I love in Japanese, how adjectival clauses are just added in front of nouns like adjectives.

The pizza that I ate = 私が食べたピザ

But last night I became confused... In english we have words to link the clause to the main sentence (my grammar knowledge and terminology is pathetic, I know).

a) The ball that I kicked
b) The pen with which I wrote
c) The building that I entered
d) The garden in which I played

Lets first look at a different form of those

A) I kicked the ball
B) I wrote with the pen
C) I entered the building
D) I played in the garden

In the Japanese we use different particles に、を、で to describe what part each word plays in the sentence. Just like the english (with, in).

When we change the english to the first form (a,b,c,d). We use (with which, in which, that). But from what I've seen, in the Japanese form the pattern is indistinguishable between a,b,c,d

a) The ball that I kicked
b) The pen with which I wrote
c) The building that I entered
d) The garden in which I played

Is that correct? It seems to lose any idea of what part the noun played in the action - it's ambiguous. To me (b) reads: "The pen I wrote". But maybe it could be read as "The pen I wrote on" or "The pen I wrote with".

Am I doing it right? Or am I missing something. It may be that the part a noun plays in the action is never ambiguous, but I'm not sure if that's true.


Relative clauses in English

In English, relative clauses are formed by removing something, leaving behind a gap.

1a) I kicked the ball
1b) I wrote with the pen
1c) I entered the building
1d) I played in the garden

In each example, we'll pull out the bolded noun phrase, leaving behind a gap:

2a) the ball [ which/that/∅ I kicked ____ ]
2b) the pen [ which/that/∅ I wrote with ____ ]
2c) the building [ which/that/∅ I entered ____ ]
2d) the garden [ which/that/∅ I played in ____ ]

Note: the ∅ symbol indicates the possibility of using neither which nor that

In examples a and c, the gaps are the objects of the verbs kicked and entered. In examples b and d, the gaps are the objects of the prepositions with and in.

The gap in each case is indicated by a relative pronoun (who, what, which, how, where, when, etc.), which is inserted at the beginning of the relative clause. In some situations, this pronoun can be replaced with the less specific that, and in some situations that can be deleted entirely.

As it happens, in all four of your examples the appropriate relative pronoun was which; in all four it can be replaced with that; and in all four that can be deleted entirely.

In two examples (b and d) the preposition before the gap can be optionally pied piped—moved before the relative pronoun which—but if you do this, you can't replace it with that:

3b) the pen [ with which I wrote with ____ ]
3d) the garden [ in which I played in ____ ]

Because in many cases that can be deleted, you can construct examples in English without any overt indicator of where the gap is:

4a) the ball [ ∅ I kicked ____ ]
4b) the pen [ ∅ I wrote with ___ ]
4c) the building [ ∅ I entered ____ ]
4d) the garden [ ∅ I played in ____ ]

Although there are no words "to link the clause to the main sentence" in examples 4a through 4d, I'm sure you have no problem understanding these examples. You immediately understand where the gap is, even if you don't consciously think about something called a "gap"—or indeed about grammar at all!

Relative clauses in Japanese

In Japanese, not every relative clause has a gap, but most do.

However, there are no relative pronouns like who, what, where, etc.; there are no words like that; and there is no pied piping. This makes gapped relative clauses in Japanese fairly simple. They're very similar to the English examples 4a through 4d above. Since you can understand those, I'm sure you can understand relative clauses in Japanese, too!

Let's take a look at your examples. But before we can relativize them, let's replace the topic particle は with the case particle が. Why? Because relative clauses in Japanese don't contain topics, so you need to use the version with が to turn them into relative clauses.

Okay, here are your non-relative examples with は replaced:

5a)   私 ボール 蹴った
5b)   私  ペン 書いた
5c)   私  ビル 入った
5d)   私   庭 遊んだ

In each example, we have two case-marked nouns. (That means they're marked with case particles like が, を, で, に, and so on.) Just like in English, we can pick one of these to remove, leaving behind a gap:

6a) [ __ ボール 蹴った ] 
6b) [ __  ペン 書いた ] 
6c) [ __  ビル 入った ] 
6d) [ __   庭 遊んだ ] 

Now we have a relative clause! The head noun in each example is 私, which has been pulled out of the relativized clause. Because in each example the case particle が has gone missing, nothing tells us what role is played by the head noun 私—we need to figure it out from context.

We can tell, for example, that in 6a the relative clause already has an を-marked noun, so the role played by 私 cannot be an を role. We must guess what the role is from the set of possible roles the relative clause does not contain, and then use our common sense to decide.

Of course, we could relativize the other element, as in your examples:

7a) [ 私 ____ 蹴った ] ボール
7b) [ 私  ___ 書いた ] ペン
7c) [ 私  ___ 入った ] ビル
7d) [ 私   __ 遊んだ ] 

Take a look back at sentences 4a through 4d again. You can understand those, right? They're just like 7a through 7d—there aren't any words specifying the role the head noun plays in the relative clause! And if you can understand those sentences in English, I'm sure you can understand their equivalents in Japanese. In each case, you can figure out which case role makes sense: を, で, に, and で.

There's another process we need to talk about to describe relative clauses in Japanese, by the way. It's called が/の conversion! When you have a noun phrase marked with が in a relative clause, you can optionally mark it with の instead. Let's rewrite the examples from 7a through 7d using の:

8a) [ 私 ____ 蹴った ] ボール
8b) [ 私  ___ 書いた ] ペン
8c) [ 私  ___ 入った ] ビル
8d) [ 私   __ 遊んだ ] 

But if the relative clause contains an を-marked noun phrase, you can't replace が with の. Example 9a is fine, but example 9b is ungrammatical with the intended meaning:

9a) [ __  ジョンが 本を 買った ] 
9b) [ __ *ジョンの 本を 買った ] 

Overall, forming these case-gapped relative clauses in Japanese is relatively simple. Although their interpretation is sometimes ambiguous, you can usually figure it out from context. In particular, looking at the set of case roles already present in the relative clause can help you narrow down where the gap is.

Gapless relative clauses

However, Japanese also contains relative clauses without a gap. In these gapless relatives, the head noun is in some way semantically restricted by the content of the relative clause. You'll commonly see these with formal nouns like こと:

10a) [ 東京に行った ] ことがない

In this example, the relative clause 東京に行った gives the formal noun こと a specific meaning. But obviously, there's nowhere in the relative clause to insert こと; it has no gap!

You'll also see this type of relative clause with lexical nouns. I'll borrow a few examples from Timothy Baldwin's thesis, along with his English translations:

11a) [ 勝つ     ] 意志 the will to win
11b) [ 魚を焼く   ]   smoke from grilling fish
11c) [ 学校に行った ] 帰り on the way back from school

In each example, the precise semantic relationship between the relative clause and the head noun is slightly different, but it can be inferred from context. But in each example, there's no gap to be filled!

So to understand a relative clause in Japanese, ask yourself these four questions:

  1. Is there a gap? Most of the time there is, but there isn't always!
  2. If so, which case roles make sense with the verb? (It can't be を with 寝る, for example.)
  3. Of those, which roles are already filled? (Where can the gap not be?)
  4. Of the remaining possibilities, which role makes the most sense in context?

Relative clauses are very common in Japanese, and the more you read, the more you'll probably do all four steps automatically. Sure, sometimes things will be ambiguous and you'll have to think about it, but most of the time you'll probably be able to figure it out from context.

My advice? Just get used to spotting the possible relationships, and you'll be well on your way to understanding relative clauses in Japanese!

  • 2
    I have to say this is one of the coolest answers I've ever read. Excellent comparative explanation. – Darius Jahandarie Feb 20 '14 at 4:38
  • 2
    +1. But: 4a through 4d [...a]re just like 7a through 7d—there aren't any words specifying the role the head noun plays in the relative clause -- I see what you're trying to do, but IMO this is a bit unconvincing since you used wrote 'with' and played 'in' in two of those sentences. I think it certainly is easy enough to infer the role, but I don't think English gives you enough evidence/analogy for this since it requires those words "with" and "in" to be present for a grammatical sentence. Put another way, you cannot delete e.g. "where" from the building 'where' I played. – Hyperworm Feb 20 '14 at 5:27
  • 1
    @Hyperworm That's true, it's not a perfect analogy. Thank you for pointing that out! I also wanted to avoid presenting the precise rules in English because they're quite complicated and I hoped the details weren't necessary here--e.g., where cannot be replaced with that or omitted (as you say) unless the head NP strongly suggests a place ("a place where/that/∅ you can relax" but "the web page where/*that/*∅ the claim was first made")--and also because I didn't want to reproduce a reference grammar of English on a site dedicated to Japanese :-) – snailboat Feb 20 '14 at 5:30
  • Interesting. I hadn't noticed that "where" can be deleted in some cases -- but as you say, that seems to be getting slightly off-topic... Thanks for expanding :) – Hyperworm Feb 20 '14 at 5:50
  • 1
    @rewolf It's true that English often has more information than Japanese does as to where the gap is. I really just want to point out that there are times when English gives no information about the gap, as in "a place [ Ø you can relax ____ ]". Here, you could optionally have where, which does indicate the role of the head noun (in this case indicating that a locative adjunct is gapped), but you can understand it just as easily without that pronoun. It's not entirely analogous to で, of course; it's different syntactically. But in any case, I'm glad the answer was helpful :-) – snailboat Feb 20 '14 at 22:53

Certainly b) is ambiguous. 私が書いたペン may mean "my illustration of a pen; I draw the figure of a pen on a paper". If I want to be precise, I will be more verbose as 私が、それを使って文章を書いたところの、ペン.

  • 5
    I would say 私が使ったペン... – user1016 Feb 19 '14 at 16:19

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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