ある's Irregular Negative ない

ある has exactly one irregularity, and it lives in a single cell of its conjugation table: the plain negative. Everywhere else — polite, past, te-form, conditional — ある behaves like a perfectly ordinary godan（五段） verb. But where the rules predict ×あらない, the actual negative is the suppletive ない. This page isolates that one cell, shows why it's shaped the way it is, and drills the over-regularization out, because ×あらない is one of the most reliable tells of a non-native speaker.

That is the correct negative of ある. Not ×あらない — a form that no native speaker produces, even though the godan rulebook seems to demand it.

Everywhere else, ある is regular

Start with what isn't broken. Run ある through the standard godan machinery and every form comes out exactly as the rules predict:

Look closely at that last one: ありません is regular. It's built the normal way — the masu-stem あり plus the negative-polite ません — because the whole polite paradigm hangs off the well-behaved あり- stem. So the irregularity is even narrower than "the negative": it's specifically the plain negative. The polite negative is fine.

What the rule "should" produce — and why it doesn't

To feel the irregularity, compare ある with a genuinely regular godan verb that ends the same way, 分かる（わかる）:

The godan rule for the plain negative is: shift the final う-row kana to its あ-row partner and add ない. る → ら, so 分かる → 分から + ない. Apply the same rule to ある and you get あら + ない = ×あらない. That form is grammatically "correct" by the rulebook and completely dead in the language. ある simply refuses it and uses ない instead.

Why ない fills the gap: a historical suppletion

This isn't random. In older Japanese, ない was an independent adjective meaning "non-existent, absent" — a word in its own right, not a piece of ある. ある had no negative form of its own, so the language did the economical thing: it drafted the ready-made adjective ない to serve as ある's missing negative. Linguists call this suppletion — plugging one word's gap with a historically unrelated form (English does it too: go / went, good / better).

Two consequences fall straight out of this history, and both are useful to know:

1. ない inflects like an い-adjective, not like a verb-negative. Its past is なかった, its te-form なくて, its adverbial なく — the い-adjective pattern (い → かった / くて / く), because that's literally what it is. That's why ×ないでした and ×ないだった are wrong: you conjugate ない the way you conjugate 高い → 高かった.

2. The negative suffix ない everywhere else is the same word. When you negate 食べる → 食べない or 行く → 行かない, the ない on the end is that same "non-existent" adjective, grammaticalized into a suffix. So ある's suppletive negative and the universal verb-negation ない aren't a coincidence of sound — they're one morpheme wearing two hats.

The full negative paradigm of ある

Putting the plain and polite negatives side by side, in every tense you'll need:

Notice the plain column is all ない-based (an い-adjective conjugating), while the polite column is all あり-based (the regular verb conjugating). The two paths never cross — which is exactly why the polite ありません feels regular and the plain ない feels irregular. They come from different stems.

Common mistakes

Every one of these traces to a single overzealous instinct: applying the godan あ-row rule to ある. The rule is right for 分かる, 作る, 帰る and every other godan verb — ある is the one exception, and it's exceptional in precisely one cell.

Key takeaways

• ある is a regular godan verb everywhere except its plain negative: あります, あった, あって, あれば, ありません all follow the rules.
• The plain negative is the suppletive ない — never the rule-predicted ×あらない.
• ない is a historical い-adjective meaning "non-existent," which is why it inflects なかった / なくて and why the universal verb-negation suffix ない shares its shape.
• The polite negative ありません is fully regular (masu-stem あり + ません); the irregularity is confined to the plain form.