models constructed from constants
The definition of a structure and of the satisfaction relation isnice, but it raises the following question : how do we get models inthe first place? The most basic construction for models offirst-order theory is the construction that uses constants. Throughoutthis entry, is a fixed first-order language.
Let be a set of constant symbols of , and be a theory in. Then we say is a set of witnesses for if and onlyif for every formula with at most one free variable ,we have for some .
Lemma.Let is any consistentset of sentences of , and is a set of new symbolssuch that . Let . Then there is a consistentset extending and which has as set ofwitnesses.
Lemma.If is a consistent theory in , and is a set of witnessesfor in , then has a model whose elements are the constantsin .
Proof:Let be the signature for . If is a consistent set ofsentences of , then there is a maximal consistent .Note that and have the same sets of witnesses. As everymodel of is also a model of , we may assume is maximalconsistent.
We let the universe of be the set of equivalence classes
, where if and only if . As ismaximal consistent, this is an equivalence relation. We interpret thenon-logical symbols as follows :
- 1.
if and only if ;
- 2.
Constant symbols are interpreted in the obvious way, i.e. if is a constant symbol, then ;
- 3.
If is an -ary relation symbol, then if and only if ;
- 4.
If is an -any function symbol, then if and only if .
From the fact that is maximal consistent, and is anequivalence relation, we get that the operations are well-defined (itis not so simple, i’ll write it out later).The proof that is a straightforward induction
on thecomplexity of the formulas of .
Corollary.(The extended completeness theorem) A set of formulas of isconsistent if and only if it has a model (regardless of whether or not has witnesses for ).
Proof:First add a set of new constants to , and expand to insuch a way that is a set of witnesses for . Then expand to a maximal consistent set . This set has a model consisting ofthe constants in , and is also a model of .
Corollary.(Compactness theorem) A set of sentences of has a model ifand only if every finite subset of has a model.
Proof:Replace “has a model” by “is consistent”, and apply the syntacticcompactness theorem.
Corollary.(Gödel’s completeness theorem)Let be a consistent set of formulas of . ThenA sentence is a theorem of if and only if it is true inevery model of .
Proof:If is not a theorem of , then is consistentwith , so has a model , in which cannot be true.
Corollary.(Downward Löwenheim-Skolem theorem) If has a model,then it has a model of power at most .
Proof:If has a model, then it is consistent. The model constructed fromconstants has power at most (because we must add at most many new constants).
Most of the treatment found in this entry can be read in more detailsin Chang and Keisler’s book Model Theory.