and produces (1). Then the object ペンを scrambles, i.e. moves across the subject 学生が. As a result, the subject now intervenes between the host noun and its quantifier.
Example (4) illustrates that the same is not possible with subject-related quantifiers. The conclusion many linguists draw is that in order for an object to be able to intervene between a subject and its floated quantifier, the subject must be base-generated in a position after the object, i.e. in OSV word order. (4) being ungrammatical, is thus seen as support that Japanese is "truly" a SOV language.

Quantifier floating in Japanese seems to be largely restricted to host nouns marked with nominative (が)or accusative (を)case. Other rare cases are discussed in Shigeru Miyagawa (1989: Structure and Case Marking in Japanese, Academic Press).

In Miyagawa & Saito (eds) The Oxford Handbook of Japanese linguistics link, there is a good section on floated quantifiers. An overview of quantifier floating in other languages can be found here.

Other quantifiers

Note that there are quantifiers that cannot appear in the constructions above. Attribution, composition, and floating are only available for quantifier adjuncts that scope over a noun. There are, however, also quantifier adjuncts and arguments over verbs:

 (5) 亀は100年間生きることがある。
     Turtles may live 100 years.
 (6) 物価は3%上がった。
     Prices rose by 3 percent.

Example (5) shows the quantifier adjunct/adverb 100年間, which scopes over the verb. (5) cannot be rephrased using attribution or composition. Since attribution is not available, floating is neither.
Example (6) shows the quantifier argument 3%. Again, attribution, composition, and/or floating are not available. Another interesting point is that quantifier arguments can express differential and absolute values. Quantifier adjuncts always express absolute values.

  • Thanks for the answer. However, from my basic understanding, the links you provided just seem to explain moving the quantifier beyond the noun. The thing I want to know is when should I place the quantifier in between a noun and a particle?
    – Razzek
    Jan 31, 2015 at 16:57
  • I've edited my answer to better reflect your wish. You want to know when, what I call, "composition" is available. My guess is that it is often available. Japanese is very liberal with nominal composition (quantifiers are nominals). Please check your fifth example again. It may be better than you think. Jan 31, 2015 at 20:18
  • (5) is okay because it adheres to the mutual c-commanding rule, that is, the host NP c-commands the FNQ and the FNQ c-commands the NP. For more info go see "The Oxford Handbook of Japanese Linguistics" and 1989: Structure and Case Marking in Japanese, Academic Press.
    – Herb
    Apr 7, 2017 at 15:02

You must log in to answer this question.

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