This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Notes on Isomorphism, Elementary Equivalence, the Generalized LöwenheimSkolemTarski Theorem, Categoricity, NonStandard Models, etc. Kenny Easwaran 05/10/05 1 Preliminaries and Setup I will use the notation “ “ ”, rather than “ F ˚˚1 ”, for the twoplace predicate whose intended (normal) interpretation is the identity relation. And, I will use the notation “ x ‰ y ”, rather than “ „p x “ y q ,” to express the claim that x and y are nonidentical. A theory is just a set of formulas in a particular language. A model for a theory is a interpretation on which every formula in the set is true (and thus every conse quence of the set is true on such an interpretation as well). We will only consider normal models here, i.e. , models where the symbol “ “ ” is interpreted as the identity relation. The size of a model is the cardinality of the domain of the interpretation. Thus, the standard model of any theory of the natural numbers is countable, while any model of the theory consisting just of ľ x ľ y p x “ y q has size 1. A consistent theory is one from which no contradictions can be derived. A max imal consistent theory is a theory T in a language L such that for every sentence S P L , if S R T then T Yt S u is inconsistent. Clearly, any maximal consistent theory is negationcomplete , meaning that for any sentence S P L , either S P T or „ S P T . T is said to be Henkinized if, for every formula A of one free variable, some sentence H – called its Henkin sentence – given as p Ac { v Ą ľ v A q is in T , where v is the one free variable in A and c is some constant. A Henkinized theory is clearly one where if T $ QS = Ad { v for all d in the language, then T $ QS = ľ v A , because some d will be identical to the c in the Henkin sentence. 2 Isomorphism and elementary equivalence Let M 1 and M 2 be two interpretations for a language L . That is, each M i will specify a domain D i , together with an object in that domain for each constant symbol in the language, a function from the ntuples of the domain to the domain for each nplace function symbol, and a set of ntuples from the domain for each nplace relation symbol. In particular, since we’re only considering normal models, both must specify that the symbol “ “ ” gets interpreted as the actual identity relation on the domain. We will say that M 1 is isomorphic to M 2 if there is a way to pair up the elements of the two domains that preserves all the structure mentioned in the language. That is, there should be a function f from D 1 to D 2 such that no two elements of D 1 get sent to the same element of D 2 and every element of D 2 is the image of some element of D 1 under f . In addition, if x is the object assigned to constant c by M 1 , then f p x q must be the object assigned to c by M 2 . Similarly, x x 1 ,..., x n y is in the interpretation of some nplace relation R under M 1 iff x f p x 1 q ,..., f p x n qy is in the interpretation of R under...
View
Full Document
 Spring '07
 FITELSON
 U0, Model theory, Firstorder logic, Henkin

Click to edit the document details