Two-variable logic: Difference between revisions

Revision as of 07:21, 14 May 2020

Latest revision as of 13:07, 13 September 2022

(5 intermediate revisions by 3 users not shown)

Line 1:

In [[mathematical logic]] and [[computer science]], '''two-variable logic''' is the [[fragment (logics)|fragment]] of [[first-order logic]] where [[formula (logics)|formulae]] can be written using only two different [[variable (logics)|variable]]s .<ref> L. Henkin. ''Logical systems containing only a finite number of symbols'', Report, Department of Mathematics, University of Montreal, 1967</ref>. This fragment is usually studied without [[function symbol]]s.

{{ref improve|date=August 2014}}

In [[mathematical logic]] and [[computer science]], '''two-variable logic''' is the [[fragment (logics)|fragment]] of [[first-order logic]] where [[formula (logics)|formulae]] can be written using only two different [[variable (logics)|variable]]s <ref> L. Henkin. ''Logical systems containing only a finite number of symbols'', Report, Department of Mathematics, University of Montreal, 1967</ref>. This fragment is usually studied without [[function symbol]]s.

== Decidability ==

Line 7 ⟶ 6:

3, No. 1 (Mar., 1997), pp. 53-69.</ref> This result generalizes results about the decidability of fragments of two-variable logic, such as certain [[description logic]]s; however, some fragments of two-variable logic enjoy a much lower [[Computational complexity theory|computational complexity]] for their satisfiability problems.

By contrast, for the three-variable fragment of [[first-order logic]] without function symbols, satisfiability is undecidable.<ref>A. S. Kahr, Edward F. Moore and Hao Wang. ''Entscheidungsproblem Reduced to the ∀ ∃ ∀ Case'', 1962, noting that their ∀ ∃ ∀ formulas use only three variables.</ref>

== Counting quantifiers ==

The two-variable fragment of first-order logic with no function symbols is known to be decidable even with the addition of [[counting quantifiers]],<ref>E. Grädel, M. Otto and E. Rosen. ''Two-Variable Logic with Counting is Decidable.'', Proceedings of Twelfth Annual IEEE Symposium on Logic in Computer Science, 1997.</ref>, and thus of [[uniqueness quantification]]. This is a more powerful result, as counting quantifiers for high numerical values are not expressible in that logic.

Counting quantifiers actually improve the expressiveness of finite-variable logics as they allow to say that there is a node with <math>n</math> neighbors, namely <math>\Phi = \exists x \exists^{\geq n} y E(x,y)</math>. Without counting quantifiers <math>n+1</math> variables are needed for the same formula.

== Connection to the Weisfeiler-Leman algorithm ==

There is a strong connection between two-variable logic and the Weisfeiler-Leman (or [[Colour refinement algorithm|color refinement]]) algorithm. Given two graphs, then any two nodes have the same stable color in color refinement if and only if they have the same <math>C^2</math> type, that is, they satisfy the same formulas in two-variable logic with counting.<ref>Grohe, Martin. "Finite variable logics in descriptive complexity theory." Bulletin of Symbolic Logic 4.4 (1998): 345-398.</ref>.

== References ==

Line 22 ⟶ 23:

[[Category:Model theory]]

[[Category:Systems of formal logic]]

{{logic-stub}}

{{comp-sci-stub}}