How to Negate a Quantified Statement

There is one rule, it is entirely mechanical, and it works on statements of any length. Most lost points here come from stopping halfway or from quietly reordering the quantifiers.

The short answer

Two equivalences do all the work:

In words: the denial of "every x satisfies P" is "some x fails P". The denial of "some x satisfies P" is "every x fails P". These are De Morgan's laws for quantifiers, and they are the same idea as the propositional versions — a universal statement is a long conjunction, an existential one is a long disjunction.

From those two lines comes the procedure:

1. Push the ¬ inward, left to right.
2. Flip every quantifier it passes. ∀ becomes ∃, ∃ becomes ∀.
3. Leave the order of the variables alone. You are moving the negation past them, not rearranging them.
4. Negate the innermost predicate last, once the ¬ has nowhere further to go.

Nothing else is involved. If your answer still has a ¬ sitting in front of a quantifier, you are not finished.

Why "not all" is not "all not"

This is the distinction the rule protects, and it is worth seeing before the symbol pushing.

Take "Every bird flies": ∀x(Bird(x) → Flies(x)).

Its negation is ∃x(Bird(x) ∧ ¬Flies(x)) — some bird does not fly. One penguin is enough.

Compare that to "Every bird does not fly": ∀x(Bird(x) → ¬Flies(x)) — no bird flies at all. That is a far stronger claim, and it is not the negation of anything on this page.

The conditional hiding inside

Almost every universally quantified statement in a real course has the shape ∀x(P(x) → Q(x)) — "all P are Q". So negating one nearly always requires the propositional fact:

Chain it with the quantifier rule:

¬∀x(P(x) → Q(x)) ≡ ∃x ¬(P(x) → Q(x)) ≡ ∃x(P(x) ∧ ¬Q(x))

Read the result: there is an x that satisfies the hypothesis and fails the conclusion. That is the definition of a counterexample. The arrow becomes an "and" — it does not survive the negation, and leaving it as → is the single most common error in this topic.

The mirror case is worth having too. "Some P is Q" is ∃x(P(x) ∧ Q(x)), and its negation is ∀x(P(x) → ¬Q(x)) — the "and" becomes an arrow going the other way. The connectives are not decoration; ∀ pairs with →, and ∃ pairs with ∧, and swapping them produces statements that are simply wrong. ∀x(P(x) ∧ Q(x)) claims everything in the universe is a P, which is almost never what was meant.

A full worked example

Here is the kind of definition that gets negated on an exam. A function f: ℝ → ℝ is periodic if

∃T > 0 (∀x ∈ ℝ, f(x + T) = f(x))

Negate it. Work outward to inward, one step per line, and change exactly one thing per line.

In English: f is not periodic if, for every positive T, there is some x where shifting by T changes the value. That is exactly what "no period works" ought to mean, which is the sanity check — a negation you cannot read back in English is usually a negation you got wrong.

Two details worth noticing. First, the bound "T > 0" is a restriction on the domain, not a separate proposition, so it rides along unchanged; you do not negate it into "T ≤ 0". Second, the variables stayed in the order T then x. Flipping them to ∃x ∀T would assert something different and much stronger.

Order matters, and the negation never changes it

∀x ∃y and ∃y ∀x are not interchangeable, and this is the other half of the topic.

Over the positive integers with the predicate "x divides y":

• ∀x ∃y (x | y) — every x divides something. True: take y = x.
• ∃y ∀x (x | y) — there is a single y divisible by everything. False: for any candidate y, the number y + 1 is larger than y and cannot divide it.

Same predicate, same variables, opposite truth values — the only difference is which quantifier is on the outside. In ∃y ∀x, the y is chosen first and must work for all x. In ∀x ∃y, the y is chosen after x and may depend on it.

So when you negate, move the ¬ through the prefix and flip each quantifier in place. Never sort them, never group them, never move an ∃ to the front because it looks tidier.

Where these go wrong

• Stopping halfway. An answer containing ¬∀ or ¬∃ is not in negated form. Push all the way to the predicate.
• Keeping the arrow. ¬∀x(P(x) → Q(x)) is not ∃x(P(x) → ¬Q(x)). The conditional becomes a conjunction: ∃x(P(x) ∧ ¬Q(x)).
• Negating the domain restriction. In ∀x > 0, the "> 0" says which x are under discussion. It stays put. Turning it into x ≤ 0 changes the universe, not the claim.
• Reordering quantifiers. ∀x∃y and ∃y∀x differ. Negation flips each quantifier where it stands and preserves the sequence.
• Pairing the wrong connective. "All P are Q" is ∀x(P(x) → Q(x)), never ∀x(P(x) ∧ Q(x)). "Some P is Q" is ∃x(P(x) ∧ Q(x)), never ∃x(P(x) → Q(x)) — the latter is satisfied by any x that simply is not a P.
• Negating only the innermost part. A negation applies to the entire statement, quantifier prefix included. Changing just the final equation while leaving ∃T and ∀x untouched is the most common half-answer.

Frequently asked questions