nLab
Löwenheim-Skolem theorem

Contents

Contents

Idea

The Löwenheim-Skolem theorem is a basic result in the model theory of first-order logic and is part of a family of closely related theorems that concern the relation between structures or models of first-order theories of different cardinality.

One idea behind it is that finitary first-order logic by itself cannot pin down the cardinality of infinite structures or models: roughly speaking, Löwenheim-Skolem says that given a such structure MM, one can construct elementary substructures or elementary extensions NN of different cardinalities. This gives an elementary reason why one cannot expect absolute categoricity for first-order theories; after this theorem was introduced, attention shifted to the question of categoricity in powers κ\kappa (i.e., how many models there are of a given cardinality κ\kappa), leading up eventually to Morley’s categoricity results and, further down the road and in much greater generality, to Shelah’s introduction of stability theory.

Historically, the theorem was one of the first results in model theory with papers published by Leopold Löwenheim (1915) and Thoralf Skolem (1920) which helped establishing first-order logic in the center of mathematical logic. It was however conceived by Skolem in view of a problem he saw with first-order set theory, now called “Skolem’s paradox”, where one could construct a countable model of ZFC, with the paradoxical consequence that (externally speaking) the model admits no uncountable sets but yet (internally speaking) it has them. As usually happens with “paradoxes” (cf. Banach-Tarski paradox), this phenomenon was over time digested and accepted as the ordinary fact of the matter without qualms, and just one more illustration of how one must simply retrain one’s intuition to fit the facts.

Although the theorem fails in second-order logic (for example, note one has absolute categoricity for the concept of complete ordered field), variants hold in infinitary logic with countably infinite disjunctions and in categorical logic for sketches. The theorem plays an important role in abstract model theory e.g. in Lindström's theorems on the characterization of standard first-order logic.1

Statements and proofs

One frequently sees the Löwenheim-Skolem theorem separated into an “upward” version (which produces elementary extensions) and a “downward” part (producing elementary substructures). The upward version can be derived just by a clever application of the compactness theorem; the downward version involves a construction called “Skolemization”. Both proofs involve modifying the language.

Let Σ\Sigma be a first-order signature and let TT be a set of sentences of cardinality β\beta. The upward version states the following: suppose TT has a model of cardinality κ 0\kappa \geq \aleph_{0}, then for every cardinal λmax( 0,β)\lambda \geq \max(\aleph_{0}, \beta), TT has a model of cardinality λ\lambda. The downward version, on the other hand, states the following: suppose TT has a model of infinite cardinality, then TT has a model of cardinality 0\aleph_{0}.

References

  • Jiří Adámek, Jiří Rosický, Locally presentable and accessible categories , Cambridge UP 1994.

  • J. van Heijenoort (ed.), From Frege to Gödel - A Source Book in Mathematical Logic 1879-1931 , Harvard UP 1967.

  • L. Löwenheim, Über die Möglichkeiten im Relativkalkül , Math. Ann. 76 (1915) pp.447-470. (gdz; English transl. in van Heijenoort (1967) pp.228-251)

  • Michael Makkai, Robert Paré, Accessible categories: The foundations of categorical model theory , Contemporary Mathematics 104. American Mathematical

    Society, Rhode Island, 1989.

  • Thoralf Skolem, Einige Bemerkungen zur axiomatischen Begründung der Mengenlehre, Mathematikerkongressen i Helsingfor 4-7 Juli 1922. English transl. pp.290-301 of van Heijenoort (ed.), From Frege to Gödel , Harvard UP 1967.

  • M. Zawadowski, The Skolem-Löwenheim theorem in toposes , Studia Logica 42 (1983) pp.461-475.

  • M. Zawadowski, The Skolem-Löwenheim theorem in toposes II , Studia Logica 44 (1985) pp.25-38.

  • M. Manzano, Model Theory, Clarendon Press (1999)


  1. As a theorem that fails to hold is hardly a theorem at all, one prefers in the context of abstract model theory to speak of the Löwenheim-Skolem property.

Last revised on December 2, 2018 at 09:12:05. See the history of this page for a list of all contributions to it.