Propositional Logic

Propositional Logic

Propositional Logic: Deriving a logical conclusion by combining many propositions and using formal logic: Hence, determining the truth of arguments.

Argument: A sequence of statements in which the conjunction of the initial statements (the premises, hypothesis) is said to imply the final statement (conclusion)

Symbolic representation of an argument: (P_1 \land P_2 \land P_{...} \land P_n) \rightarrow Q

Example of a logically correct argument:

Valid Argument: Whenever the truth of hypothesis (P_n) leads to the conclusion (Q).

Example of an invalid argument

Proof Sequence (How to Arrive at a Valid Argument)

Proof Sequence; A sequence of well-formed formula in which each well-formed formula is either a hypothesis or the result of applying on the formal system’s derivation rules to earlier well-formed formulas in the sequence.

Example of Proof in Propositional Logic: Prove (A \to B) \land A \to B:

Rules for Propositional Logic

Derivation rules for propositional logic:

Equivalence RulesInference Rules
Allows individual well-formed formulas to be rewrittenAllows new well-formed formulas to be derived
Truth-preserving rulesWork only in one direction

Professor’s Note: Knowledge of all of these rules is vital and will become natural through experience.

★ Equivalence Rules

(Let R and S be statement variables)

ExpressionEquivalent toAbbreviate for Rule
R \lor SS \lor RCommutative, comm.
(R \lor S) \lor Q; (R \land S) \land QR \lor (S \lor Q); R \land (S \land Q)Associative, ass.
(R \lor S)'; (R \land S)'R' \land S; R' \lor S'De-Morgan’s Laws, dm
R \to SR' \lor SImplication, Imp
R(R')'Double Negation, dn
P \leftrightarrow Q(P \to Q) \land (Q \to P)Equivalence, equ

Example of using equivalence rule in part of a proof sequence:

Problem: Simplify (A ' \lor B') \lor C

  1. (A ' \lor B') \lor C
  2. (A \land B)' \lor C; 1, De Morgan
  3. (A \land B) \to C; 2, De Morgan

Professor’s Note: When working on proofs there are no hard-and-fast rules. Trial and error, reason backward and forward, rearrange things to look at them differently, etc.

★ Inference Rules

(Let R and S be statement variables)

FromCan DeriveAbbreviation for Rule
R, R \to SS is trueModus Ponens; mp
R \to S, S'R'Modus tollens, mt
R, SR \land SConjunction; con
R \land SR, SSimplification; sim
RR \lor SAddition; add

Note: Reading the “From” column, each statement is assumed to be true.

Example: “If it is bright and sunny today, then I will wear my sunglasses”

  1. Modus Ponens: It is bright and sunny today. Therefore, I will wear my sunglasses.
  2. Modus Tollens: I will not wear my sunglasses. Therefore, it is not bright and sunny today.

Example: Prove (A \to B) \land (B \to C) \land A \to C

  1. A \to B; hyp
  2. B \to C; hyp
  3. A; hyp
  4. B; 1,3, mp
  5. C; 2,4, mp

Professor’s Note: Every line you write down is assumed or derived to be true

Note: Everything to the left of the main implication are the hypothesis (P_n)

Example: Prove ((A \land B) \to C) \land C' \to A' \lor B'

  1. C'; hyp
  2. (A \land B \to C); hyp
  3. (A \land B)'; 1, 2, mt
  4. A ' \lor B'; 3, dm

Deduction Method

To prove an argument of the form:

Deduction method allows for the use of R as an additional hypothesis and thus prove:

Example: Prove (A \to B) \land (B \to C) \to (A \to C)

By deduction method, prove:

(A \to B) \land (B \to C) \land A \to C

  1. A \to B; hyp
  2. B \to C; hyp
  3. A; hyp
  4. B; 1, 3 mp
  5. C; 2, 4 mp

★ More Inference Rules

(These rules can be derived using the previous rules. Memorizing these allow us to solve proofs with fewer steps.)

FromCan DeriveAbbreviation
P \to Q, Q \to RP \to RHypothetical syllogism; hs
P \lor Q, P'QDisjunctive syllogism; ds
P \to QQ' \to P'Contraposition; cont
Q' \to P'P \to QContraposition; cont
PP \land PSelf-reference; self
P \lor PPSelf-reference; self
(P \land Q) \to RP \to (Q \to R)Exportation; exp
P, P'QInconsistency; inc
P \land (Q \lor R)(P \land Q) \lor (P \land R)Distributive; dist
P \lor (Q \land R)(P \lor Q) \land (P \lor R)Distributive; dist

More on Inference Rules

Some Notes:

Reminder: Only a couple inference rules are bidirectional, most are one-directional

Examples: Solving Proofs

Example: Proving Inconsistency

  1. P, hyp
  2. P', hyp
  3. P \lor Q, 1, add
  4. Q, 3, 2, ds
  1. P, hyp
  2. P', hyp
  3. P \lor Q, 1, add
  4. Q \lor P, 3, comm
  5. (Q')' \lor P, 4, dn
  6. Q' \to P, 5, imp
  7. (Q')', 2, 6, mt
  8. Q, 7, dn

Example:

  1. A, hyp
  2. B \to C, hyp
  3. (A \land B) \to (D \lor C'), hyp
  4. B, hyp
  5. C, 2,4, mp
  6. A \land B, 1, 4, con
  7. D \lor C', 3,6, mp
  8. D, 5,7, ds

Professor’s Note: Stay focused on the result. A longer path is fine so long as it is a correct path.

Example:

  1. B \to A', hyp
  2. B' \to A', hyp
  3. A \to B, 3, cont
  4. A \to A', 1, 3, hs
  5. A' \lor A', 4, imp
  6. A', 5, self

Professor’s Tip: Turning implications (\to) to disjunctions (\lor) (implication rule) gives us more blocks to play with.

Exercises: Writing & Proving Verbal Arguments

Instructions: Rewrite each verbal argument as a wff, then validate the wff.

Problem I: “Russia was a superior power, and either France was not strong or Napoleon made an error. Napoleon did not make an error, but if the army did not fail, then France was strong. Hence the army failed and Russia was a superior power.”

Let the statement variables be defined as:

Translation: [ A \land (B' \lor C)] \land [C' \land (D' \to B)] \to (D \land A)

Proof:

  1. A \land (B' \lor C), hyp
  2. C' \land (D' \to B), hyp
  3. A, 1, sim
  4. D' \to B, 2, sim
  5. B' \lor C, 1, sim
  6. C', 2, sim
  7. B', 5, 6, ds
  8. D, 4, 7, mt
  9. (D \land A), 3, 8, con

Problem II: “If the program is efficient, it executes quickly. Either the program is efficient, or it has a bug. However, the program does not execute quickly. Therefore it has a bug. E, Q ,B

Let the statement variables be defined as:

Translation: ( E \to Q ) \land (E \lor B) \land (\lnot Q \to B)

By deduction method, prove ( E \to Q ) \land (E \lor B) \land \lnot Q \to B:

  1. E \to Q, hypothesis
  2. E \lor B, hypothesis
  3. \lnot Q, hypothesis
  4. \lnot E, 1,3 modus tollens
  5. B 2,4 disjunctive syllogism

Problem III: “The crop is good, but there is not enough water. If there is a lot of rain or not a lot of sun, then there is enough water. Therefore the crop is good and there is a lot of sun. C, W, R, S

Let the statement variables be defined as:

Translation: (C \land W') \land [(R \lor S') \to W] \to (C \land S)

Proof:

  1. C \land W', hypothesis
  2. (R \lor S') \to W, hypothesis
  3. W', 1 simplification
  4. C, 1 simplification
  5. (R \lor S')' 2,3 disjunctive syllogism
  6. R' \land S 5 de-morgan
  7. S 6 simplification
  8. C \land S 4,7 conjunction