Sign in

Libre University uses your GitHub account. Signing in is only needed to sit a final test, so the score is kept on your profile.

Clauses inside sentences

English builds "the book that I bought yesterday" out of a relative pronoun, a change of word order and a comma habit, and Japanese does the whole job by putting the clause in front of the noun and stopping. There is no word for "that", "which" or "who", nothing is moved, and the only requirement is the one the previous lesson supplied: the verb inside the clause must be in the plain form.

The slot in front of a noun

Japanese has one rule about modification and it has no exceptions: whatever describes a noun comes before it. You already rely on this without noticing. たかほん is an expensive book, しずかなまち is a quiet town, 日本語にほんごほん is a Japanese book. Adjective before noun, な adjective before noun, noun plus の before noun: three different kinds of modifier, one slot.

A clause goes in exactly that same slot. Nothing new is added to the grammar; a bigger object is simply dropped into a hole that was already there.

ったほん katta hon the book I bought

That is a complete noun phrase, and it can now be used anywhere a noun can. ったほんむ, I read the book I bought. ったほんたかかった, the book I bought was expensive. The clause has not become a sentence of its own; it is doing the work an adjective does.

The cost of the rule is real and worth naming early, because it is what makes intermediate Japanese hard to listen to. The modifier can be as long as you like, and the noun it belongs to arrives only at the end. In きのうえきまえともだちとはなしていたひと, "the person who was talking with a friend in front of the station yesterday", sixteen characters of clause go by before ひと tells you what the whole thing is about. English gets to say "the person" second, out of fourteen words, and then fill in the details. Japanese makes you hold everything in memory until the head noun lands. That is a listening skill, and it is trained by hearing long sentences rather than by learning another rule.

Building a modifying clause

Take any sentence, put its verb in the plain form, and place it in front of the noun. All four plain forms work, so the modifier carries tense and polarity of its own, independent of the main sentence.

ははつくった料理りょうりはおいしかった haha ga tsukutta ryouri wa oishikatta the food my mother made was delicious

まだんでいないほんさんさつある mada yonde inai hon ga sansatsu aru there are three books I have not read yet

田中たなかさんがんでいるまちしずかだ Tanaka-san ga sunde iru machi wa shizuka da the town Tanaka lives in is quiet

Look at what happened to the particle in each case. In the first, the underlying sentence is はは料理りょうりつくった, and when 料理りょうり is pulled out to be the head noun, its を simply disappears. In the third, the underlying sentence is 田中たなかさんがまちんでいる, and に disappears the same way. The head noun leaves a gap inside the clause, and the particle that marked the gap is not written anywhere. English does the same thing with "the town Tanaka lives in", except that English strands the preposition at the end, and Japanese has nothing to strand.

Any role can be the gap. ほんったひと is the person who bought the book, with the subject as the gap. ほんったみせ is the shop where I bought the book, with a place as the gap. ほんった is the day I bought the book. The clause is identical each time and only the head noun changes, which is a useful thing to drill, because it forces you to read the clause as a unit rather than as a sentence that has gone wrong.

Example. Turn these two sentences into one noun phrase meaning "the letter Tanaka wrote": 田中たなかさんが手紙てがみいた.

The head noun is 手紙てがみ, so pull it out of the clause, and its を goes with it. What is left is 田中たなかさんがいた, whose verb is already in the plain past. Put the clause in front of the head and you have 田中たなかさんがいた手紙てがみ. Nothing is reordered and nothing is inserted. Used in a sentence: 田中たなかさんがいた手紙てがみんだ, I read the letter Tanaka wrote.

Now you. Turn 山田やまださんが京都きょうと写真しゃしんった into a noun phrase meaning "the photograph Yamada took in Kyoto", and then into one meaning "the city where Yamada took photographs".

Answer

山田やまださんが京都きょうとった写真しゃしん, and 山田やまださんが写真しゃしんったまち. In the first, 写真しゃしん leaves the clause and takes を with it. In the second, the place leaves and takes で with it, and the head noun まち replaces 京都きょうと. The verb is identical in both, which is the point: the clause does not change shape according to which role the head noun plays.

The particle that changes

There is one substitution to learn and it catches everybody. は cannot appear inside a noun-modifying clause. は marks the topic of a sentence, and a clause sitting in front of a noun is not a sentence and has no topic of its own to mark. Its subject takes が instead.

So ちち新聞しんぶんんでいる, my father is reading the newspaper, becomes 新聞しんぶんんでいるちち when ちち is the head, and becomes ちちんでいる新聞しんぶん when the newspaper is. The moment a clause goes into the modifier slot, any は inside it turns into が. Writing ちちんでいる新聞しんぶん is one of the most recognisable intermediate mistakes there is.

A second substitution is optional and worth recognising when you read it. In a short clause, が may be replaced by の with no change of meaning: わたしつくった料理りょうり is the same as わたしつくった料理りょうり. This works only when the clause is short and its subject is a simple noun, it is more common in writing than in speech, and it never applies to the が of a main clause. Produce が and understand の.

Modifiers that leave no gap

Here is where Japanese does something English cannot, and where translation stops helping.

電車でんしゃおくれた理由りゆう densha ga okureta riyuu the reason the train was late

さかなくにおい sakana o yaku nioi the smell of grilling fish

In neither of these does the head noun play any role inside the clause. 理由りゆう is not the subject or object of おくれる, and no particle has been deleted: the underlying sentence 電車でんしゃおくれた is complete on its own. The clause simply states the situation, and the head noun names some aspect of it. Japanese allows this freely with nouns like 理由りゆう (reason), おと (sound), におい (smell), はなし (story), 約束やくそく (promise) and 写真しゃしん (photograph). English has to insert a preposition or a whole clause, "the reason why", "the smell of ... doing", which is why literal translation produces monsters.

The practical consequence is that you should stop testing modifier clauses by asking which word was removed. The test that actually works is simpler: is there a plain-form predicate immediately in front of a noun? If so, everything from the start of that clause up to the noun is describing it, whatever the internal relationship.

Turning a clause into a noun

The same construction, with a nearly empty head noun, converts a clause into a noun. The two empty nouns are の and こと.

毎日まいにち漢字かんじくのは大変たいへんだ mainichi kanji o kaku no wa taihen da writing kanji every day is hard work

わたし趣味しゅみほんむことです watashi no shumi wa hon o yomu koto desu my hobby is reading books

Once nominalised, the clause takes particles like any other noun: は in the first, and in the second it is sitting where a noun would sit after です. This is what lets you make a whole event the subject or object of another sentence, and it is the last structural piece a B1 speaker is missing.

The choice between の and こと is genuinely fuzzy, and much of the time either is acceptable. Three tendencies do most of the work. is preferred for a concrete event that the speaker directly perceives or is directly involved in, which is why verbs of perception demand it: どもがみちわたるのをた, I saw the child cross the road, where ことをた is simply wrong. こと is preferred for something abstract, reported or general, which is why it appears with できる, 約束やくそくする and ある: 日本にほんったことがある, I have been to Japan. And こと is required after です in a definition, as in the hobby sentence above, where むのです would instead be the explaining んです of the next section and would mean something else entirely.

Example. Which of の and こと fits each blank, and why? (1) となりの部屋へやだれかがうたっている_がこえる. (2) 毎朝まいあさはしる_にした.

The first is a perception verb, こえる, reporting a concrete event happening now in the next room, so it takes の: うたっているのがこえる. The second is a decision about a general practice rather than a perceived event, and it is the fixed pattern ことにする, so it takes こと: 毎朝まいあさはしることにした, I have decided to run every morning. That pattern gets a lesson of its own later, and it never accepts の.

Now you. Which fits each blank? (1) 日本語にほんご毎日まいにち勉強べんきょうする_はむずかしい. (2) ゆきっている_をた.

Answer

The first accepts either, and both are heard: 勉強べんきょうするのはむずかしい is the more conversational, 勉強べんきょうすることはむずかしい the more written and slightly more general. This is the ordinary case, where the choice carries almost nothing. The second is a perception verb with a concrete event, so it must be の: ゆきっているのをた. Getting a genuinely free choice and a genuinely forced one side by side is the useful part of this pair.

んです, and the explaining mood

Now the payoff. Take a plain clause, add の, and add です, and you have the single most common sentence ending in spoken Japanese after ます itself. In speech の contracts to ん, giving んです, and in writing the full のです is usual.

The construction is not a politeness level and not a tense. It marks the sentence as an account of the situation rather than a bare report of a fact. Something is on the table, visible or already mentioned, and this sentence explains it.

どうしてなかったんですか doushite konakatta n desu ka why didn't you come?

電車でんしゃおくれたんです densha ga okureta n desu the train was late (that's the explanation)

あたまいたいんです atama ga itai n desu I've got a headache (that's why I'm like this)

Attachment follows the rule you already have, with one wrinkle. Verbs and い adjectives attach the plain form directly: くんです, かなかったんです, いたいんです. Nouns and な adjectives cannot use だ here and take な instead, giving 学生がくせいなんです and しずかなんです. That な is the same な you already put between a な adjective and a noun, and it is here for the same reason: の is a noun.

Two things learners get wrong with it. The first is using it for a plain statement of fact. Asked 仕事しごとなんですか, answer 教師きょうしです; answering 教師きょうしなんです implies that being a teacher accounts for something the listener has just noticed. The second is that the question form can be blunt. どうしてなかったんですか demands an explanation and can sound like an accusation, where ませんでしたか merely asks. Used well, though, んですが is the politest opening in the language for a request: ちょっときたいことがあるんですが, "there's something I wanted to ask", trailing off and leaving the listener to offer.

Example. You arrive at a friend's flat and they are holding a suitcase. What do you ask, and why is the plain question wrong?

どこかへくんですか, "are you going somewhere?". The suitcase is the situation on the table, and your question asks them to account for it, which is exactly what んです marks. どこかへきますか is grammatical but reads as an out-of-nowhere enquiry about their travel plans, as though you had not seen the suitcase at all. This is the clearest single test of the form: if something in front of you prompted the sentence, it wants んです.

Now you. Rewrite 明日あした試験しけんがあります as an explanation of why you cannot go out tonight, and rewrite 学生がくせいです as an explanation of why you have no money.

Answer

明日あした試験しけんがあるんです and 学生がくせいなんです. The verb ある goes into its plain form before の, and the noun 学生がくせい takes な rather than だ, because の is a noun and nouns are linked to nouns by な here for the same reason しずかなまち takes な. Writing 学生がくせいだんです is the standard error.

Vocabulary

WordReadingMeaning
理由りゆうriyuureason
趣味しゅみshumihobby
手紙てがみtegamiletter
写真しゃしんshashinphotograph
torutake (a photograph)
わたwatarucross
yakugrill, bake
においnioismell
おくれるokurerube late
大変たいへんtaihenhard going, serious
むずかしいmuzukashiidifficult
約束やくそくyakusokupromise, arrangement
試験しけんshikenexamination
いたitaipainful

A clause can now describe a noun, be a noun, or explain a situation. What it still cannot do is say when it happened relative to anything else. The next lesson takes the construction you have just built, puts a time noun such as とき or まえ in the head slot, and gets the whole apparatus of before, after, while and during out of it for almost nothing.