Maggie Johnson Handout #17
CS103A
The Logic of
Quantifiers
Key Topics
· Tautologies with Quantifiers
· First-Order Validity, Equivalence & Consequence
· DeMorgan’s & Other Equivalences for Quantifiers
What quantified sentences are logical truths?
What arguments involving quantifiers are valid?
Can we ignore quantifiers in determining truth value?
1. "x (Professor(x) ® Smart(x))
"x Professor(x)
---
"x Smart(x)
2. "x Professor(x)
"x Smart(x)
---
"x Professor(x) ^ Smart(x)
3. $x (Professor(x) ® Smart(x))
$x Professor(x)
---
$x Smart(x)
4. $x Professor(x)
$x Smart(x)
---
$x Professor(x) ^ Smart(x)
One way of arriving at tautologies is to substitute complex, quantified sentences into a known tautology from prepositional logic. Here is a known tautology:
P v ~P
The following is also a tautology:
("x
(Professor(x) ® Smart(x))) v
~("x
(Professor(x) ® Smart(x)))
Truth-Functional Form: All quantified sentences and atomic sentences have been replaced by variables, and the truth-functional connectives outside the scope of the quantifiers or connecting atomic sentences have been maintained.
A quantified sentence in FOL is a tautology iff its truth-functional form is a tautology.
$x (Professor(x) ® Smart(x)) $x Professor(x) ® $x Smart(x)
$x Professor(x) $x Professor(x)
--- ---
$x Smart(x) $x Smart(x)
|
Propositional Logic |
First-Order Logic |
General Notion |
|
Tautology |
FO Validity |
Logical Truth |
|
Tautological Consequence |
FO Consequence |
Logical Consequence |
|
Tautological Equivalence |
FO Equivalence |
Logical Equivalence |
A sentence of FOL is FO valid if it is a logical truth when you ignore the meanings of the names, function symbols and predicates (other than the identity symbol).
A sentence of FOL is an FO consequence of premises P1,…,Pn if it is a logical consequence of these premises truth when you ignore the meanings of the names, function symbols and predicates (other than the identity symbol).
Two wff’s are FO equivalent if, in any possible circumstance, they are satisfied by the same objects. That is, if, when you replace the free variables in the wff’s with new names, the resulting sentences are logically equivalent.
~"x P(x) ó $x ~P(x)
~$x P(x) ó "x ~P(x)
"x (P(x) ^ Q(x)) ó "x P(x) ^ "x Q(x))
$x (P(x) v Q(x)) ó $x P(x) v $x Q(x)
For any wff P in which x is not free:
"x P ó P
$x P ó P
"x (P v Q(x)) ó P v "x Q(x)
$x (P ^ Q(x)) ó P ^ $x Q(x)
For any wff P(x) and variable y that does not occur in P(x):
"x P(x) ó"y P(y)
$x P(x) ó$y P(y)