Löb's theorem in nLab

5 min read Original article ↗
Contents

Context

Foundations

foundations

The basis of it all

 Set theory

set theory

Foundational axioms

foundational axioms

Removing axioms

Type theory

natural deduction metalanguage, practical foundations

  1. type formation rule
  2. term introduction rule
  3. term elimination rule
  4. computation rule

type theory (dependent, intensional, observational type theory, homotopy type theory)

syntax object language

homotopy levels

semantics

Contents

Idea

Löb’s theorem in its original form is a generalization of the Gödel incompleteness theorem whose formulation lends itself to the tools of type theory and modal logic.

Löb’s theorem states that to prove that a proposition is provable, it is sufficient to prove the proposition under the assumption that it is provable. Since the Curry-Howard isomorphism identifies formal proofs with abstract syntax trees of programs; Löb’s theorem implies, for total languages which validate it, that self-interpreters are impossible. (Gross-Gallagher-Fallenstein 16)

In provability logic the abstract statement is considered in itself as an axiom on a modal operator \Box interpreted as the modality “is provable”. In this form the statement reads formally:

(PP)P \Box(\Box P \to P) \to \Box P

for any proposition PP (“Löb’s axiom”).

This reduces to an incompleteness theorem when taking P=P = false and using that

  1. negation is ¬P=(Pfalse)\not P = (P \to false);

  2. consistency means that P¬¬P\Box P \to \not \Box \not P

(falsefalse)false (¬false)false (¬false)¬¬false (¬false)false ¬¬false \begin{aligned} & \Box (\Box false \to false) \to \Box false \\ \Rightarrow \;\; & \Box ( \not \Box false ) \to \Box false \\ \Rightarrow \;\; & \Box ( \not \Box false ) \to \not \Box \not false \\ \Rightarrow \;\; & \Box ( \not \Box false ) \to false \\ \Rightarrow \;\; & \not \Box \not \Box false \end{aligned}

Where the last line reads in words “It is not provable that false is not provable.”

Guarded Recursion Variant

A variant of the Löb axiom is used in guarded recursion and synthetic guarded domain theory, which uses a modality \blacktriangleright, usually pronounced “later”. Then the Löb induction axiom is for any proposition PP,

(PP)P(\blacktriangleright P \to P) \to P

In a setting with the principle of unique choice, this can be used to prove the existence of all guarded fixed points.

Note that the assumptions about the later modality \blacktriangleright are usually quite different from the provability modality \Box. For instance, \Box is usually assumed to have some subset of the properties of a comonadic modality, but \blacktriangleright typically satisfies PPP \to \blacktriangleright P.

References

See also

Unified account of Löb's theorem/Gödel's second incompleteness theorem, Kripke semantics for provability logic, and guarded recursion, via essentially algebraic theories which “self-internalize”:

Last revised on October 5, 2024 at 18:49:29. See the history of this page for a list of all contributions to it.