I am trying to figure out head-internal relative clauses. A paper I looked at presented two versions of the same sentence, one with the head word (りんご) outside the relative clause:


And one version where it is inside the relative clause:


Do these two sentences mean exactly the same thing, or are there differences in nuance? When are head-internal relative clauses usually used in Japanese?

2 Answers 2


They are slightly different, if not much. The former sounds saying a fact relatively objectively.

On the other hand, the latter rather means "although an apple was on the plate, s/he stole it" and it sounds somehow accusive in the sense that it should have been there. In grammar for old Japanese, a similar form is considered a conjunction.

When are head-internal relative clauses usually used in Japanese?

I forgot to answer to this part. Speaking of this usage of を, not specifically to "head-internal relative clause", I personally use it like a weaker version of のに in the point of paradoxical sense, mostly in the form of "…、それをさぁ~、…。". So, I think I've been using it to some degree. I personally am used to the form of HIRC with を too. But I'm not sure other native speakers share that sense.

  • I personally don't feel the sentence with the head-internal relative clause is more disapprobative than the other. It would seem mistaken at least to ascribe an adversative sense to the HIRC, given it can be used in sentences like「財布を落としたのを拾ってあげた」「困っているのを助けてあげた」 unproblematically.
    – goldbrick
    May 19, 2019 at 2:47
  • @goldbrick Good point. It's only a result of answering to the specific example. There's the same problem when we call 「名前を呼ばれた」"adversative passive".
    – user4092
    May 19, 2019 at 16:09

I could have the wrong end of the stick here, but your second example doesn't make sense to me.

Part 1: regular relative clause

Let's look at your first example first.


At its core (haha, pun not intended), we have:

I pilfered the apple.

The relative clause portion tells us more about the apple:

[(it) was on the plate]

Looking at the whole utterance, we get:

I pilfered the apple [(that) was on the plate].

Part 2: nominalized clause

Now let's look at your second example.


The core here is more complicated, because we don't have a simple concrete noun as the object of our verb くすねた. Instead, we have the の, which here is used to nominalize (make a noun out of) the entire preceding clause.

So let's look at the embedded clause.

[The apple was on the plate]

Okay, simple enough.

After this, though, we have that の, turning our entire embedded clause into a nominalized phrase. This can be a bit messy to translate into English; it comes through somewhat similar to "the fact that", or "the act of", or sometimes by turning a verb into the "-ing" form. Some examples:

I like (the act of) eating ramen.

Note that this is different from just "I like ramen". We're not talking about "ramen" as the main noun, but rather about the whole clause that contains "ramen" -- in this case, about "eating" it.

I didn't know (the fact) that Michiko went to Tokyo.

Again, this is different from "I didn't know Michiko". We're not talking about "Michiko" as the main noun, but rather about the whole clause that contains "Michiko" -- in this case, that she "went to Tokyo".

Looking again at the whole second example sentence then:


The key is that it turns the whole embedded clause into a kind of noun: we're not talking about "the apple" anymore, but rather the fact that "the apple was on the plate".

Because of the verb here, くすねた / "pilfered", nothing quite makes sense -- just due to the meaning of the words, this doesn't fit together. The best translation I can come up with would be something like:

I pilfered (the fact that) [the apple was on the plate].

...??? That doesn't make sense in English. Nor does the Japanese make sense. (At least, as I understand it.)

If you change the verb from くすねた to 見た, that would work:

I saw (the fact) that [the apple was on the plate].

The paper in question appears to be this one:

This seems to have been written by a native speaker of Japanese, which makes that second sample sentence a bit of a head-scratcher for me.

For the verb くすねる, I'm only aware of the sense "pilfer, filch, sneak, pinch, swipe", with the core underlying meaning of "to steal something sneakily". There might be a sense of くすねる that I'm missing, which could make the second sample sentence work better.

  • 3
    The second sentence is fine like user4092 says in their answer. May 17, 2019 at 23:58
  • @DariusJahandarie, I'm passingly familiar with the 古文 usage (similar to modern のに), but I'm unfamiliar with that usage in the modern language. How does it parse? May 18, 2019 at 0:20
  • 1
    HIRCs (sometimes IHRCs) have been discussed quite often in the literature – see e.g. The mechanism of inverted relativization in Japanese: A silent linker and inversion (Hiraiwa 2012) – but their grammatical status is somewhat controversial. Educated native speakers use them, even in professional writing, but not everyone is happy with the way they sound. If you ask for acceptability judgments, native speakers sometimes rate them poorly.
    – user1478
    May 18, 2019 at 1:09
  • @snailboat, is the のを here similar in function to the 古文 variety, parsing out as similar to のに? If so, I'd argue that this isn't a relative clause at all... May 18, 2019 at 1:10
  • 2
    @EiríkrÚtlendi Consider the examples from the first page of this paper, perhaps: lingref.com/cpp/wccfl/31/paper3018.pdf
    – user1478
    May 18, 2019 at 1:38

You must log in to answer this question.

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