• Aucun résultat trouvé

Intuitionistic Logic

N/A
N/A
Protected

Academic year: 2022

Partager "Intuitionistic Logic"

Copied!
89
0
0

Texte intégral

(1)

Intuitionistic Logic

Akim Demaille [email protected]

EPITA — École Pour l’Informatique et les Techniques Avancées

June 10, 2016

(2)

Intuitionistic Logic

1 Constructivity

2 Intuitionistic Logic

A. Demaille Intuitionistic Logic 2 / 34

(3)

Constructivity

1 Constructivity

2 Intuitionistic Logic

(4)

Constructivity

Classical logic does not build truth it discovers a preexisting truth

Classical logic assumes facts are either true or false

` A ∨ ¬A Excluded middle, tertium non datur

A. Demaille Intuitionistic Logic 4 / 34

(5)

Constructivity

Classical logic does not build truth it discovers a preexisting truth

Classical logic assumes facts are either true or false

` A ∨ ¬A Excluded middle, tertium non datur

(6)

Constructivity

Classical logic does not build truth it discovers a preexisting truth

Classical logic assumes facts are either true or false

` A ∨ ¬A Excluded middle, tertium non datur

A. Demaille Intuitionistic Logic 4 / 34

(7)

Constructivity

Classical logic does not build truth it discovers a preexisting truth

Classical logic assumes facts are either true or false

` A ∨ ¬A Excluded middle, tertium non datur

(8)

Excluded Middle

XM A ∨ ¬A

· ·

¬¬A ·

¬¬

A [¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction A

A ` A

` ¬A, A ` ¬

`r ∨

` A ∨ ¬A, A

`l ∨

` A ∨ ¬A, A ∨ ¬A

`C

` A ∨ ¬A

A. Demaille Intuitionistic Logic 5 / 34

(9)

Reductio ad Absurdum

In Real Life

A You should respect C’s belief, for all beliefs are of equal validity and cannot be denied.

B What about D’s belief? (Where D believes something that is considered to be wrong by most people, such as nazism or the world being flat)

A I agree it is right to deny D’s belief.

B If it is right to deny D’s belief, it is not true that no belief

can be denied. Therefore, I can deny C’s belief if I can give

reasons that suggest it too is incorrect.

(10)

Reductio ad Absurdum

In Real Life

A You should respect C’s belief, for all beliefs are of equal validity and cannot be denied.

B

1

I deny that belief of yours and believe it to be invalid.

2

According to your statement, this belief of mine (1) is valid, like all other beliefs.

3

However, your statement also contradicts and invalidates mine, being the exact opposite of it.

4

The conclusions of 2 and 3 are incompatible and contradictory, so your statement is logically absurd.

A. Demaille Intuitionistic Logic 7 / 34

(11)

Reductio ad Absurdum

Mathematics: The Smallest Positive Rational

There is no smallest positive rational.

1

Suppose there exists one such rational r

2

r/2 is rational and positive

3

r/2 < r

4

Contradiction

(12)

Reductio ad Absurdum

Mathematics: The Smallest Positive Rational

There is no smallest positive rational.

1

Suppose there exists one such rational r

2

r/2 is rational and positive

3

r/2 < r

4

Contradiction

A. Demaille Intuitionistic Logic 8 / 34

(13)

Reductio ad Absurdum

Mathematics: The Smallest Positive Rational

There is no smallest positive rational.

1

Suppose there exists one such rational r

2

r/2 is rational and positive

3

r/2 < r

4

Contradiction

(14)

Reductio ad Absurdum

Mathematics: The Smallest Positive Rational

There is no smallest positive rational.

1

Suppose there exists one such rational r

2

r/2 is rational and positive

3

r/2 < r

4

Contradiction

A. Demaille Intuitionistic Logic 8 / 34

(15)

Reductio ad Absurdum

Mathematics: The Smallest Positive Rational

There is no smallest positive rational.

1

Suppose there exists one such rational r

2

r/2 is rational and positive

3

r/2 < r

4

Contradiction

(16)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(17)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

(18)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(19)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

(20)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(21)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

(22)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(23)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

(24)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(25)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

(26)

Reductio ad Absurdum

Mathematics: √

2 is irrational

√ 2 is irrational.

1

Assume √

2 is rational: ∃a, b integers st. a/b = √ 2

2

a, b can be taken coprime

3

∴ a 2 /b 2 = 2 and a 2 = 2b 2

4

∴ a 2 is even (a 2 = 2b 2 )

5

∴ a is even

6

Because a is even, ∃k st. a = 2k.

7

We insert the last equation of (3) in (6): 2b 2 = (2k) 2 is equivalent to 2b 2 = 4k 2 is equivalent to b 2 = 2k 2 .

8

Since 2k 2 is even, b 2 is even, hence, b is even

9

By (5) and (8) a, b are even

10

Contradicts 2

A. Demaille Intuitionistic Logic 9 / 34

(27)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2

But it is not known which numbers.

We proved A ∨ B, but neither A nor B .

(28)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2

But it is not known which numbers. We proved A ∨ B, but neither A nor B .

A. Demaille Intuitionistic Logic 10 / 34

(29)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2

But it is not known which numbers.

We proved A ∨ B, but neither A nor B .

(30)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2

But it is not known which numbers. We proved A ∨ B, but neither A nor B .

A. Demaille Intuitionistic Logic 10 / 34

(31)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2

But it is not known which numbers.

We proved A ∨ B, but neither A nor B .

(32)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2 But it is not known which numbers.

We proved A ∨ B, but neither A nor B .

A. Demaille Intuitionistic Logic 10 / 34

(33)

Constructivity

Mathematics: Rationality and Power

There are irrational positive numbers a, b such that a b is rational.

1

√ 2 is known to be irrational

2

Consider √ 2

√ 2 :

1

If it is rational, take a = b = √ 2

2

Otherwise, take a = √ 2

√2

, b = √

2, a

b

= 2 But it is not known which numbers.

We proved A ∨ B, but neither A nor B.

(34)

Constructivity

Mathematics: Unknown Numbers

Let σ be the number defined below. Its value is unknown, but it is rational.

For each decimal digit of π, write 3. Stop if the sequence 0123456789 is found.

1

If 0123456789 occurs in π, then σ = 0, 3 . . . 3 = 10 3.10

k

−1

k

2

If it does not, σ = 0, 3 . . . = 1/3

We proved ∃x.P (x), but know no t : P (t).

A. Demaille Intuitionistic Logic 11 / 34

(35)

Constructivity

Mathematics: Unknown Numbers

Let σ be the number defined below. Its value is unknown, but it is rational.

For each decimal digit of π, write 3. Stop if the sequence 0123456789 is found.

1

If 0123456789 occurs in π, then σ = 0, 3 . . . 3 = 10 3.10

k

−1

k

2

If it does not, σ = 0, 3 . . . = 1/3

We proved ∃x.P (x), but know no t : P (t).

(36)

Constructivity

Mathematics: Unknown Numbers

Let σ be the number defined below. Its value is unknown, but it is rational.

For each decimal digit of π, write 3. Stop if the sequence 0123456789 is found.

1

If 0123456789 occurs in π, then σ = 0, 3 . . . 3 = 10 3.10

k

−1

k

2

If it does not, σ = 0, 3 . . . = 1/3

We proved ∃x.P (x), but know no t : P (t).

A. Demaille Intuitionistic Logic 11 / 34

(37)

Constructivity

Mathematics: Unknown Numbers

Let σ be the number defined below. Its value is unknown, but it is rational.

For each decimal digit of π, write 3. Stop if the sequence 0123456789 is found.

1

If 0123456789 occurs in π, then σ = 0, 3 . . . 3 = 10 3.10

k

−1

k

2

If it does not, σ = 0, 3 . . . = 1/3

We proved ∃x.P (x), but know no t : P (t).

(38)

Constructivity

Disjunction Property

If A ∨ B is provable, then either A or B is provable, and reading the proof tells which one.

Existence Property

If ∃x · A(x) is provable, then reading the proof allows to exhibit a witness t (i.e., such that A(t)).

A. Demaille Intuitionistic Logic 12 / 34

(39)

Constructivity

Disjunction Property

If A ∨ B is provable, then either A or B is provable, and reading the proof tells which one.

Existence Property

If ∃x · A(x) is provable, then reading the proof allows to exhibit a witness t

(i.e., such that A(t)).

(40)

Intuitionistic Logic

1 Constructivity

2 Intuitionistic Logic

NJ: Intuitionistic Natural Deduction LJ: Intuitionistic Sequent Calculus

A. Demaille Intuitionistic Logic 13 / 34

(41)

Intuitionistic Logic

Luitzen Egbertus Jan Brouwer (1881–1966)

(42)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

A. Demaille Intuitionistic Logic 15 / 34

(43)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

(44)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

A. Demaille Intuitionistic Logic 15 / 34

(45)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

(46)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

A. Demaille Intuitionistic Logic 15 / 34

(47)

Intuitionistic Logic

Classical logic focuses on truth (hence truth values) Intuitionistic logic focuses on provability (hence proofs) A is true if it is provable

The excluded middle is. . . excluded

A ` A

` ¬A, A ` ¬

` r∨

` A ∨ ¬A, A

` l ∨

` A ∨ ¬A, A ∨ ¬A

` C

` A ∨ ¬A

(48)

NJ: Intuitionistic Natural Deduction

1 Constructivity

2 Intuitionistic Logic

NJ: Intuitionistic Natural Deduction LJ: Intuitionistic Sequent Calculus

A. Demaille Intuitionistic Logic 16 / 34

(49)

Intuitionistic Natural Deduction

Natural deduction supports very well intuistionistic logic.

In fact, classical logic does not fit well in natural deduction.

[A] ·

· ·

¬A ¬I

· ·

· A

· ·

¬A ·

⊥ ¬E

XM A ∨ ¬A

· ·

¬¬A ·

¬¬

A

[¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction

A

(50)

Intuitionistic Natural Deduction

Natural deduction supports very well intuistionistic logic.

In fact, classical logic does not fit well in natural deduction.

[A] ·

· ·

¬A ¬I

· ·

· A

· ·

¬A ·

⊥ ¬E

XM A ∨ ¬A

· ·

¬¬A ·

¬¬

A

[¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction A

A. Demaille Intuitionistic Logic 17 / 34

(51)

Intuitionistic Natural Deduction

Natural deduction supports very well intuistionistic logic.

In fact, classical logic does not fit well in natural deduction.

[A] ·

· ·

¬A ¬I

· ·

· A

· ·

¬A ·

⊥ ¬E

XM A ∨ ¬A

· ·

¬¬A ·

¬¬

A

[¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction

A

(52)

Intuitionistic Natural Deduction

Natural deduction supports very well intuistionistic logic.

In fact, classical logic does not fit well in natural deduction.

[A] ·

· ·

¬A ¬I

· ·

· A

· ·

¬A ·

⊥ ¬E

XM A ∨ ¬A

· ·

¬¬A ·

¬¬

A

[¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction A

A. Demaille Intuitionistic Logic 17 / 34

(53)

Intuitionistic Natural Deduction

Natural deduction supports very well intuistionistic logic.

In fact, classical logic does not fit well in natural deduction.

[A] ·

· ·

¬A ¬I

· ·

· A

· ·

¬A ·

⊥ ¬E

XM A ∨ ¬A

· ·

¬¬A ·

¬¬ A

[¬A] ·

· · B

[¬A] ·

· ·

¬B

Contradiction

A

(54)

Intuitionistic Negation

[A] ·

· ·

⊥ ¬I

¬A

· · A ·

· ·

¬A ·

¬E

⊥ [A] ·

· ·

B ⇒I A ⇒ B

· ·

· A

· ·

· A ⇒ B

⇒E B

So define ¬A := A ⇒ ⊥.

A. Demaille Intuitionistic Logic 18 / 34

(55)

Intuitionistic Negation

[A] ·

· ·

⊥ ¬I

¬A

· · A ·

· ·

¬A ·

¬E

⊥ [A] ·

· ·

B ⇒I A ⇒ B

· ·

· A

· ·

· A ⇒ B

⇒E B

So define ¬A := A ⇒ ⊥.

(56)

Prove A ` ¬¬A

A [A ⇒ ⊥] 1

⊥ ⇒E

⇒I 1 (A ⇒ ⊥) ⇒ ⊥

A. Demaille Intuitionistic Logic 19 / 34

(57)

Prove A ` ¬¬A

A [A ⇒ ⊥] 1

⊥ ⇒E

⇒I 1

(A ⇒ ⊥) ⇒ ⊥

(58)

Prove ¬¬¬A ` ¬A

[A] 2 [A ⇒ ⊥] 1

⇒E

⊥ ⇒I 1

(A ⇒ ⊥) ⇒ ⊥ ((A ⇒ ⊥) ⇒ ⊥) ⇒ ⊥

⇒E

⊥ ⇒I 2 A ⇒ ⊥

A. Demaille Intuitionistic Logic 20 / 34

(59)

Prove ¬¬¬A ` ¬A

[A] 2 [A ⇒ ⊥] 1

⇒E

⊥ ⇒I 1

(A ⇒ ⊥) ⇒ ⊥ ((A ⇒ ⊥) ⇒ ⊥) ⇒ ⊥

⇒E

⊥ ⇒I 2

A ⇒ ⊥

(60)

Intuitionistic Natural Deduction

[A] ·

· ·

B ⇒I A ⇒ B

· · A ·

· · A ⇒ · B

⇒E B

· ·

⊥ ·

⊥E A A B

∧I A ∧ B

A ∧ B

∧l E A

A ∧ B

∧r E B

· ·

· A ∨lI A ∨ B

· ·

· B ∨rI A ∨ B

· ·

· A ∨ B

[A] ·

· · C

[B ]

· ·

·

C ∨E

C

(61)

LJ: Intuitionistic Sequent Calculus

1 Constructivity

2 Intuitionistic Logic

NJ: Intuitionistic Natural Deduction

LJ: Intuitionistic Sequent Calculus

(62)

LJ — Intuitionistic Sequent Calculus

LJ — Gentzen 1934

Logistischer intuitionistischer Kalkül

A. Demaille Intuitionistic Logic 23 / 34

(63)

LK: Identity Group

Id A ` A

Γ ` A, ∆ Γ 0 , A ` ∆ 0

Cut

Γ, Γ 0 ` ∆,∆ 0

(64)

LK: Identity Group

Id A ` A

Γ ` A, ∆ Γ 0 , A ` ∆ 0 Cut Γ, Γ 0 ` ∆,∆ 0

A. Demaille Intuitionistic Logic 24 / 34

(65)

LJ: Identity Group

Id A ` A

Γ ` A

, ∆

Γ 0 , A ` B Cut Γ, Γ 0 `

∆,

B

(66)

LK: Structural Group

Γ ` ∆

`X Γ ` τ (∆)

Γ ` ∆ X ` σ(Γ) ` ∆ Γ ` ∆

`W Γ ` A, ∆

Γ ` ∆ W`

Γ, A ` ∆ Γ ` A, A, ∆

` C Γ ` A, ∆

Γ, A, A ` ∆ C ` Γ, A ` ∆

A. Demaille Intuitionistic Logic 25 / 34

(67)

LK: Structural Group

Γ ` ∆

`X Γ ` τ (∆)

Γ ` ∆ X ` σ(Γ) ` ∆ Γ ` ∆

`W Γ ` A, ∆

Γ ` ∆ W`

Γ, A ` ∆ Γ ` A, A, ∆

` C Γ ` A, ∆

Γ, A, A ` ∆

C `

Γ, A ` ∆

(68)

LJ: Structural Group

Γ ` ∆

` X Γ ` τ (∆)

Γ ` B X ` σ(Γ) ` B

Γ ` ∆

` W Γ ` A, ∆

Γ ` B W ` Γ, A ` B

Γ ` A, A, ∆

` C Γ ` A, ∆

Γ, A, A ` B C`

Γ, A ` B

A. Demaille Intuitionistic Logic 25 / 34

(69)

LK: Logical Group: Negation

Γ, A ` ∆

` ¬ Γ ` ¬A, ∆

Γ ` A, ∆

¬ `

Γ, ¬A ` ∆

(70)

LK: Logical Group: Negation

Γ, A ` ∆

` ¬ Γ ` ¬A, ∆

Γ ` A, ∆

¬ ` Γ, ¬A ` ∆

A. Demaille Intuitionistic Logic 26 / 34

(71)

LJ: Logical Group: Negation

Γ, A ` ∆

` ¬ Γ ` ¬A, ∆

Γ ` A, ∆

¬ `

Γ, ¬A ` ∆

(72)

LK: Logical Group: Conjunction

Γ ` A, ∆ Γ ` B , ∆

` ∧ Γ ` A ∧ B, ∆

Γ, A ` ∆ l ∧ ` Γ, A ∧ B ` ∆

Γ, B ` ∆ r ∧ ` Γ, A ∧ B ` ∆

A. Demaille Intuitionistic Logic 27 / 34

(73)

LK: Logical Group: Conjunction

Γ ` A, ∆ Γ ` B , ∆

` ∧ Γ ` A ∧ B, ∆

Γ, A ` ∆ l ∧ ` Γ, A ∧ B ` ∆

Γ, B ` ∆

r ∧ `

Γ, A ∧ B ` ∆

(74)

LJ: Logical Group: Conjunction

Γ ` A

, ∆

Γ ` B

, ∆

` ∧ Γ ` A ∧ B

, ∆

Γ, A ` C

l∧ ` Γ, A ∧ B ` C

Γ, B ` C

r∧ ` Γ, A ∧ B ` C

A. Demaille Intuitionistic Logic 27 / 34

(75)

LK: Logical Group: Disjunction

Γ ` A, ∆

` l∨

Γ ` A ∨ B, ∆ Γ ` B , ∆

` r∨

Γ ` A ∨ B, ∆

Γ, A ` ∆ Γ, B ` ∆

∨ `

Γ, A ∨ B ` ∆

(76)

LK: Logical Group: Disjunction

Γ ` A, ∆

` l∨

Γ ` A ∨ B, ∆ Γ ` B , ∆

` r∨

Γ ` A ∨ B, ∆

Γ, A ` ∆ Γ, B ` ∆

∨ ` Γ, A ∨ B ` ∆

A. Demaille Intuitionistic Logic 28 / 34

(77)

LJ: Logical Group: Disjunction

Γ ` A

, ∆

` l∨

Γ ` A ∨ B

, ∆

Γ ` B

, ∆

` r∨

Γ ` A ∨ B

, ∆

Γ, A ` C Γ, B ` C

∨ `

Γ, A ∨ B ` C

(78)

LK: Logical Group: Implication

Γ ` A, ∆ Γ 0 , B ` ∆ 0

⇒`

Γ, Γ 0 , A ⇒ B ` ∆,∆ 0

Γ, A ` B, ∆

`⇒

Γ ` A ⇒ B, ∆

A. Demaille Intuitionistic Logic 29 / 34

(79)

LK: Logical Group: Implication

Γ ` A, ∆ Γ 0 , B ` ∆ 0

⇒`

Γ, Γ 0 , A ⇒ B ` ∆,∆ 0

Γ, A ` B, ∆

`⇒

Γ ` A ⇒ B, ∆

(80)

LJ: Logical Group: Implication

Γ ` A

, ∆

Γ 0 , B ` B 0

⇒`

Γ, Γ 0 , A ⇒ B `

∆,

B 0

Γ, A ` B

, ∆

`⇒

Γ ` A ⇒ B

, ∆

A. Demaille Intuitionistic Logic 29 / 34

(81)

LJ

Id A ` A

Γ ` A Γ 0 , A ` B Cut Γ, Γ 0 ` B

Γ ` B X ` σ(Γ) ` B

Γ ` B W`

Γ, A ` B

Γ, A, A ` B C ` Γ, A ` B Γ ` A Γ ` B

` ∧ Γ ` A ∧ B

Γ, A ` C

l ∧ ` Γ, A ∧ B ` C

Γ, B ` C

r∧ ` Γ, A ∧ B ` C Γ ` A

` l ∨ Γ ` A ∨ B

Γ ` B

` r∨

Γ ` A ∨ B

Γ, A ` C Γ, B ` C

∨ ` Γ, A ∨ B ` C

Γ ` A Γ 0 , B ` C

⇒`

Γ, Γ 0 , A ⇒ B ` C

Γ, A ` B

`⇒

Γ ` A ⇒ B

(82)

Prove A ` ¬¬A

A − ` A + ⊥ ` ⊥ +

⇒` A − , A + ⇒ ⊥ ` ⊥ +

`⇒ A − ` (A + ⇒ ⊥ ) ⇒ ⊥ +

A. Demaille Intuitionistic Logic 31 / 34

(83)

Prove A ` ¬¬A

A ` A ⊥ ` ⊥

⇒`

A, A ⇒ ⊥ ` ⊥

`⇒

A ` (A ⇒ ⊥) ⇒ ⊥

(84)

Prove A ` ¬¬A

A − ` A + ⊥ ` ⊥ +

⇒`

A − , A + ⇒ ⊥ ` ⊥ +

`⇒

A − ` (A + ⇒ ⊥ ) ⇒ ⊥ +

A. Demaille Intuitionistic Logic 31 / 34

(85)

Prove ¬¬¬A ` ¬A

A − ` A + ` ⊥ +

⇒` A − , A + ⇒ ⊥ ` ⊥ +

`⇒

A − ` (A + ⇒ ⊥ − ) ⇒ ⊥ +0 ` ⊥ 0 +

⇒` A − , ((A + ⇒ ⊥ − ) ⇒ ⊥ + ) ⇒ ⊥ 0 ` ⊥ 0 +

`⇒

((A + ⇒ ⊥ − ) ⇒ ⊥ + ) ⇒ ⊥ 0 ` A − ⇒ ⊥ 0 +

Therefore, in intuistionistic logic ¬¬¬A ≡ ¬A, but ¬¬A 6≡ A.

(86)

Prove ¬¬¬A ` ¬A

A ` A ⊥ ` ⊥

⇒`

A, A ⇒ ⊥ ` ⊥

`⇒

A ` (A ⇒ ⊥) ⇒ ⊥ ⊥ 0 ` ⊥ 0

⇒`

A, ((A ⇒ ⊥) ⇒ ⊥) ⇒ ⊥ 0 ` ⊥ 0

`⇒

((A ⇒ ⊥) ⇒ ⊥) ⇒ ⊥ 0 ` A ⇒ ⊥ 0

Therefore, in intuistionistic logic ¬¬¬A ≡ ¬A, but ¬¬A 6≡ A.

A. Demaille Intuitionistic Logic 32 / 34

(87)

Prove ¬¬¬A ` ¬A

A − ` A + ` ⊥ +

⇒`

A − , A + ⇒ ⊥ ` ⊥ +

`⇒

A − ` (A + ⇒ ⊥ − ) ⇒ ⊥ +0 ` ⊥ 0 +

⇒`

A − , ((A + ⇒ ⊥ − ) ⇒ ⊥ + ) ⇒ ⊥ 0 ` ⊥ 0 +

`⇒

((A + ⇒ ⊥ ) ⇒ ⊥ + ) ⇒ ⊥ 0 ` A − ⇒ ⊥ 0 +

Therefore, in intuistionistic logic ¬¬¬A ≡ ¬A, but ¬¬A 6≡ A.

(88)

Recommended Readings

[van Atten, 2009]

The history of intuitionistic logic.

A. Demaille Intuitionistic Logic 33 / 34

(89)

Bibliography I

van Atten, M. (2009).

Stanford Encyclopedia of Philosophy, chapter The Development of Intuitionistic Logic.

The Metaphysics Research Lab.

Références

Documents relatifs

reasonable form of analyticity that can be demanded in Gilp ∗ : for any deriva- tion d, containing both cuts at the sequent and the TND-sequent level, there exists a cut-

In dialogical logic, the particle rules are said to state the local semantics: what is at stake is only the request and the answer corresponding to the utterance of a given

If we can independently prove that only a finite number of different types need to be considered to find a valid proof (this is the case for propositional logic because of

In Section 3 we study the well-known Harrop’s rule, and more precisely the logic KP obtained by adding Harrop’s principle to IPC: we prove good properties like subject reduction,

The protocol to follow in the creation of an abstract machine for a given focused sequent calculus is in the end quite mechanical, as it consists essentially in the identification of

In the Kripke semantics defined for dual-intuitionistic logic (see [Gor00]), the dual of intuitionistic implication is interpreted as being valid in a possible world if it is valid

The arbitrary assignment of positive (synchronous) and negative (asynchronous) bias to atomic formulas can have a major impact on, not the existence of focused proofs, but the shape

The missing piece to decompose cuts to their atomic form as shown in (9) is the rule s↑, dual to the switch s↓ rule, which allows to move a substructure in positive position outside