A complete theory is categorical theory in the infinite cardinal , or -categorical, when it has a model of cardinality and any two of its models of cardinality are isomorphic. Equivalently, has exactly one model of cardinality up to isomorphism.
The theory is not aleph-zero-categorical. Introduce a constant and add the formulasEvery finite subset is realized in by taking to be a nonzero common multiple of the finitely many displayed integers. The compactness theorem therefore gives a model of containing a nonzero infinitely divisible element of an abelian group. The Downward Lowenheim-Skolem theorem gives such a model that is countable.
No nonzero integer is divisible by every positive integer, so this countable model is not isomorphic to . This is the nonstandard model of the additive integers obstruction to categoricity.
The group is a vector space over a finite field, namely , because every element has order at most two. It is infinite-dimensional. The complete first-order theory of infinite-dimensional -vector spaces says, for each , that there are linearly independent vectors; the usual elimination argument for vector spaces shows that all infinite-dimensional -vector spaces are elementarily equivalent.
Every countably infinite model of this theory has dimension : finite dimension would make it finite, while uncountable dimension would make its underlying set uncountable. Any two vector spaces over the same field with the same dimension are isomorphic. Therefore is aleph-zero-categorical, as recorded by the aleph-zero-categoricity of an infinite-dimensional vector space over a finite field.
- ;
- if , then ;
- if and , then .
These conditions imply .
An ultrafilter is a proper filter maximal under inclusion. Equivalently, for every , exactly one of and belongs to .
If contains no finite set, then every cofinite set belongs to it: for a finite , one has , so the ultrafilter alternative forces . Thus contains the cofinite filter.
Suppose instead that a finite set belongs to . If none of its singleton subsets belonged to , all their complements would belong to , and intersecting those complements with would put the empty set in . Hence for some .
Upward closure then puts every subset containing in , while no subset omitting can belong to it. Thereforethe principal ultrafilter at . Together with part i, this proves the dichotomy.
Yes. Choose the principal ultrafilter at any . Evaluation at the th coordinate giveswhich is a finite cyclic group of order .
No. Suppose a formula with parameters defined a function that was surjective and not injective. The assertions that the formula defines a function, that the function is surjective, and that it is not injective are all first-order statements about that formula and those parameters. By Łoś theorem, they would hold simultaneously in for -almost every .
Every surjective self-map of a finite set is injective, so no finite factor can satisfy those statements. This contradiction shows that the ultraproduct has no such definable function.
Yes. Choose a nonprincipal ultrafilter on and letFor each fixed positive integer , the equality holds in the th factor exactly when divides . Only finitely many positive integers divide , so the cofinite set of indices for which belongs to . Łoś theorem gives in for every . Thus has infinite order of a group element.
Use the finite-partial-isomorphism criterion for quantifier elimination. Let and let be an isomorphism between finite suborders. For , its position relative to is one of the finitely many open intervals determined by , or one of the two exterior rays. The corresponding interval or ray determined by is nonempty because the orders are dense and have no endpoints. Choose there. Then remains a partial order isomorphism.
The same argument extends in the other direction. The back-and-forth method criterion therefore applies, proving quantifier elimination for dense linear orders without endpoints. Hence DLO eliminates quantifiers.
The type space is the set of all complete types in one free variable over the parameter set that are consistent with together with the diagram of those parameters. Thus each member chooses, for every formula with from , exactly one of and , consistently and completely.
A type is an isolated type when some formula isolates it: is the unique complete type containing . Equivalently, implies every formula in modulo the complete theory with the named parameters.
By quantifier elimination, a one-type is determined entirely by the position of relative to the natural-number parameters. Assuming , the isolated types and isolating formulas are:
- , for each ;
- ;
- , for each .
Each formula fixes every comparison of with every natural number, so it determines a complete type. These are precisely the isolated members of the one-types over the natural numbers in the rational order.
There is exactly one non-isolated type:It is consistent by the compactness theorem, since every finite subset is realized by a sufficiently large rational number. It is complete by quantifier elimination, because it decides every comparison with a parameter from .
No formula isolates it. Any formula belongs to only through finitely many natural-number parameters; after quantifier elimination it holds throughout some final ray. It is consequently also satisfied by a sufficiently large natural number, whose equality type differs from . Thus is non-isolated, and the list in part i exhausts all other cuts of .
Articles by others on the same topic
There are currently no matching articles.