
The て form: how it's made and why it matters
The て form has no tense of its own and can't be translated on its own, yet it is the piece of Japanese grammar that opens the most doors: making requests, saying what you're doing, asking permission, joining two clauses.
Learning it pays twice: the rule that builds it is the same one that gives you the plain past, and the constructions built on it cover half the N5 and N4 syllabus.
Why a tenseless form matters so much
On its own the て form doesn't say when anything happens: it takes its tense from the verb that follows. That is exactly what makes it useful — it is a joint. 食べて means nothing finished, but 食べています, 食べてください and 食べてもいい are three complete and very different sentences.
The one exception
行く ends in く but doesn't give 行いて: it gives 行って. It is the only irregularity in building the て form, and it is worth learning on day one because "to go" is everywhere. Every other く verb behaves: 書く gives 書いて, 聞く gives 聞いて.
ている is not the English present continuous
With action verbs it is: 食べている is "I'm eating". But with verbs describing a change, it marks the resulting state: 結婚している isn't "I'm getting married" but "I'm married", and 知っている is "I know". It is the commonest mistake among people translating by ear.
The little song
Almost everyone learns group 1 by chanting うつる → って, むぬぶ → んで, く → いて, ぐ → いで, す → して. It feels childish and it works: the て form has to come out in half a second while you speak, and at that speed the rule has to be automatic, not reasoned.
How it's built
Group 2 drops る and takes て. する becomes して, 来る becomes 来て. All the difficulty is in group 1, where the verb's last syllable decides — and there is exactly one exception in the whole language.
| Ending | Becomes | Example |
|---|---|---|
| う・つ・る | って | 買う \u2192 買って |
| む・ぶ・ぬ | んで | 読む \u2192 読んで |
| く | いて | 書く \u2192 書いて |
| ぐ | いで | 泳ぐ \u2192 泳いで |
| す | して | 話す \u2192 話して |
| 行く | 行って | 行く \u2192 行って(exception) |
The same rules give the plain past: swap て for た and で for だ (読んで → 読んだ). One table learned, two used.
The ten constructions built on it
This is why it repays learning properly: each of these is something you cannot say without it. The first six are N5, the rest arrive at N4.
〜てくださいN5
Making a request: "please wait"
待ってください
〜ているN5
Ongoing action or state: "I'm eating"
食べている
〜てもいいN5
Asking permission: "may I come in?"
入ってもいいですか
〜てからN5
After doing something: "after eating"
食べてから
〜て、〜N5
Linking actions in sequence: "I get up and eat"
起きて、食べます
〜てはいけないN5
Prohibition: "you must not enter"
入ってはいけません
〜てみるN4
Trying something out: "I'll try eating it"
食べてみる
〜ておくN4
Doing in advance, leaving ready: "I'll buy it (for later)"
買っておく
〜てしまうN4
Completion, or regret: "I've gone and forgotten it"
忘れてしまった
〜てあげる / 〜てくれるN4
Giving and receiving favours: "he taught me"
教えてくれた
Make it automatic, not memorized
The て form is learned by producing it dozens of times. With Yukigo you drill it on real verbs, inside sentences that come back through spaced repetition, with audio and instant correction. Free and ad-free.
Free beta, installs in two minutes. How do I install it?
Frequently asked questions
Because て is the base form and で is its voiced variant, appearing after verbs in む, ぬ and ぶ. It is the same phenomenon that makes the past either た or だ: 読んで and 読んだ follow one logic.
Neither: it is neutral. Politeness comes from what you attach — 待ってください is polite, while 待って on its own is a casual imperative used between friends.
い-adjectives become 〜くて (安くて美味しい), な-adjectives take で (静かできれい). Same linking job, different construction.
Yes, and it is among the most present topics: 〜てください, 〜ている and 〜てもいい already appear in the first level's syllabus, in grammar and in listening alike.
Don't memorize the table, produce it: take twenty verbs you already know and transform them out loud, every day, until you stop thinking about it. Recognizing it is easy; getting it out mid-sentence is the only thing that counts.