In this note I sketch the proof of the Löwenheim-Skølem (or Löwenheim-Skølem-Tarski) theorem for first order theories. This basic result of model theory is really a consequence of a set theoretic combinatorial lemma, as the proof will demonstrate.

Let be a first order language, understood as a set of constant, function, and relation symbols. Let

so is unless is finite, in which case we take Talking about rather than simplifies the presentation slightly.

The Löwenheim-Skølem theorem is concerned with the possible infinite sizes of models of first order theories. Of course, a theory could only have finite models; the result does not say anything about if that is the case.

Theorem 1If is a first order theory in a language and there is at least one infinite model of then there are models of of size for all

We will prove a more precise statement. Before stating it, note that it is possible to have a theory in some uncountable language such that has models of certain infinite sizes, but not all. Theorem 1 does not say anything about infinite models of of size What cardinals in this range are the possible sizes of models of is actually a rather difficult problem, and we will not address it.

To state the more precise version that we will prove, we need to review the notion of *elementary substructure*. Thoughout the note we use the notational convention that if is a -structure, its underlying *universe* is

Recall that if and are -structures, we say that

( *is a substructure of* ) iff:

- .
- for any and any -ary relation symbol in
- for all constant symbols in
- for any and any -ary function symbol in

For example, a subgroup of a group is simply a subset such that and is closed under A subgroup may have significantly different properties than the group it comes from. Consider, for example,

A substructure of is *elementary*, in symbols

iff for any -formula and any we have that

iff

This is a much more strict notion than simply being a substructure. It is easy to check that is not elementary in for example.

Note that if then, for any -theory we have that iff (Simply consider sentences in the definition above.)

Here is a natural example of how to obtain elementary extensions of a given model Expand the language into a larger language by adding to a set of many new constant symbols for We can expand to an -structure simply by setting Let denote this expanded structure. Consider the theory where the theory is of course considered in the language Let be any model of Then there is a natural way of identifying with an elementary extension of More precisely, the map given by is *elementary*, meaning that

iff

for any -formula and any

It easily follows that the pointwise image is the universe of an elementary substructure of that is isomorphic to so we might as well assume that is the identity. By considering the restriction of to language (i.e., by “forgetting” about the new constant symbols), we obtain this way an elementary extension of

In fact, it is easy to check that if then there is a natural expansion of to an -structure that models so all elementary extensions of arise in this fashion. From the observations above, Theorem 1 then follows from the following more precise version, where we write for and say that is infinite iff is.

Theorem 2Let be an infinite -structure.

- For any cardinal there is with
- For any cardinal and any with there is such that and

We note first that item 1 follows from item 2 and compactness: Consider the language obtained by adding to a set of many new constant symbols and many new constant symbols In this language, consider the theory obtained by adding to the axioms for all where is as above. This theory is consistent, by compactness (using that is infinite), so it has models. But any model of can be naturally identified with an elementary extension of and has size at least By item 2, there is an elementary substructure of that contains and has size precisely Since it follows that the reduct of to language is indeed an elementary extension of of size

To prove item 2, we expand the language in a different manner.

For this, we need the notion of *Skølem functions*. Syntactically, a Skølem function for a formula is simply a new function symbol for an -ary function. Semantically: Given a structure in a language containing and we call the interpretaion a Skølem function iff, whenever and

then in fact

We say that a language *is closed under Skølem functions* if a function symbol is in the language for each in the language. We say that a structure in this language *has Skølem functions* if it satisfies all axioms of the form

Note that using the axiom of choice it is straightforward to define interpretations of all the function symbols so that these axioms are satisfied: Simply fix a well-ordering of and take as the least (in the sense of the well-ordering) that witnesses if such is the case, or as the least element of otherwise.

It is straightforward to expand a language to one closed under Skølem functions: Given let and set is an -formula for all Finally, set and note that is as wanted.

Moreover, as is easily verified by induction on This means that to prove item 2, we may as well assume that is closed under Skølem functions and has Skølem functions to begin with.

The point of all this is that if has Skølem functions and , then automatically This follows from the following test for elementarity:

Lemma 3 (Tarski-Vaught criterion)Suppose and that whenever and

then there exists such that

Then

*Proof:* By induction in the complexity of

The usefulness of the test derives from the fact that it only mentions satisfaction in rather than in both and

To see how Lemma 3 allows us to conclude the claim made above, suppose that has Skølem functions, and that is a substructure of Then, automatically, the hypothesis of Lemma 3 is satisfied, since must be closed under the restrictions of all the functions being a substructure, meaning that whenever It then follows that as claimed.

Moreover, note that if and is (nonempty and) closed under all the Skølem functions from then is the universe of a (necessarily elementary) substructure of To see this, note that from the definition of substructure given above, all that we require is that for all constant symbols in and that whenever for all function symbols in

But consider Then for any so if is nonempty and as above, then contains the interpretaions of all the constant symbols from the language. (The only reason we added to the formula rather than just taking is to avoid having to be -ary, since our official definition of function symbols requires this not to be the case. Similarly, consider Then

We have finally arrived at the combinatorial core of Theorem 2. Some notation is useful (we follow Kunen **Set theory. An introduction to independence proofs** in the remainder of this note).

For say that is an -ary function on a set iff and, as usual, or and (This is a natural extension of the notion for Note that and any is simply picking an element of here we are just identifying and this element.) Say that is a *finitary function on * iff is -ary on for some

Say that is *closed under *a finitary function iff where if is -ary, and if is -ary, If is a set of finitary functions on we say that is closed under iff for all The *closure* of under is the -smallest such that and is closed under

From the preliminaries above, it is clear that the following completes the proof of Theorem 2, since we can take as the set in item 2 (or rather, a superset of of size exactly ), as the cardinal as the set and as the set of (interpretations of) Skølem functions on

Lemma 4Suppose that is infinite, that and that is a set of at most many finitary functions on Then the closure of under exists and has size at most

*Proof:* Given write for Set and for all Finally, set Then, by induction on each is contained in any subset of closed under that contains Since itself is closed under it follows that *is* the closure of By induction, for all and so the same holds for

*Typeset using LaTeX2WP. Here is a printable version of this post.*

[…] This can be established by an argument very similar to the one we used in the notes on the Löwenheim-Sko lem theorem. Namely, for and each of its subformulas consider Sko lem functions from Let be the set of […]