Quantifiers and precise mathematical claims

Begin with the objects and a defined predicate

Start with the three numbers −1, 0 and 1. Ask which have square equal to 1: −1 and 1 work, while 0 does not. Write the allowed set as D = {−1, 0, 1} and the condition as P(x): x² = 1. P is a predicate: it gives a true-or-false condition once an allowed object x has been supplied. We write T for true and F for false.

Before saying every or some, state the domain D: the set of allowed objects. Here ℤ denotes all integers and ℝ all real numbers. Braces list the members of a finite set. Changing the domain can change a quantified claim even when its condition is unchanged: x² = 1 holds at every member of {−1, 1}, but not at every member of {−1, 0, 1}.

Use a total predicate on the declared domain: each allowed input must give one defined classical truth value. The predicate x² = 1 is defined for every real x. The reciprocal comparison Q(x): 1/x > 0 is defined on ℝ ∖ {0}, the real numbers except zero. It is not a real-valued comparison at zero, so zero cannot be silently classified as a false instance of that same predicate.

Restrict the domain before quantifying the reciprocal. Writing a material implication with x ≠ 0 as an antecedent does not itself define the comparison 1/0 > 0. A separately stated total extension could assign a value at zero, but that would be a different predicate. All quantifier equivalences below assume that their predicates are defined throughout the stated domains.

A quantifier binds a variable within its scope

The symbol ∀ means ‘for every’ and ∃ means ‘there exists’. In ∀x ∈ D, P(x), the quantifier binds the occurrences of x in P(x). Parentheses show where its scope ends when the quantified claim is part of a larger formula. An occurrence outside that scope is not bound by that quantifier.

For example, ∀x ∈ ℤ, x + y = 4 still has the free variable y. Until y is assigned or quantified, it is a condition about y rather than one closed claim. In ∀x ∈ ℤ, ∃y ∈ ℤ, x + y = 4, both variables are bound, and the sentence is closed. Once all free parameters and meanings are fixed, a quantified sentence has a truth value.

A bound variable is a local name. Replacing ∀x ∈ D, P(x) by ∀t ∈ D, P(t) preserves the claim if every bound occurrence is renamed consistently and the new name does not capture an existing free variable. In particular, rename the bound x in ∀x ∈ ℤ, x < y to t rather than y: changing it to ∀y ∈ ℤ, y < y would capture the formerly free y and change the meaning.

Every means that no allowed counterexample exists

The statement ∀x ∈ D, P(x) is true exactly when P(x) is true for every member of D. It is false if an allowed member a has P(a) false. Such an a is a counterexample. Showing that one chosen member works cannot establish a universal claim unless the domain contains only that member.

On a finite domain, inspect every distinct member. On an infinite domain, use an argument for an arbitrary allowed input. For example, ∀n ∈ ℤ, n² + 6n + 9 ≥ 0 is true because n² + 6n + 9 = (n + 3)², and the square of any real number is nonnegative. An arbitrary integer n is real, so the argument covers every integer rather than only several tested values.

The word arbitrary matters: the proof has used no special choice of n. A symbol standing for an arbitrary input does not mean a secret numerical sample. Keep every hypothesis on that input visible while deriving the required conclusion.

Some means at least one, and permits all

The statement ∃x ∈ D, P(x) is true exactly when at least one allowed member satisfies P. A verified member a ∈ D with P(a) true is a witness. The existential claim does not say that a is the only witness, nor that any other member fails the condition.

In mathematical logic, ‘some’ means at least one. If every member of a nonempty domain satisfies P, then the existential statement is also true. To say ‘some but not all’, write both requirements: (∃x ∈ D, P(x)) ∧ (∃y ∈ D, ¬P(y)). Here ∧ means ‘and’, and ¬ means ‘not’. The second witness is a failing member; it is not implied by the word some alone.

For D = {1, 2, 3} and P(x): x > 0, both every and some are true. For D = {−1, 1}, some but not all are positive: 1 witnesses P and −1 witnesses its negation. A proposed witness must actually belong to the declared domain; a real solution does not automatically witness a claim restricted to integers.

Exactly one requires existence and uniqueness

The symbol ∃! means ‘there exists exactly one’. The statement ∃!x ∈ D, P(x) says that exactly one member of D satisfies P. First establish existence with a witness a. Then establish uniqueness: if any b ∈ D also satisfies P, prove b = a. Finding one witness alone establishes existence, but leaves open the possibility of another.

Equivalently, unique existence means ∃a ∈ D, (P(a) ∧ (∀b ∈ D, P(b) → b = a)). The arrow → reads ‘if … then …’. Both directions follow from the same requirement: the displayed a must work, and every other working member must be that a. Since P is total on D, this implication involves defined truth values.

For integers, n + 4 = 9 has the witness n = 5, and subtracting 4 shows every solution equals 5. By contrast, x² = 9 has two real witnesses, −3 and 3, so its real solution is not unique. On the nonnegative real domain, only 3 remains. The negation of unique existence means either no witness exists or at least two distinct witnesses exist; it does not mean exactly two.

A witness and a counterexample prove different things

To establish an existential claim, supply a witness and verify its membership and predicate. To refute a universal claim, supply a counterexample and verify its membership and failure. These single objects are decisive because they match the meaning of the respective quantifiers.

To establish a universal claim, cover all allowed objects by a general argument or by complete enumeration of a finite domain. To refute existence, establish that every allowed object fails. An unsuccessful search among several candidates cannot prove that an infinite domain has no witness.

To establish uniqueness, a witness and a domain-wide ‘any other solution equals it’ argument are both needed. To refute uniqueness, either rule out all witnesses or exhibit two distinct allowed witnesses. Record what the evidence establishes before deciding that a calculation has settled the claim.

Every does not create an object

A bounded quantifier may range over an empty set ∅. Then ∀x ∈ ∅, P(x) is true: there is no member violating the condition. This is called vacuous truth. It does not assert that P has an actual instance, and it remains true even when a supplied total condition would be false at every input outside that empty domain.

On the same empty domain, ∃x ∈ ∅, P(x) and ∃!x ∈ ∅, P(x) are false because no witness can belong to ∅. These conventions agree with the meanings ‘no failing allowed member’, ‘at least one satisfying allowed member’ and ‘exactly one satisfying allowed member’.

Therefore a universal statement entails existence only when a nonempty domain is also known. With D nonempty and every member satisfying P, choose a member and obtain a witness. Without nonemptiness, that choice cannot be made. We use explicit bounded domains, including empty ones; this does not require the entire universe of a first-order interpretation to be empty.

Later witnesses may depend on earlier choices

For a total predicate R(x, y) on X × Y, ∀x ∈ X, ∃y ∈ Y, R(x, y) lets the witness y depend on the already chosen x. In ∃y ∈ Y, ∀x ∈ X, R(x, y), one fixed y must work for every x. Read the scopes from outside to inside; these requirements generally differ.

Over integers, ∀x ∈ ℤ, ∃y ∈ ℤ, x + y = 5 is true: given any integer x, choose the integer y = 5 − x. The reversed ∃y ∈ ℤ, ∀x ∈ ℤ, x + y = 5 is false. Its fixed y would have to equal 5 at x = 0 and 4 at x = 1, which is impossible.

A uniform witness does imply a possibly dependent witness: if one y works for all x, use that same y for each x. The reverse implication fails, as the example shows. This forward argument remains valid for empty bounded sets: an existing uniform y already belongs to Y, and if X is empty the resulting universal requirement is vacuous.

The finite matrix below uses the complete domains X = {0, 1, 2} and Y = {3, 4, 5}. Every row has a true entry, so each x has a suitable y. No column is entirely true, so no single y serves all x. The matrix proves these finite claims; the integer-wide conclusion above relies on the formula y = 5 − x, not on the three rows.

Negation switches the quantifier and negates its condition

The symbol ≡ means that the two claims have the same truth value under the stated assumptions. For a total classical predicate on the same domain D, ¬(∀x ∈ D, P(x)) ≡ ∃x ∈ D, ¬P(x). Both sides say an allowed failing member exists. Conversely, such a member prevents every member from satisfying P. Also ¬(∃x ∈ D, P(x)) ≡ ∀x ∈ D, ¬P(x): no satisfying member exists exactly when every allowed member fails. These are two-way equivalences and hold on empty domains too.

Push a negation through each nested quantifier without changing the domains or their order. For example, ¬(∀x ∈ X, ∃y ∈ Y, R(x, y)) ≡ ∃x ∈ X, ∀y ∈ Y, ¬R(x, y). A counterexample to the outer every is one x for which no y works. Likewise, ¬(∃y ∈ Y, ∀x ∈ X, R(x, y)) ≡ ∀y ∈ Y, ∃x ∈ X, ¬R(x, y): every proposed uniform y has a failing x.

Negate the inner condition as well. The negation of x < y is x ≥ y, including equality. For unique existence, a complete negation is (∀x ∈ D, ¬P(x)) ∨ (∃a ∈ D, ∃b ∈ D, (a ≠ b ∧ P(a) ∧ P(b))). If exactly one witness does not exist, either none exists or a first witness has another distinct companion; either alternative rules out exactly one.

Complete means complete for the declared domain

A finite list gives an exhaustive check only when it contains every allowed distinct member. A complete X × Y matrix covers every pair from those finite domains. A few entries from the integers or reals are samples, even if all their calculations are exact.

For example, (n − 1)(n − 3)(n − 5) = 0 holds at n = 1, 3 and 5, but fails at n = 7 because 6 × 4 × 2 = 48. Those successful samples cannot prove the universal integer claim. In contrast, the input n = 3 is a verified witness for the existential integer claim because it makes a factor zero.

A symbolic witness rule can cover infinitely many inputs: for every integer n, the integer m = n + 3 satisfies m − n = 3. An arbitrary-input argument can also establish that no witness exists, such as x² < 0 on the real numbers because every real square is nonnegative. The conclusion must name its domain and the argument that covers it.

Explore domains, witnesses, order and full negations

In the domains scene, compare a total finite predicate with a reciprocal that includes an excluded zero and a comparison with a free parameter. Assigning the parameter supplies a defined instance; it does not bind that variable. Retaining an undefined reciprocal does not supply a value. A withheld evaluation is displayed with its reason rather than as F.

The quantifier scene checks every member of each stated finite domain and reports every, at least one and exactly one separately. It retains all satisfying and failing members, so an empty set of witnesses is different from one witness or several. The empty-domain cases have no invented member or evaluation.

The order scene displays every pair of the chosen finite sets. Read the row-wise dependent witnesses and the column-wise uniform candidates separately. Empty X, empty Y and both empty are distinct cases; the nested truth values follow their explicit scopes.

The negation scene compares each whole claim’s opposite value with its fully transformed unary or nested formula. The diagrams and selected entries illustrate those exact finite domains. The written semantic equivalences and arbitrary-integer arguments establish the broader statements, including domains that cannot be exhaustively drawn.

Worked examples

Try it yourself

1. Identify the bound and free variables in ∀x ∈ ℤ, x + y = 4. Rename the bound variable safely, and decide the closed claim ∀x ∈ ℤ ∃y ∈ ℤ, x + y = 4.

   Show the answer and reasoning

   x is bound and y is free. A safe renaming is ∀t ∈ ℤ, t + y = 4. The closed dependent-witness claim is T.

   Only occurrences in the quantifier’s scope are bound; y has no assignment here. Renaming x to y would capture that free variable and change the formula. For the closed claim, an arbitrary integer x has the integer witness y = 4 − x, and their sum is 4.

2. For D = {−3, −1, 1, 3} and P(x): x² = 1, give the complete truth column and decide every, some, exactly one and some-but-not-all.

   Show the answer and reasoning

   The column is F, T, T, F. Every is F, some is T, exactly one is F, and some-but-not-all is T.

   The squared values are 9, 1, 1 and 9. The two distinct witnesses −1 and 1 establish existence and disprove uniqueness. The counterexample −3 disproves every and, together with a witness, proves some-but-not-all. Logical some alone does not require that counterexample.

3. Prove ∃!n ∈ ℤ, n + 4 = 9. Then decide whether x² = 9 has a unique solution on ℝ and on the nonnegative reals.

   Show the answer and reasoning

   The integer solution is uniquely n = 5. The square equation is not unique on ℝ, but is uniquely solved by x = 3 on the nonnegative real domain.

   Five belongs to ℤ and satisfies the equation; any other solution must equal 9 − 4 = 5. For x² = 9, factoring (x − 3)(x + 3) = 0 gives exactly −3 and 3 over the reals. Both work, so uniqueness fails there. Restricting to x ≥ 0 excludes −3 and retains the verified witness 3.

4. Negate ∃x ∈ ℤ ∀y ∈ ℤ, x + y = 5. Prove the negation directly with a witness rule and retain both domains.

   Show the answer and reasoning

   The negation is ∀x ∈ ℤ ∃y ∈ ℤ, x + y ≠ 5, and it is T.

   Move the negation through ∃ and then ∀, changing them to ∀ and ∃ and negating the equation. For each arbitrary integer x, choose the integer y = 6 − x. Then x + y = 6, which is not 5. This formula covers all x and is a general proof rather than a finite search.

5. With D = ∅, decide ∀x ∈ D, x² < 0, ∃x ∈ D, x² < 0 and ∃!x ∈ D, x² < 0. If X = Y = ∅, also decide ∀x ∈ X ∃y ∈ Y, x = y and ∃y ∈ Y ∀x ∈ X, x = y.

   Show the answer and reasoning

   The unary values are T, F and F. The two nested values are T and F respectively.

   No domain member violates the unary universal, but no member can witness either existential. In the first nested formula the outer universal has no x to test, so it is T. The reversed formula requires an actual y in the empty Y before considering its inner universal, so it is F. Vacuity supplies no object.

6. For X = Y = {0, 1} and R(x, y): x = y, list the complete two-by-two matrix. Decide ∀x ∃y R and ∃y ∀x R, then give the truth value of each full negation.

   Show the answer and reasoning

   The rows are T, F and F, T. The dependent-witness claim is T and the uniform-witness claim is F; their full negations are F and T.

   Each row has its matching y = x. Neither column works for both different x values. The first full negation is ∃x ∀y, x ≠ y, false because every x has its matching y. The second is ∀y ∃x, x ≠ y, true because choose x = 1 for y = 0 and x = 0 for y = 1. All four pairs belong to the declared complete domains.

7. Prove ∀n ∈ ℤ, n² + 6n + 9 ≥ 0. A different learner tests (n − 1)(n − 3)(n − 5) = 0 at 1, 3 and 5 and claims it holds for every integer. Give a counterexample and an existential conclusion those tests do establish.

   Show the answer and reasoning

   The first claim is T because n² + 6n + 9 = (n + 3)² ≥ 0 for arbitrary n. The second universal is F: n = 7 gives 48. Its existential version is T, for example at n = 3.

   The square identity and nonnegativity cover every integer input, including inputs not sampled. Each successful sample in the second calculation makes a factor zero, but at 7 the product is 6 × 4 × 2 = 48. One verified allowed witness proves existence; only a domain-wide argument or exhaustive finite domain can establish every.

What to carry forward

• State the domain and require a defined predicate at every allowed input. An undefined reciprocal at zero cannot be repaired by an unguarded truth-functional implication.
• A quantifier binds occurrences only within its scope. A remaining free variable needs assignment or quantification; safe renaming must avoid capturing it.
• Every needs all allowed inputs; some means at least one and permits all. Exactly one combines a witness with a proof that any other witness is the same.
• A verified witness establishes existence and a counterexample refutes a universal. Complete finite enumeration and arbitrary-input arguments have different, explicitly stated scopes.
• An empty-domain universal is vacuously true, while existence and unique existence are false. Every alone does not establish a witness without a nonempty domain.
• In ∀x ∃y, the witness y may depend on x; in ∃y ∀x, one fixed y must work throughout. Quantifier order and original domain membership matter.
• Negate whole quantified claims by switching each ∀ and ∃ and negating the inner condition, retaining scopes and domains. Negated uniqueness permits no witnesses or at least two distinct witnesses.