Deductive System for Propositional Calculus: Axioms, Rules, and Proofs, Study notes of Mathematics

The concept of a deductive system for propositional calculus, focusing on the language l0, its axioms, and rules of inference. It also covers the definition of deducibility and provides examples of deductions. The document concludes with the soundness theorem, stating that the system is sound.

Typology: Study notes

2010/2011

Uploaded on 09/08/2011

dukenukem
dukenukem 🇬🇧

3.9

(8)

240 documents

1 / 8

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
6. A deductive system for
propositional calculus
We have indtroduced logical consequence’:
Γ|=φ whenever (each formula of) Γ is
true so is φ
But we don’t know yet how to give an ac-
tual proof of φfrom the hypotheses Γ.
Aproof should be a finite sequence
φ1, φ2, . . . , φnof statements such that
either φiΓ
or φiis some axiom (which should clearly
be true)
or φishould follow from previous φj’s by
some rule of inference
AND φ=φn
Lecture 5 - 1/8
pf3
pf4
pf5
pf8

Partial preview of the text

Download Deductive System for Propositional Calculus: Axioms, Rules, and Proofs and more Study notes Mathematics in PDF only on Docsity!

6. A deductive system for

propositional calculus

  • We have indtroduced ‘logical consequence’: Γ |= φ – whenever (each formula of) Γ is true so is φ
  • But we don’t know yet how to give an ac- tual proof of φ from the hypotheses Γ.
  • A proof should be a finite sequence φ 1 , φ 2 ,... , φn of statements such that - either φi ∈ Γ - or φi is some axiom (which should clearly be true) - or φi should follow from previous φj’s by some rule of inference - AND φ = φn

6.1 Definition Let L 0 := L[{¬, →}] (which is an adequate lan-

guage). Then the system L 0 consists of the

following axioms and rules:

Axioms

An axiom of L 0 is any formula of the following

form (α, β, γ ∈ Form(L 0 )):

A1 (α → (β → α))

A2 (((α → (β → γ)) → ((α → β) → (α → γ)))

A3 ((¬β → ¬α) → (α → β))

Rules of inference

Only one: modus ponens

(for any α, β ∈ Form(L 0 )) MP From α and (α → β) infer β.

6.3 Example For any φ ∈ Form(L 0 )

(φ → φ)

is a theorem of L 0.

Proof:

α 1 (φ → (φ → φ)) [A1 with α = β = φ] α 2 (φ → ((φ → φ) → φ)) [A1 with α = φ, β = (φ → φ)] α 3 ((φ → ((φ → φ) → φ)) → → ((φ → (φ → φ)) → (φ → φ))) [A2 with α = φ, β = (φ → φ), γ = φ] α 4 ((φ → (φ → φ)) → (φ → φ)) [MP α 2 , α 3 ] α 5 (φ → φ) [MP α 1 , α 4 ]

Thus, α 1 , α 2 ,... , α 5 is a deduction of (φ → φ) in L 0.

2

6.4 Example

For any φ, ψ ∈ Form(L 0 ):

{φ, ¬φ} ⊢ ψ

Proof:

α 1 (¬φ → (¬ψ → ¬φ))

[A1 with α = ¬φ, β = ¬ψ]

α 2 ¬φ [∈ Γ]

α 3 (¬ψ → ¬φ) [MP α 1 , α 2 ]

α 4 ((¬ψ → ¬φ) → (φ → ψ))

[A3 with α = φ, β = ψ] α 5 (φ → ψ) [MP α 3 , α 4 ]

α 6 φ [∈ Γ]

α 7 ψ [MP α 5 , α 6 ]

i = 1 either α 1 is an axiom, so v˜(α 1 ) = T or α 1 ∈ Γ, so, by hypothesis, v˜(α 1 ) = T.

Induction step Suppose (⋆) is true for some i < m. Consider αi+1.

Either αi+1 is an axiom or αi+1 ∈ Γ, so v˜(αi+1) = T as above,

or else there are j 6 = k < i + 1 such that αj = (αk → αi+1).

By induction hypothesis

v^ ˜(αk) = v˜(αj) = v˜((αk → αi+1)) = T.

But then, by tt →, v˜(αi+1) = T (since T → F is F ).

2

For the proof of the converse

Completeness Theorem

If Γ |= α then Γ ⊢ α.

we first prove

6.6 The Deduction Theorem for L 0

For any Γ ⊆ Form(L 0 ) and

for any α, β ∈ Form(L 0 ):

if Γ ∪ {α} ⊢ β then Γ ⊢ (α → β).