You are here
Home ›a property of truth-value semantics for intuitionistic propositional logic
Primary tabs
a property of truth-value semantics for intuitionistic propositional logic
In this entry, we show the following: if is a tautology of , then is a theorem. First, we need the following lemma, which is the intuitionistic version of one for classical propositional logic, found here. Given an interpretation , define
It is easy to see that for any , for any , so that is always true. In addition, we have the following table:
The proofs of the following lemmas use instances of the theorem schemas below (proofs here):
Lemma 1.
.
Proof.
Since and , by modus ponens and instances of the theorem schema 1 above, we have and . This proves the first three cases.
For the last case, we start with the axiom , or by the deduction theorem. Apply modus ponens twice to instances of schema 1, we get , or by the deduction theorem twice. Again, applying modus ponens twice to instances of 1, we have , or by the deduction theorem. One application of modus ponens to an instance of schema 2, we have , as desired. ∎
Lemma 2.
.
Proof.
Since and , by modus ponens twice to instances of the schema 1, we have and . This settles the last three cases.
For the first case, we use the axiom , which is just , or by the deduction theorem twice. ∎
Lemma 3.
.
Proof.
For the first two, all we need is . To see this, we have deduction
so , or by the deduction theorem. Since is an instance of schema 3, by modus ponens, as desired.
For the third, by the deduction theorem, it is enough to show . Now,
is a deduction of from , and , where is a theorem.
For the last, all we need to show is . We start with , which is an axiom. Applying modus ponens twice to instances of 1, we have , or . ∎
Lemma 4.
Suppose are all the propositional variables in a wff . Then
Proof.
We use induction on the number of primitive logical connectives (, and ) in . If , then is either or a propositional variable . If is , then , or , or . If is , then clearly . Now, if has connectives, and is either , , or , then and has no more than connectives. By induction,
or
By the first three lemmas above, , so by modus ponens twice,
∎
We are now ready for the main result:
Theorem 1.
If is a tautology of , then .
Proof.
Let be any interpretation, then by the last lemma, where are all the propositional variables in . Since is a tautology,
If , then we are done. Otherwise, let and be two interpretations such that for , and and , so that
By applying the deduction theorem twice to each of the above deductive relations, we get
Apply schema 2 to the second deductive relation above, we get
By the deduction theorem once more, we have
With the axiom instance , apply modus ponens to each of the last two deductive relations, we get
so that is removed from the original deductive relation. Continue this process until all of the are removed on the left, and we get
∎
We record to immediate corollaries:
Corollary 1.
If is a tautology of , then .
Proof.
By the theorem, . But , by modus ponens. ∎
In the next corollary, we use and to distinguish that is a theorem of classical and intuitionistic propositional logic respectively.
Corollary 2.
(Glivenko’s Theorem) iff .
Proof.
If , then by the soundness theorem of classical propositional logic, is a tautology of truth-value semantics, which is just , and therefore by the theorem above, .
Conversely, if , then certainly , as PL is a subsystem of PL. Since is a theorem of PL, we get by modus ponens. ∎
In particular, iff , since . In other words, PL is consistent iff PL is.
Mathematics Subject Classification
03B20 Subsystems of classical logic (including intuitionistic logic)- Forums
- Planetary Bugs
- HS/Secondary
- University/Tertiary
- Graduate/Advanced
- Industry/Practice
- Research Topics
- LaTeX help
- Math Comptetitions
- Math History
- Math Humor
- PlanetMath Comments
- PlanetMath System Updates and News
- PlanetMath help
- PlanetMath.ORG
- Strategic Communications Development
- The Math Pub
- Testing messages (ignore)
- Other useful stuff
Recent Activity
new image: sinx_approx.png by jeremyboden
new image: approximation_to_sinx by jeremyboden
new image: approximation_to_sinx by jeremyboden
new question: Solving the word problem for isomorphic groups by mairiwalker
new image: LineDiagrams.jpg by m759
new image: ProjPoints.jpg by m759
new image: AbstrExample3.jpg by m759
new image: four-diamond_figure.jpg by m759
May 16
new problem: Curve fitting using the Exchange Algorithm. by jeremyboden
new question: Undirected graphs and their Chromatic Number by Serchinnho


