Classical 活用: The Six Bases (Bridge Page)

Open any Japanese-language grammar reference, dictionary appendix, or 活用表（かつようひょう, conjugation chart) and you'll meet six labels the rest of this guide has quietly avoided: 未然形（みぜんけい）・連用形（れんようけい）・終止形（しゅうしけい）・連体形（れんたいけい）・已然形（いぜんけい）・命令形（めいれいけい）. They are the traditional "six bases" of Japanese conjugation, inherited from classical grammar and still the standard vocabulary of every Japanese school textbook (学校文法, がっこうぶんぽう). This page is the bridge: it takes 書（か）く through all six and shows exactly which modern form each base produces — so the labels stop being a foreign code and become a map of things you already conjugate.

The six bases, worked through 書く

Each base is a stem-shape that a set of endings attaches to. Here is the complete column for 書く, with the modern forms each base feeds.

Two of the six have "slash" shapes, and that's not sloppiness — it's real. The 未然形 splits into 書か (before ない・れる・せる・ず) and 書こ (before volitional う); school grammar lists both under one label because they're the same slot playing two roles. The 連用形 splits into 書き (before ます・たい) and its euphonic (音便, おんびん) variant 書い (before た・て) — the sound change explained on the te/ta sound-change chart. Everything else is a single clean shape.

Why 終止形 and 連体形 look identical — and once didn't

Here is the fact that makes this page more than a vocabulary list. In modern Japanese, 終止形 and 連体形 are the same shape — both 書く, whether the verb ends the sentence (手紙を書く。) or sits before a noun (手紙を書く人). Beginners reasonably ask why the chart bothers to list two separate rows for one form. The answer is historical: in classical Japanese these two were genuinely different, and the modern language collapsed them.

Watch it happen with a classical 下二段（しもにだん）verb, 受（う）く ("to receive"):

Read the 終止形 and 連体形 rows: classically they diverged (終止 受く but 連体 受くる; 終止 く but 連体 くる; 終止 す but 連体 する). Then — over centuries — the 連体形 won. The attributive shape displaced the old terminal one and became the modern dictionary form. That is why modern する, 来る (くる), and every 一段 verb like 受ける have the shapes they do: they are the old 連体形 promoted to be the whole verb's citation form. The two rows still appear in charts as a fossil of a distinction the spoken language erased.

Fossils you say every day are frozen 連体形

Because the 連体形 was the "before a noun" form, some classical attributives got locked into the modern vocabulary as unchangeable words — and native speakers rarely notice. Two are everywhere:

• いわゆる ("so-called") is the classical 連体形 of the old verb 言（い）はゆ — literally "that which is said."
• あらゆる ("every, all") is the classical 連体形 of 有（あ）らゆ — "all that exists."

Neither has a dictionary form in modern Japanese; they survive only in their attributive shape, always glued to a following noun. They are living proof that the 連体形 was once a distinct base.

已然形 → 仮定形: same slot, opposite meaning

The fifth base carries two names because its meaning flipped. In classical Japanese the base was the 已然形 ("already-realised form"), and 已然形＋ば meant a factual "because / when": 書けば ≈ "because one writes." Hypothetical "if" in classical Japanese used the 未然形＋ば (書か＋ば → 書かば, "if one were to write"). Modern Japanese dropped 未然形＋ば entirely and reinterpreted 已然形＋ば as the hypothetical, so today 書けば means "if one writes." Because the meaning became hypothetical, school grammar renamed the base 仮定形 ("hypothetical form"). The shape (書け) never moved; only the label and the sense did. This is exactly the shift traced on the classical conditional ば page.

The old realis sense still echoes in the concessive 已然形＋ど／ども ("even though"), fossilised in words like されど ("and yet") — literally さ ＋ あれ（已然）＋ ど.

The 未然形 is the launchpad for negation, passive, causative, volitional

The 未然形 (書か-) is the busiest base for auxiliaries — which auxiliary attaches to which base is the whole subject of the auxiliary-attachment table. It feeds the negative ない, the passive れる, the causative せる, the volitional (via the 書こ sub-form), and the classical negative ず — which survives all over modern Japanese in fossils like 思（おも）わず and 絶（た）えず, covered on the classical ず/ぬ page.

For English speakers: bases are stems, not tenses

English verbs have a handful of forms (write, writes, wrote, written, writing), and each is more or less a finished word. The Japanese six bases are not finished words at all — they are stems waiting for an ending. 書か on its own means nothing; it's a launch-pad that becomes 書かない, 書かれる, 書かせる, or 書かず depending on what you bolt on. So the mental model is not "six tenses" but "six sockets": pick the socket the auxiliary demands, then plug in. That's why grammar references specify attachment by base — "ず attaches to 未然形" tells you the socket, and 書か is the plug.

Common mistakes

1. Attaching ば to the 未然形 instead of the 已然形/仮定形. Modern ば rides the 書け base, not 書か. ×書かば is classical (and archaic); modern is 書けば.

2. Building the 五段 volitional off the wrong 未然形 sub-form. The volitional う takes the 書こ sub-form (o-row), not 書か. So it's 書こう, never ×書かう.

3. Hunting for a separate modern attributive form. In modern Japanese the 連体形 equals the 終止形 — the plain dictionary form goes straight before a noun. Don't invent a special ending.

4. Calling the modern ば-base 已然形. In modern school grammar the base is 仮定形; 已然形 is its classical name. Using the classical label for a modern-Japanese analysis mixes up the meaning (realis vs hypothetical).

Key takeaways

• The six 活用形 are stems (sockets), not tenses: 未然（書か／書こ）・連用（書き／書い）・終止（書く）・連体（書く）・已然/仮定（書け）・命令（書け）.
• 終止形 and 連体形 merged in modern Japanese (both 書く); classically they differed (受く vs 受くる), and the 連体形 became the modern dictionary form — which is why する・来る・食べる look as they do.
• いわゆる・あらゆる are frozen classical 連体形 forms, living proof of the old attributive base.
• The ば-base was the 已然形 ("because/when") in classical Japanese and was renamed 仮定形 ("if") when its meaning flipped — same shape 書け, new sense.
• The 未然形 is the launchpad for ない・れる・せる・volitional・ず; grammar references specify auxiliary attachment by base, so knowing the six labels lets you read any 活用表.