Discrete Mathematics

(Predicate Logic)

Pramod Ganapathi
Department of Computer Science
State University of New York at Stony Brook

January 24, 2021

Contents

Contents
Predicates and Quantified Statements
Statements with Multiple Quantifiers
Arguments with Quantified Statements
Predicates and Quantified Statements
What is a propositional function or predicate?

Definition
A propositional function or predicate is a sentence that contains
one or more variables
A predicate is neither true nor false
A predicate becomes a proposition when the variables are sub-
stituted with specific values
The domain of a predicate variable is the set of all values that
may be substituted for the variable
Examples
Symbol Predicate Domain Propositions
p(x) x>5 x∈R p(6), p(−3.6), p(0), . . .
p(x, y) x + y is odd x ∈ Z, y ∈ Z p(4, 5), p(−4, −4), . . .
√
p(x, y) x2 + y 2 = 4 x ∈ R, y ∈ R p(−1.7, 8.9), p(− 3, −1), . . .
What is a truth set?

Definition
A truth set of a predicate is the set of all values of the predicate
that makes the predicate true
If p(x) is a predicate and x has domain D, then the truth set
of p(x) is the set of all elements of D that makes p(x) true
when the values are substituted for x. That is,

Truth set of p(x) = {x ∈ D | p(x)}

Examples
Symbol Predicate Domain Truth set
p(x) x>5 x∈R {p(6), p(13.6), p(5.001), . . .}
p(x, y) x + y is odd x ∈ Z, y ∈ Z {p(4, 5), p(−4, −3), . . .}
√
p(x, y) x2 + y 2 = 4 x ∈ R, y ∈ R {p(−2, 2), p(− 3, −1), . . .}
Predicates to propositions

There are two methods to obtain propositions from predicates

1. Assign specific values to variables
2. Add quantifiers

1. Assign specific values to variables

Predicates Propositions

2. Add quantifiers
What are quantifiers?

Definition
Quantifiers are words that refer to quantities such as “all” or
“some” and they tell for how many elements a given predicate
is true

Introduced into logic by logicians Charles Sanders Pierce and

Gottlob Frege during late 19th century
Two types of quantifiers:
1. Universal quantifier (∀)
2. Existential quantifier (∃)
Universal quantifier (∀)

Definition
Let p(x) be a predicate and D be the domain of x
A universal statement is a statement of the form

∀x ∈ D, p(x)

Forms:
“p(x) is true for all values of x”
“For all x, p(x)”
“For each x, p(x)”
“For every x, p(x)”
“Given any x, p(x)”
It is true if p(x) is true for each x in D; It is false if p(x) is
false for at least one x in D
A counterexample to the universal statement is the value of x
for which p(x) is false
Universal quantifier (∀)

Examples
Universal st.s Domain Truth value Method
∀x ∈ D, x2 ≥ x D = {1, 2, 3} True Method of exhaustion
∀x ∈ R, x2 ≥ x R False Counterexample
x = 0.1

Caution
Method of exhaustion cannot be used to prove universal state-
ments for infinite sets
Existential quantifier (∃)

Definition
Let p(x) be a predicate and D be the domain of x
An existential statement is a statement of the form

∃x ∈ D, p(x)

Forms:
“There exists an x such that p(x)”
“For some x, p(x)”
“We can find an x, such that p(x)”
“There is some x such that p(x)”
“There is at least one x such that p(x)”
It is true if p(x) is true for at least one x in D; It is false if p(x)
is false for all x in D
A counterproof to the existential statement is the proof to show
that p(x) is true is for no x
Existential quantifier (∃)

Examples
Universal st.s Domain Truth value Method
∃x ∈ D, x2 ≥ x D = {1, 2, 3} True Method of exhaust.
∃x ∈ R, x2 ≥ x R True Example
∃x ∈ Z, x + 1 ≤ x Z False How?
Formal and informal languages

Example
∀x ∈ R, x2 ≥ 0
Every real number has a nonnegative square
All real numbers have nonnegative squares
Any real number has a nonnegative square
The square of each real number is nonnegative
No real numbers have negative squares
x2 is nonnegative for every real x
x2 is not less than zero for each real number x
Universal conditional statement (∀, →)

Definition
A universal conditional statement is of the form

∀x, if p(x) then q(x)

Examples
∀x ∈ R, if x > 2 then x2 > 4
∀ real number x, if x is an integer then x is rational
∀ integer x, x is rational B Logically equivalent
∀x, if x is a square then x is a rectangle
∀ square x, x is a rectangle B Logically equivalent
∀x ∈ U , if p(x) then q(x)
∀x ∈ D, q(x) B Logically equivalent
(where, D = {x ∈ U | p(x) is true})

Can be extended to existential conditional statement (∃, →)

Implicit quantification (⇒, ⇔)

Examples
If a number is an integer, then it is a rational number
Implicit meaning: ∀ number x, if x is an integer, x is rational
The number 10 can be written as a sum of two prime numbers
Implicit meaning: ∃ prime numbers p and q such that 10 = p+q
If x > 2, then x2 > 4
Implicit meaning: ∀ real x, if x > 2, then x2 > 4
Definition
Let p(x) and q(x) be predicates and D be the common domain
of x. Then implicit quant. symbols ⇒, ⇔ are defined as:

p(x) ⇒ q(x) ≡ ∀x, p(x) → q(x)

p(x) ⇔ q(x) ≡ ∀x, p(x) ↔ q(x)
Implicit quantification (⇒, ⇔)

Problem
q(n): n is a factor of 8; r(n): n is a factor of 4
s(n): n < 5 and n 6= 3
Domain of n is Z+ (i.e., positive integers)
What are the relationships between q(n), r(n), and s(n) using
symbols ⇒ and ⇔?
Implicit quantification (⇒, ⇔)

Problem
q(n): n is a factor of 8; r(n): n is a factor of 4
s(n): n < 5 and n 6= 3
Domain of n is Z+ (i.e., positive integers)
What are the relationships between q(n), r(n), and s(n) using
symbols ⇒ and ⇔?
Solution
Truth set of q(n) = {1, 2, 4, 8}; Truth set of r(n) = {1, 2, 4};
Truth set of s(n) = {1, 2, 4}
∀n in Z+ , r(n) → q(n) i.e., r(n) ⇒ q(n)
i.e., “n is a factor of 4” ⇒ “n is a factor of 8”
∀n in Z+ , r(n) ↔ s(n) i.e., r(n) ⇔ s(n)
i.e., “n is a factor of 4” ⇔ “n < 5 and n 6= 3”
∀n in Z+ , s(n) → q(n) i.e., s(n) ⇒ q(n)
i.e., “n < 5 and n 6= 3” ⇒ “n is a factor of 8”
Negation of quantified statements (∼)
Definition
Formally,

∼ (∀x ∈ D, p(x)) ≡ ∃x ∈ D, ∼ p(x)

∼ (∃x ∈ D, p(x)) ≡ ∀x ∈ D, ∼ p(x)

Negation of a universal statement (“all are”) is logically equiv-

alent to an existential statement (“there is at least one that is
not”)
Negation of an existential statement (“some are”) is logically
equivalent to a universal statement (“all are not”)
Methods
Two methods to avoid errors while finding negations:
1. Write the statements formally and then take negations
2. Ask “What exactly would it mean for the given statement to
be false?”
Negation of quantified statements (∼)

Examples
All mathematicians wear glasses
Negation (incorrect): No mathematician wears glasses
Negation (incorrect + ambiguous): All mathematicians do not
wear glasses
Negation (correct): There is at least one mathematician who
does not wear glasses
Some snowflakes are the same
Negation (incorrect):: Some snowflakes are different
Negation (correct):: All snowflakes are different
Negation of quantified statements (∼)

Examples
∀ primes p, p is odd
Negation: ∃ primes p, p is even
∃ triangle T , sum of angles of T equals 200◦
∀ triangles T , sum of angles of T does not equal 200◦
No politicians are honest
Formal statement: ∀ politicians x, x is not honest
Formal negation: ∃ politician x, x is honest
Informal negation: Some politicians are honest
1357 is not divisible by any integer between 1 and 37
Formal statement: ∀n ∈ [1, 37], 1357 is not divisible by n
Formal negation: ∃n ∈ [1, 37], 1357 is divisible by n
Informal negation: 1357 is divisible by some integer between 1
and 37
Negation of universal conditional statements

Definition
Formally,

∼ (∀x, p(x) → q(x)) ≡ ∃x, ∼ (p(x) → q(x))

≡ ∃x, (p(x)∧ ∼ q(x))

Examples
∀ real x, if x > 10, then x2 > 100.
Negation: ∃ real x such that x > 10 and x2 ≤ 100.
If a computer program has more than 100,000 lines, then it
contains a bug.
Negation: There is at least one computer program that has
more than 100,000 lines and does not contain a bug.
Relation between quantifiers (∀, ∃) and (∧, ∨)

Relation
Universal statements are generalizations of and statements
Existential statements are generalizations of or statements
If p(x) is a predicate and D = {x1 , x2 , . . . , xn } is the domain
of x, then

∀x ∈ D, p(x) ≡ p(x1 ) ∧ p(x2 ) ∧ · · · ∧ p(xn )

∃x ∈ D, p(x) ≡ p(x1 ) ∨ p(x2 ) ∨ · · · ∨ p(xn )
Vacuous truth of universal statements
Problem
Consider the bowl and the balls
Consider the statement:
All the balls in the bowl are blue
Bowl Balls
Is the statement true?
Solution
The statement is false iff its negation is true
Negation: There exists a ball in the bowl that is not blue
The negation is false; So, the statement is true, by default
Definition
A statement of the form

∀x in D, if p(x), then q(x)

is vacuously true or true by default, if and only if p(x) is false

for all x in D
Universal conditional statements (∀x, p(x) → q(x))
Definitions
Statement: ∀x, if p(x) then q(x)
Contrapositive of the statement is ∀x, if ∼ q(x) then ∼ p(x)
Converse of the statement is ∀x, if q(x) then p(x)
Inverse of the statement is ∀x, if ∼ p(x) then ∼ q(x)
Identities
Conditional ≡ Contrapositive B Useful for proofs
Conditional 6≡ Converse
Conditional 6≡ Inverse
Converse ≡ Inverse
Formulas
∀x, p(x) → q(x) ≡ ∀x, ∼ q(x) →∼ p(x) B Useful for proofs
∀x, p(x) → q(x) 6≡ ∀x, q(x) → p(x)
∀x, p(x) → q(x) 6≡ ∀x, ∼ p(x) →∼ q(x)
∀x, q(x) → p(x) ≡ ∀x, ∼ p(x) →∼ q(x)
Universal conditional statement ∀x, p(x) → q(x)

Definitions
∀x, p(x) is a sufficient condition for q(x) means
∀x, if p(x) then q(x)
∀x, p(x) is a necessary condition for q(x) means
∀x, if ∼ p(x) then ∼ q(x) ≡ ∀x, if q(x) then p(x)
∀x, p(x) only if q(x) means
∀x, if ∼ q(x) then ∼ p(x) ≡ ∀x, if p(x) then q(x)
Example
For real x, x = 1 is a sufficient condition for x2 = 1
i.e., ∀x, if x = 1 then x2 = 1 B True
For real x, x2 = 1 is a necessary condition for x = 1
i.e., ∀x, if x2 6= 1 then x 6= 1 B True
For real x, x = 1 only if x2 = 1
i.e., ∀x, if x2 6= 1 then x 6= 1 B True
Statements with Multiple Quantifiers
Statements with multiple quantifiers

Problem
What is the interpretation for the following statement?
“There is a person supervising every detail of the production
process.”
Ambiguous interpretations
1. There is one single person who supervises all the details of the
production process.
∃ person p such that ∀ detail d, p supervises d
2. For any particular production detail, there is a person who
supervises that detail, but there might be different supervisors
for different details.
∀ detail d, ∃ person p such that p supervises d
Statements with multiple quantifiers

Definitions
1. Statement form:

∀x ∈ D, ∃y ∈ E such that P (x, y)

Interpretation: Allow someone else to pick whatever element

x in D they wish. Then, you must find an element y in E that
“works” for that particular x.
2. Statement form:

∃x ∈ D such that ∀y ∈ E, P (x, y)

Interpretation: Your job is to find one particular x in D that

will “work” no matter what y in E anyone might choose to
challenge you with.
Example: Tarski world

Problem
For all triangles x, there is a square y such that x and y have
the same color. Truth value?
Answer
True. How?
Example: Tarski world

Problem
There is a triangle x such that for all circles y, x is to the right
of y. Truth value?
Answer
True. How?
Example: College cafeteria

Problem
∃ an item I such that ∀ students S, S chose I.
Informal statement? Truth value?
Solution
There is an item that was chosen by every student.
True. How?
Example: College cafeteria

Problem
∃ a student S such that ∀ items I, S chose I.
Informal statement? Truth value?
Solution
There is a student who chose every available item.
False. How?
Example: College cafeteria

Problem
∃ a student S such that ∀ stations Z, ∃ an item I in Z such
that S chose I.
Informal statement? Truth value?
Solution
There is a student who chose at least one item from every
station.
True. How?
Example: College cafeteria

Problem
∀ students S and ∀ stations Z, ∃ an item I in Z such that S
chose I.
Informal statement? Truth value?
Solution
Every student chose at least one item from every station.
False. How?
Translating from informal to formal language

Problem
Every nonzero real number has a reciprocal.
There is a real number with no reciprocal.
There is a smallest positive integer.
There is no smallest positive real number.
Translating from informal to formal language

Problem
Every nonzero real number has a reciprocal.
There is a real number with no reciprocal.
There is a smallest positive integer.
There is no smallest positive real number.
Solution
∀ nonzero real numbers u, ∃ a real number v such that uv = 1.
∃ a real number c such that ∀ real numbers d, cd 6= 1.
∃ a positive integer m such that ∀ positive integers n, m ≤ n.
∀ positive real numbers x, ∃ a positive real number y such that
y < x.
Negations of multiply-quantified statements

Definitions
∼ (∀x in D, ∃y in E such that P (x, y))
≡ ∃x in D such that ∼ (∃y in E such that P (x, y))
≡ ∃x in D such that ∀y in E, ∼ P (x, y)
∼ (∃x in D such that ∀y in E, P (x, y))
≡ ∀x in D, ∼ (∀y in E, P (x, y))
≡ ∀x in D, ∃y in E such that ∼ P (x, y)
Example: Tarski world

Problem
For all squares x, there is a circle y such that x and y have the
same color. Negation?
Answer
∃ a square x such that ∀ circles y, x and y do not have the
same color. True. How?
Example: Tarski world

Problem
There is a triangle x such that for all squares y, x is to the
right of y. Negation?
Answer
∀ triangles x, ∃ a square y such that x is not to the right of y.
True. How?
Order of quantifiers

Order
The order of quantifiers are important when multiple quantifiers
are involved
Example
∀ people x, ∃ a person y such that x loves y.
Quite possible.
∃ a person y such that ∀ people x, x loves y.
Quite impossible.
Order of quantifiers

Example
For every square x there is a triangle y such that x and y have
different colors B True
There exists a triangle y such that for every square x, x and y
have different colors. B False
Order of quantifiers

Example
Suppose R∗ is a set of nonzero real numbers.
∀x ∈ Z, ∃y ∈ R∗ (xy < 1) B True
Two cases:
a. For x ≤ 0, let y = 1, then xy < 1
b. For x > 0, let y = 1/(x + 1), then xy < 1
∃y ∈ R∗ , ∀x ∈ Z (xy < 1) B False
Two cases:
a. For y > 0, if integer x ≥ 1/y, then xy ≮ 1
b. For y < 0, if integer x ≤ 1/y, then xy ≮ 1
In both the cases, an adversary can choose an integer that
makes the predicate false. Hence, the quantified statement is
false.
Formal logical notation

Definitions
∀x in D, P (x)
≡ ∀x(x in D → P (x))
∃x in D such that P (x)
≡ ∃x(x in D ∧ P (x))
Example: Tarski world

Definitions
Triangle(x): x is a triangle
Circle(x): x is a circle
Square(x): x is a square
Blue(x): x is blue
Gray(x): x is gray
Black(x): x is black
RightOf(x, y): x is to the right of y
Above(x, y): x is above y
SameColor(x, y): x has the same color as y
Example: Tarski world

Problem
For all circles x, x is above f .
Formal statement? Formal negation?
Solution
Formal statement
∀x(Circle(x) → Above(x, f ))
Formal negation
∼ (∀x(Circle(x) → Above(x, f )))
≡ ∃x ∼ (Circle(x) → Above(x, f ))
≡ ∃x(Circle(x)∧ ∼ Above(x, f ))
Example: Tarski world

Problem
There is a square x such that x is black.
Formal statement? Formal negation?
Solution
Formal statement
∃x(Square(x) ∧ Black(x))
Formal negation
∼ (∃x(Square(x) ∧ Black(x)))
≡ ∀x ∼ (Square(x) ∧ Black(x))
≡ ∀x(∼ Square(x)∨ ∼ Black(x))
Example: Tarski world

Problem
For all circles x, there is a square y such that x and y have the
same color.
Formal statement? Formal negation?
Solution
Formal statement
∀x(Circle(x) → ∃y(Square(y) ∧ SameColor(x, y)))
Formal negation
∼ (∀x(Circle(x) → ∃y(Square(y) ∧ SameColor(x, y))))
≡ ∃x ∼ (Circle(x) → ∃y(Square(y) ∧ SameColor(x, y)))
≡ ∃x(Circle(x)∧ ∼ (∃y(Square(y) ∧ SameColor(x, y))))
≡ ∃x(Circle(x) ∧ ∀y(∼ (Square(y) ∧ SameColor(x, y))))
≡ ∃x(Circle(x) ∧ ∀y(∼ Square(y)∨ ∼ SameColor(x, y)))
Example: Tarski world

Problem
There is a square x such that for all triangles y, x is to right
of y.
Formal statement? Formal negation?
Solution
Formal statement
∃x(Square(x) ∧ ∀y(Triangle(y) → RightOf(x, y)))
Formal negation
∼ (∃x(Square(x) ∧ ∀y(Triangle(y) → RightOf(x, y))))
≡ ∀x ∼ (Square(x) ∧ ∀y(Triangle(y) → RightOf(x, y)))
≡ ∀x(∼ Square(x)∨ ∼ (∀y(Triangle(y) → RightOf(x, y))))
≡ ∀x(∼ Square(x) ∨ ∃y(∼ (Triangle(y) → RightOf(x, y))))
≡ ∀x(∼ Square(x) ∨ ∃y(Triangle(y)∧ ∼ RightOf(x, y)))
Arguments with Quantified Statements
Universal instantiation

Definition
If some property is true of everything in a set, then it is true of
any particular thing in the set.
Example
All men are mortal.
Socrates is a man.
∴ Socrates is mortal.
Rule of inference: Universal modus ponens

Definition
It has the form:
∀x, if P (x) then Q(x)
P (a) for a particular a
∴ Q(a)
Used in direct proofs
Example
Informal argument
If an integer is even, then its square is even.
k is a particular integer that is even.
∴ k 2 is even
Formal argument
∀x, if E(x) then S(x) B E(x)? S(x)? k?
E(k) for a particular k
∴ S(k)
Rule of inference: Universal modus tollens

Definition
It has the form:
∀x, if P (x) then Q(x)
∼ Q(a) for a particular a
∴ ∼ P (a)
Used in proof by contradiction
Example
Informal argument
All human beings are mortal.
Zeus is not mortal.
∴ Zeus is not human.
Formal argument
∀x, if H(x) then M (x) B H(x)? M (x)? Z?
∼ M (Z)
∴ ∼ H(Z)
Fallacy: Converse and inverse errors

Definition
Converse error has the form:
∀x, if P (x) then Q(x)
Q(a) for a particular a
∴ P (a)
Inverse error has the form:
∀x, if P (x) then Q(x)
∼ P (a) for a particular a
∴ ∼ Q(a)
Fallacy: Converse error

Example
Law
All the town criminals frequent the Hot Life bar.
John frequents the Hot Life bar.
∴ John is one of the town criminals.
Suspect John but don’t convict him.
Medicine
For all x, if x has pneumonia, then x has a fever and chills,
coughs deeply, and feels exceptionally tired and miserable.
John has a fever and chills, coughs deeply, and feels exception-
ally tired and miserable.
∴ John has pneumonia.
Diagnosis of pneumonia is a strong possibility, though not a
certainty.
Using diagrams to test validity: Example 1

Example
All human beings are mortal.
Zeus is not mortal.
∴ Zeus is not human. B Valid (Modus tollens)
Using diagrams to test validity: Example 2
Example
All human beings are mortal.
Felix is mortal.
∴ Felix is a human being. B Invalid (Converse error)
Using diagrams to test validity: Example 3
Example
No polynomial functions have horizontal asymptotes.
This function has a horizontal asymptote.
∴ This function is not a polynomial function. B Valid

Equivalence
P (x) : x is a polynomial function
Q(x) : x does not have a horizontal asymptote
∀x, if P (x) then Q(x)
∼ Q(a) for a particular a
∴ ∼ P (a) B Modus tollens