Evanalysis
1.3Estimated reading time: 29 min

1.3 Quantifiers and negation

Read predicates, domains, and quantifiers carefully, then negate quantified statements without losing scope, dependencies, or meaning.

Course contents

Motivation

Propositional logic treats a complete sentence as one proposition, but it does not expose the structure of “every human is mortal” or “some human dislikes cheese.” Predicate logic adds variables, predicates, and quantifiers so an argument can say which objects are discussed and how many satisfy a condition. This language expresses definitions of primes, prerequisites, functions, and relations without leaving the objects or conditions implicit.

The reliable habit is to read a formula from the outside in. First identify the domain and the scope of each quantifier. Then translate the logical connective inside that scope. When negating, work from the outside inward and preserve every condition attached to a variable.

To formalize the argument about humans and mortality, write ∀x (H(x)→M(x))\forall x\,(H(x)\to M(x)), together with H(s)H(s), to derive M(s)M(s) for the named person ss. The universal premise is instantiated at ss before the implication is used.

Definitions

Definition

Predicate and proposition

A predicate is a formula whose truth may depend on variables. An occurrence is free when no quantifier controls it. Assigning values to all free variables lets us evaluate the formula under that assignment, but leaves those occurrences syntactically free. Binding every free occurrence with a suitable quantifier produces a closed sentence.

Definition

Domain and assignment

The domain is the set from which quantified variables are chosen. An assignment gives values to free variables; thus P(x,y)P(x,y) with x=1,y=2x=1,y=2 is false when P(x,y)P(x,y) means x=yx=y. In ∀x P(x,y)\forall x\,P(x,y), xx is bound but yy is free. Quantifiers control variables locally, so an inner quantifier may reuse a letter without referring to an outer variable with that letter.

The domain is part of the statement. The expression x2=1x^2=1 has one natural number solution and two integer, rational, or real solutions. A formula such as ∀x P(x)\forall x\,P(x) is incomplete until its domain is stated or fixed by context. Bounded notation records the domain: ∀x∈S P(x)\forall x\in S\,P(x) and ∃x∈S P(x)\exists x\in S\,P(x).

Definition

Universal and existential quantifiers

The statement ∀x P(x)\forall x\,P(x) requires PP for every domain value; ∃x P(x)\exists x\,P(x) requires one allowed witness. A universal proof starts with an arbitrary element, an existential proof checks a particular witness, a universal refutation gives one counterexample, and an existential refutation rules out all candidates.

Worked example

Evaluate a predicate before and after quantifying

Let the bounded domain be D={1,2}D=\{1,2\}, let the ambient domain for assignments be Z\mathbb Z, and let P(x,y)P(x,y) mean x=yx=y. Under the assignment x=1x=1, y=2y=2, the open formula P(x,y)P(x,y) is false. In ∀x∈D P(x,y)\forall x\in D\,P(x,y), xx is bound but yy is free, so the formula is syntactically open; this does not imply that its truth must vary with yy. For example, over the ambient integers, ∀x∈Z (x=y)\forall x\in\mathbb Z\,(x=y) is open for the same syntactic reason but is false for every integer assignment to yy, since x=y+1x=y+1 is a counterexample. Returning to the bounded formula, y=1y=1 makes ∀x∈D P(x,y)\forall x\in D\,P(x,y) false because x=2x=2 fails, and y=3y=3 is an allowed external assignment even though it is not a bounded witness. Finally, ∀x∈D ∃y∈D P(x,y)\forall x\in D\,\exists y\in D\,P(x,y) is a proposition: for each xx, choose y=xy=x. Binding all relevant variables, rather than changing PP, is what closes the formula.

Reading syntax and scope

The scope of a quantifier is the formula immediately following it, unless parentheses make it larger. In

∀x (P(x)→∃y (Q(x,y)∧R(y))),\forall x\,(P(x)\to\exists y\,(Q(x,y)\land R(y))),

the outer quantifier controls xx throughout its parentheses, while the inner one controls yy only there. Thus xx is available when choosing yy, but yy cannot be used outside. Parentheses make the intended connective scope explicit.

A bound variable is a placeholder whose name can be changed when the replacement is safe: ∀x P(x)\forall x\,P(x) and ∀z P(z)\forall z\,P(z) agree when zz is absent from the body. Reusing a letter bound in an inner scope can capture outer occurrences and change the truth value.

Proof obligations for quantified claims

Quantifier words determine proof shape. For ∀x∈D P(x)\forall x\in D\,P(x), write “let x∈Dx\in D be arbitrary” and derive P(x)P(x) without choosing a convenient value; this is why a proof of injectivity starts with arbitrary x1,x2∈Xx_1,x_2\in X. For ∃x∈D P(x)\exists x\in D\,P(x), announce an allowed candidate and verify membership and PP. A universal refutation needs one allowed counterexample; an existential refutation must rule out every candidate (or exhaust a finite domain).

For a nested claim, ∀x ∃y P(x,y)\forall x\,\exists y\,P(x,y) first fixes arbitrary xx and permits a dependent choice yy; ∃y ∀x P(x,y)\exists y\,\forall x\,P(x,y) chooses one yy before the arbitrary xx. Choosing y=xy=x in the second formula proves a different statement.

Conditions, vacuous truth, and counterexamples

An implication in ∀x (Student(x)→P(x))\forall x\,(Student(x)\to P(x)) restricts the obligation to students; non-students make it vacuously true. But a non-student can also witness ∃x (Student(x)→P(x))\exists x\,(Student(x)\to P(x)) without satisfying PP, so require Student(x)∧P(x)Student(x)\land P(x) when membership is intended. Likewise, ∀x∈S P(x)\forall x\in S\,P(x) expands with antecedent x∈Sx\in S, while ∃x∈S P(x)\exists x\in S\,P(x) requires membership and PP; a negated universal needs an element inside SS where PP fails.

If the bounded domain is empty, these expansions make the edge case explicit: ∀x∈∅ P(x)\forall x\in\varnothing\,P(x) is true because every implication x∈∅→P(x)x\in\varnothing\to P(x) has a false antecedent, while ∃x∈∅ P(x)\exists x\in\varnothing\,P(x) is false because no conjunction x∈∅∧P(x)x\in\varnothing\land P(x) can be true.

Relations and proof obligations

Quantifiers also organize definitions built from several clauses. Writing xRyxRy for (x,y)∈R(x,y)\in R, a partial order on XX requires universal clauses for reflexivity, antisymmetry, and transitivity; an equivalence relation replaces antisymmetry with symmetry. Injectivity is expressed by

∀x1,x2∈X (f(x1)=f(x2)→x1=x2).\forall x_1,x_2\in X\,(f(x_1)=f(x_2)\to x_1=x_2).

Its proof starts with arbitrary x1,x2x_1,x_2 and its negation supplies two distinct inputs with equal outputs. Each clause has a clear proof obligation: universal definitions use arbitrary elements, while failed definitions use an explicit counterexample. The English words “only” and “all” must also keep their direction: “only students may submit” is ∀x (Submitted(x)→Student(x))\forall x\,(Submitted(x)\to Student(x)), whereas “all students submit” reverses the implication.

Written in full, the three partial-order obligations are

∀x∈X (xRx),∀x,y∈X (xRy∧yRx→x=y),∀x,y,z∈X (xRy∧yRz→xRz).\forall x\in X\,(xRx),\quad \forall x,y\in X\,(xRy\land yRx\to x=y),\quad \forall x,y,z\in X\,(xRy\land yRz\to xRz).

For an equivalence relation, replace the middle clause by symmetry, ∀x,y∈X (xRy→yRx)\forall x,y\in X\,(xRy\to yRx). These clauses are an example of several independent universal proof obligations joined by a conjunction.

Negation reverses the quantifier

Theorem

Negating a quantifier

For any predicate PP over a fixed domain,

¬∀x P(x)≡∃x ¬P(x),¬∃x P(x)≡∀x ¬P(x).\neg\forall x\,P(x)\equiv\exists x\,\neg P(x), \qquad \neg\exists x\,P(x)\equiv\forall x\,\neg P(x).

Negation flips the outer quantifier and then negates the formula in its scope. It does not change the domain.

Theorem

Bounded quantifiers and De Morgan's laws

Bounded notation is an abbreviation:

∀x∈S P(x)  ≡  ∀x (x∈S→P(x)),∃x∈S P(x)  ≡  ∃x (x∈S∧P(x)).\forall x\in S\,P(x)\;\equiv\;\forall x\,(x\in S\to P(x)), \qquad \exists x\in S\,P(x)\;\equiv\;\exists x\,(x\in S\land P(x)).

Consequently,

¬∀x∈S P(x)≡∃x∈S ¬P(x),¬∃x∈S P(x)≡∀x∈S ¬P(x).\neg\forall x\in S\,P(x)\equiv\exists x\in S\,\neg P(x), \qquad \neg\exists x\in S\,P(x)\equiv\forall x\in S\,\neg P(x).

Inside a scope, use the propositional laws ¬(A∧B)≡(¬A∨¬B)\neg(A\land B)\equiv(\neg A\lor\neg B) and ¬(A∨B)≡(¬A∧¬B)\neg(A\lor B)\equiv(\neg A\land\neg B). In particular, the negation of A→BA\to B is A∧¬BA\land\neg B.

Proof sketch or proof idea

The first quantifier law follows directly from what it means for a universal statement to fail: “not every domain element satisfies PP” means that at least one domain element makes PP fail. Similarly, “there is no domain element satisfying PP” means every domain element fails PP.

For a nested formula, apply one law at a time from the outside inward. For example, over a fixed domain,

¬∀x ∃y P(x,y)≡∃x ¬∃y P(x,y)≡∃x ∀y ¬P(x,y).\begin{aligned} \neg\forall x\,\exists y\,P(x,y) &\equiv\exists x\,\neg\exists y\,P(x,y)\\ &\equiv\exists x\,\forall y\,\neg P(x,y). \end{aligned}

The counterexample has the same outer xx as the original universal claim, but for that xx every possible yy fails. Stopping after the first flip would leave the second quantifier and its scope wrong.

Keep the domain and track the witness

Worked example

Negate the definition of a prime number

Fix a natural number n≥2n\ge 2 throughout. The lower bound is essential because 1 is not prime. A standard definition says that every natural divisor of nn is trivial:

∀d∈N (d∣n→(d=1∨d=n)).\forall d\in\mathbb N\,\bigl(d\mid n\to(d=1\lor d=n)\bigr).

Negate in three steps. First flip the quantifier. Next negate the implication, which keeps its hypothesis and negates its conclusion. Finally apply De Morgan's law:

¬∀d∈N (d∣n→(d=1∨d=n))≡∃d∈N ¬(d∣n→(d=1∨d=n))≡∃d∈N (d∣n∧¬(d=1∨d=n))≡∃d∈N (d∣n∧d≠1∧d≠n).\begin{aligned} \neg\forall d\in\mathbb N\,\bigl(d\mid n\to(d=1\lor d=n)\bigr) &\equiv\exists d\in\mathbb N\,\neg\bigl(d\mid n\to(d=1\lor d=n)\bigr)\\ &\equiv\exists d\in\mathbb N\,\bigl(d\mid n\land\neg(d=1\lor d=n)\bigr)\\ &\equiv\exists d\in\mathbb N\,\bigl(d\mid n\land d\ne1\land d\ne n\bigr). \end{aligned}

In English, there is a natural number dd which divides nn and is neither 1 nor nn. That is a concrete non-trivial divisor, exactly the witness needed to show that nn is not prime. The assumption n≥2n\ge2 is a hypothesis of this example, not a condition that should disappear during negation.

Worked example

Negate a course prerequisite statement

Let Student(x)Student(x) mean “xx is a student,” Enrolls(x)Enrolls(x) mean “xx enrolls,” and Prereq(x)Prereq(x) mean “xx satisfies the prerequisite.” Consider the statement

∀x ((Student(x)∧Enrolls(x))→Prereq(x)).\forall x\,\bigl((Student(x)\land Enrolls(x))\to Prereq(x)\bigr).

The negation is

∃x (Student(x)∧Enrolls(x)∧¬Prereq(x)).\exists x\,\bigl(Student(x)\land Enrolls(x)\land\neg Prereq(x)\bigr).

Its witness is a student who enrolls and does not satisfy the prerequisite. A student who does not enroll is not a counterexample, because the original implication makes no claim about that person. This is why replacing the implication by a conjunction, or dropping the hypothesis while negating, gives the wrong condition.

Worked example

Bounded universal versus bounded existential

The two bounded forms have different connectives. For a set SS,

∀x∈S P(x)≡∀x (x∈S→P(x)),∃x∈S P(x)≡∃x (x∈S∧P(x)).\forall x\in S\,P(x)\equiv\forall x\,(x\in S\to P(x)), \qquad \exists x\in S\,P(x)\equiv\exists x\,(x\in S\land P(x)).

Negating ∀n∈Z (n>0→n2>n)\forall n\in\mathbb Z\,(n\gt 0\to n^2\gt n) gives

∃n∈Z (n>0∧n2≤n).\exists n\in\mathbb Z\,(n\gt 0\land n^2\le n).

The witness n=1n=1 satisfies both requirements. The integer 0 satisfies 02≤00^2\le0 but fails n>0n\gt 0, so it cannot witness the negation. This example shows why the antecedent of an implication becomes part of a counterexample.

One witness or a choice for each input

Worked example

Quantifier order and dependent witnesses

Over the natural numbers, compare

∀x∈N ∃y∈N (x<y)\forall x\in\mathbb N\,\exists y\in\mathbb N\,(x\lt y)

with

∃y∈N ∀x∈N (x<y).\exists y\in\mathbb N\,\forall x\in\mathbb N\,(x\lt y).

The first is true: after an arbitrary xx is presented, choose y=x+1y=x+1. The choice depends on xx. The second is false: a proposed fixed natural number yy is defeated by the allowed choice x=yx=y, for which x<yx\lt y is false. A proof of the first statement therefore cannot be reused as a proof of the second; it constructs a function of the input rather than one uniform witness.

The finite example makes the same point without an unbounded domain. Let X={1,2}X=\{1,2\} and P(x,y)P(x,y) mean x=yx=y. Then

∀x∈X ∃y∈X P(x,y)\forall x\in X\,\exists y\in X\,P(x,y)

is true by choosing y=xy=x. But

∃y∈X ∀x∈X P(x,y)\exists y\in X\,\forall x\in X\,P(x,y)

is false: y=1y=1 fails at x=2x=2, while y=2y=2 fails at x=1x=1.

Worked example

Safe distribution and permutation of quantifiers

Over a fixed domain, the following identities

∀x (P(x)∧Q(x))≡(∀x P(x))∧(∀x Q(x)),∃x (P(x)∨Q(x))≡(∃x P(x))∨(∃x Q(x))\forall x\,(P(x)\land Q(x))\equiv(\forall x\,P(x))\land(\forall x\,Q(x)), \qquad \exists x\,(P(x)\lor Q(x))\equiv(\exists x\,P(x))\lor(\exists x\,Q(x))

follow from the proof obligations. For the first, take an arbitrary xx: the conjunction is true exactly when both predicates hold for that same element. For the second, a witness for the disjunction lies in one case, and a witness for either case works in reverse. Same-type blocks also commute, but their meanings differ:

∀x ∀y P(x,y)≡∀y ∀x P(x,y),∃x ∃y P(x,y)≡∃y ∃x P(x,y).\forall x\,\forall y\,P(x,y)\equiv\forall y\,\forall x\,P(x,y), \qquad \exists x\,\exists y\,P(x,y)\equiv\exists y\,\exists x\,P(x,y).

The universal statement checks every ordered pair; the existential statement asks for one ordered pair. Neither fact licenses swapping mixed quantifiers. For the crossed failure on {1,2}\{1,2\}, let P(x)P(x) mean x=1x=1 and Q(x)Q(x) mean x=2x=2. Then ∀x(P(x)∨Q(x))\forall x(P(x)\lor Q(x)) is true while (∀xP(x))∨(∀xQ(x))(\forall xP(x))\lor(\forall xQ(x)) is false. Likewise, ∃x(P(x)∧Q(x))\exists x(P(x)\land Q(x)) is false while (∃xP(x))∧(∃xQ(x))(\exists xP(x))\land(\exists xQ(x)) is true.

Theorem

Witness transfer

If ∀x (P(x)→Q(x))\forall x\,(P(x)\to Q(x)) and ∃x P(x)\exists x\,P(x), then ∃x Q(x)\exists x\,Q(x). Choose a witness aa with P(a)P(a). Instantiating the universal premise at aa gives P(a)→Q(a)P(a)\to Q(a), hence Q(a)Q(a); the same aa is a witness for the conclusion.

Worked example

Negating a conjunction under existence

The outside-in rules and De Morgan's law give

¬∃x (P(x)∧Q(x))≡∀x ¬(P(x)∧Q(x))≡∀x (¬P(x)∨¬Q(x)).\neg\exists x\,(P(x)\land Q(x)) \equiv\forall x\,\neg(P(x)\land Q(x)) \equiv\forall x\,(\neg P(x)\lor\neg Q(x)).

The final sentence says that every object fails at least one of the two properties; it does not say that every object fails both.

Worked example

Read a bounded-order formula in its stated domain

The formula

∃y ∀x (x≤y)\exists y\,\forall x\,(x\le y)

says that one value yy is at least as large as every domain value xx. Its truth depends on the stated domain, so report the domain with the translation.

Worked example

Different books, different borrowers

Let PeoplePeople and BooksBooks be the sets of people and books. The two statements are

∀b∈Books ∃x∈People Borrows(x,b)\forall b\in Books\,\exists x\in People\,Borrows(x,b)

and

∃x∈People ∀b∈Books Borrows(x,b).\exists x\in People\,\forall b\in Books\,Borrows(x,b).

The first says every book has a borrower; the second says one person borrows all books. With two books, separate borrowers can borrow one each, neither both, so the first is true while the second is false.

“Everyone is reading at least two books” requires distinct book witnesses:

∀x∈People ∃b1,b2∈Books (b1≠b2∧Reads(x,b1)∧Reads(x,b2)).\forall x\in People\,\exists b_1,b_2\in Books\,(b_1\ne b_2\land Reads(x,b_1)\land Reads(x,b_2)).

Worked example

Translation with a student domain

Let the domain be all students at a university. If S(x)S(x) means “xx studies mathematics” and P(x)P(x) means “xx passed the exam,” consider these three statements, respectively: every mathematics student passed; at least one student passed; and at least one mathematics student did not pass. Their translations are

∀x (S(x)→P(x)),∃x P(x),∃x (S(x)∧¬P(x)).\forall x\,(S(x)\to P(x)),\qquad \exists x\,P(x),\qquad \exists x\,(S(x)\land\neg P(x)).

The domain supplies “student”; the predicates add the requested properties.

Translate the dependency before the symbols

Worked example

Translate nested conditions into logic

Assume the domain contains students, professors, courses, exams, and questions, and use predicates to restrict each kind of object. The following sentences illustrate three different dependencies.

“Every student has taken at least one mathematics course” becomes

∀x (Student(x)→∃m (MathCourse(m)∧Taken(x,m))).\forall x\,\bigl(Student(x)\to\exists m\,(MathCourse(m)\land Taken(x,m))\bigr).

“There is a professor who teaches every course in the department” becomes

∃p (Professor(p)∧∀c (DepartmentCourse(c)→Teaches(p,c))).\exists p\,\bigl(Professor(p)\land\forall c\,(DepartmentCourse(c)\to Teaches(p,c))\bigr).

“Every exam has a question that every student finds difficult” becomes

∀e (Exam(e)→∃q (Question(q,e)∧∀s (Student(s)→Difficult(s,q)))).\forall e\,\bigl(Exam(e)\to\exists q\,(Question(q,e)\land\forall s\,(Student(s)\to Difficult(s,q)))\bigr).

The difficult question may depend on the exam, but it must work for every student for that exam. Moving ∃q\exists q before ∀e\forall e claims one question works for all exams, a stronger statement.

Diagnose a claim by testing its witnesses

Worked example

Diagnose an incorrect formalization

Consider three errors that change the intended claim. The formula

∃d∈N (d∣n→(d=1∨d=n))\exists d\in\mathbb N\,\bigl(d\mid n\to(d=1\lor d=n)\bigr)

does not define primality. Keep the intended hypothesis n≥2n\ge2. The witness d=n+1d=n+1 is natural and does not divide nn, so the implication is true vacuously; the existential formula is satisfied without checking all divisors. (For n=0n=0, do not use this particular witness: 11 divides 00.) The prime definition requires ∀d\forall d.

In fact, d=1d=1 witnesses this existential formula for every nn, since both 1∣n1\mid n and d=1d=1 hold. The d=n+1d=n+1 choice highlights the separate vacuous truth under the intended n≥2n\ge2 hypothesis.

The formula ∀x ∃y Attends(x,y)\forall x\,\exists y\,Attends(x,y) says that each xx attends at least one yy, with the types and dependency left unclear. If the intended statement is that every lecture has a student attending it, write

∀ℓ (Lecture(ℓ)→∃s (Student(s)∧Attends(s,ℓ))).\forall \ell\,\bigl(Lecture(\ell)\to\exists s\,(Student(s)\land Attends(s,\ell))\bigr).

Finally, ∀a ∃d SubmittedBefore(a,d)\forall a\,\exists d\,SubmittedBefore(a,d) allows the deadline to depend on the assignment. A single deadline before which all assignments are submitted is

∃d ∀a (Assignment(a)→SubmittedBefore(a,d)).\exists d\,\forall a\,(Assignment(a)\to SubmittedBefore(a,d)).

Each correction makes the dependency order explicit.

Worked example

Rename a bound variable without capture

Bound variable names are local labels. If zz does not occur anywhere in the body of ∀x P(x)\forall x\,P(x), replacing the occurrences controlled by that outer quantifier gives the equivalent formula ∀z P(z)\forall z\,P(z). Occurrences bound by an inner quantifier are left alone. The same rule applies to ∃\exists.

Global absence of the replacement letter from the body is a sufficient safeguard, not a necessary rule. What is necessary is that the renaming avoid capturing an occurrence whose binding or free status changes. On X={1,2}X=\{1,2\},

∀x∈X ∃y∈X (x≠y)\forall x\in X\,\exists y\in X\,(x\ne y)

is true because each element has the other element as a witness. Replacing the outer xx by the already-used yy produces

∀y∈X ∃y∈X (y≠y),\forall y\in X\,\exists y\in X\,(y\ne y),

which is false: the inner quantifier captures both occurrences. A genuinely fresh zz gives ∀z∈X ∃y∈X (z≠y)\forall z\in X\,\exists y\in X\,(z\ne y) and preserves the truth conditions.

Worked example

Existence and uniqueness are separate proof obligations

“There exists a unique x∈Dx\in D satisfying P(x)P(x)” means both existence and at-most-one:

∃x∈D (P(x)∧∀y∈D (P(y)→y=x)).\exists x\in D\,\bigl(P(x)\land\forall y\in D\,(P(y)\to y=x)\bigr).

To prove it, first exhibit one witness xx and verify P(x)P(x). Then let yy be an arbitrary element of DD satisfying P(y)P(y) and prove y=xy=x. To disprove uniqueness, two distinct witnesses are enough. Over Z\mathbb Z, x2=1x^2=1 has witnesses 11 and −1-1, so existence holds but uniqueness fails. Over Z\mathbb Z, x2=0x^2=0 has the single witness 00: substitution verifies existence, and y2=0y^2=0 implies y=0y=0. The statement “my classmate has exactly one friend” uses the same pattern: ∃!x Friend(o,x)\exists!x\,Friend(o,x) abbreviates existence plus at-most-one.

Worked example

Translate predicate logic back into English

With people as the domain,

∀x ∃y (x≠y∧Friend(x,y))\forall x\,\exists y\,(x\ne y\land Friend(x,y))

says every person has a distinct friend; the distinctness condition is inside the existential scope.

With a domain containing people and students,

∃x ∀y (Student(y)→Older(x,y))\exists x\,\forall y\,(Student(y)\to Older(x,y))

says one person is older than every student. It does not require xx to be a student unless that predicate is added.

Common mistakes

Common mistake

Do not lose a hypothesis when negating an implication

The negation of A→BA\to B is A∧¬BA\land\neg B, not ¬A→¬B\neg A\to\neg B and not ¬A∧B\neg A\land B. For a bounded universal implication, a counterexample must lie in the intended restricted class and fail the conclusion. In the prerequisite example, a non-student or a student who does not enroll cannot witness the negation.

Common mistake

Do not swap mixed quantifiers

Same-type blocks commute over a fixed domain, but for different reasons: ∀x∀y\forall x\forall y checks every pair, whereas ∃x∃y\exists x\exists y needs one pair. Mixed blocks generally cannot be swapped. The finite set X={1,2}X=\{1,2\} with P(x,y)P(x,y) meaning x=yx=y is a concrete counterexample.

Common mistake

A truth table does not enumerate an infinite domain

Predicate logic extends propositional logic. A truth table can track the truth of connectives after particular atomic statements are assigned, but it does not replace a proof over an infinite domain such as N\mathbb N. For a universal claim, prove it for an arbitrary element or find one counterexample; for an existential claim, provide or rule out witnesses.

Follow a negation step by step

Use the stepper to move between a quantified statement and its negation. Treat each step as a scope check: flip one quantifier, negate its complete scope, and then simplify the propositional connective. The widget supports the article's reasoning but does not replace the written proof.

Read and try

Negate one quantified statement carefully

The worked sequence reveals one quantifier-negation move at a time.

Example

For every real number x, x^2 >= 0.

  1. 1. Start with the outer quantifier: “for every x.”

Summary

Quantifiers bind variables over an explicit domain. Free variables need an assignment; bounded universals use implication, bounded existentials use conjunction, and domains stay fixed under negation. Negate outside in, changing ∀\forall to ∃\exists and applying De Morgan's laws and ¬(A→B)≡A∧¬B\neg(A\to B)\equiv A\land\neg B. Order records dependency: ∀x∃y\forall x\exists y permits dependence, while ∃y∀x\exists y\forall x requires a uniform witness. Universal proofs use arbitrary elements, existential proofs use checked witnesses, and uniqueness needs existence plus at-most-one.

Exercises: scope, witnesses, and proof

Exercise 1

Over a fixed domain, negate ∀x ∃y P(x,y)\forall x\,\exists y\,P(x,y) and simplify completely.

Solution · Hint

Apply the quantifier-negation law one quantifier at a time, then stop only when the connective inside has been negated.

Solution · Model solution
¬∀x ∃y P(x,y)≡∃x ¬∃y P(x,y)≡∃x ∀y ¬P(x,y).\begin{aligned} \neg\forall x\,\exists y\,P(x,y) &\equiv\exists x\,\neg\exists y\,P(x,y)\\ &\equiv\exists x\,\forall y\,\neg P(x,y). \end{aligned}

There is an xx for which every yy fails P(x,y)P(x,y).

Exercise 2

Formalize: there is one student who finds at least one question difficult in every exam. Use Student, Exam, Question, and Difficult.

Solution · Hint

Keep the student fixed before choosing an exam; the question may depend on the exam.

Solution · Model solution
∃s (Student(s)∧∀e (Exam(e)→∃q (Question(q,e)∧Difficult(s,q)))).\exists s\,\bigl(Student(s)\land\forall e\,(Exam(e)\to\exists q\,(Question(q,e)\land Difficult(s,q)))\bigr).

The student is chosen before the exam. For each exam, a question can then be chosen for that same student.

Exercise 3

Assume n≥2n\ge2. Why is ∃d∈N (d∣n→(d=1∨d=n))\exists d\in\mathbb N\,(d\mid n\to(d=1\lor d=n)) too weak to define a prime number?

Solution · Hint

Find a natural number for which the implication is true for a vacuous reason.

Solution · Model solution

Under the stated hypothesis n≥2n\ge2, choose d=n+1d=n+1. This natural number does not divide nn, so the implication has a false antecedent and is true. An existential implication therefore does not require all divisors to be trivial; the correct prime definition uses ∀d\forall d.

Exercise 4

Can a finite truth table by itself decide ∀x P(x)\forall x\,P(x) when the domain is N\mathbb N? State the correct proof alternatives.

Solution · Hint

Separate propositional truth assignments from quantification over domain elements.

Solution · Model solution

No. A truth table handles finitely many truth values of atomic propositions; it does not enumerate all natural numbers. Prove a universal statement for an arbitrary natural number, or refute it with one natural counterexample. For an existential statement, exhibit an allowed witness or show that every candidate fails.

Read this first

For the propositional version of these ideas, review 1.1 Propositional logic.

Practice

Work out your answer, then check it. You can revise and try again.

Loading…

Key terms in this unit