There are two ways to measure the genericity of expansions: the first is via existential closures (Lascar's way) and the other via Baire categoricity (Truss). The authors' work here is more in the second way, but also relates both ways to each other. Let \(T\) be a theory in \(L\) and let \(N\) be a countable model of \(T\). Further let: \(L_0\supseteq L\) contain finitely many new functions and relations, \(T_0\) be a theory in \(L_0\), and \((N,\sigma)\) be an expansion of \(N\) where \(\sigma\) is a finite sequence of interpretations of the new symbols. The set of all these \(\sigma\)s is topologized by taking \(\{\sigma\mid(N, \sigma)\models\varphi\}\) as a basic open set, \(\varphi\) being a sentence with parameters from \(N\). This topology is also induced by the following metric: since \(N\) is countable, enumerate it, and define \(\text{dist}(\sigma,\tau)\) by means of the first element where \(\sigma\) and \(\tau\) differ. This metric is complete and so the Baire category theorem holds. By definition, \(\sigma\) is Truss generic when \(\{\sigma^g\mid g\in \Aut(N)\}\) is comeager in \(\{\tau\mid (N,\tau)\models T_0\}\). To relate this genericity to existential closures etc., another topology is called for. Let \(Y\) be the set of existentially closed and smooth expansions. Here, the smoothness of \(\sigma\) is defined by: \((N,\sigma)\) realizes all possible types of the form \(p(x)\wedge\psi(x)\) where \(p(x)\) is a type \(\subseteq L\) and \(\psi(x)\in L_0\) is quantifier-free. For a finite tuple of elements of \(N\), \(b\), and \(\sigma\in Y\), define the 1-diagram of \(\sigma\) at \(b\), \(\text{diag}(\sigma,b)\), to be \(\{\varphi(b)\mid q(x)\) is universal or existential and \((N,\sigma)\models\varphi(b)\}\), and let \(D_b\) be the set of all 1-diagrams at \(b\). Further, let \(\pi(b)\) be an infinitary formula \(\exists y[p(x,y)\wedge\psi(x, y)]\), where \(p(x,y)\subseteq L\) is a type and \(\psi(x,y)\in L_0\) is quantifier-free. Topologize \(D_b\) by taking \(\{\text{diag}(\sigma,b)\mid (N,\sigma)\models\pi(b)\}\) as a basic open set. Define \(\sigma\) to be e-atomic when \(\text{diag}(\sigma,b)\) is an isolated point in \(D_b\) for each \(b\). Main results: Assume that \(N\models T\) is saturated and all existentially closed models of \(T_0\) have the same existential theory. Then, a Truss generic exists iff the set of isolated points is dense in \(D_b\) for each \(b\), and \(\sigma\) is Truss generic iff it is e-atomic. When \(T\) is \(\omega\)-categorical and \(T_0\) has a model companion \(T_{\mathrm{mc}}\), \((N,\sigma)\) is an atomic model of \(T_{\mathrm{mc}}\) exactly when \(\sigma\) is Truss generic. Throughout the paper, the authors give a few examples showing existence/non-existence of generics, etc.
- Autour De La Propriété Du Petit Indice
- Comeagre conjugacy classes and free products with amalgamation
- Extending partial isomorphisms of graphs
- Generic Automorphisms of Homogeneous Structures
- Generic expansions of ω-categorical structures and semantics of generalized quantifiers
- Generic structures and simple theories
- scientific article; zbMATH DE number 3908418 (Why is no real title available?)
- scientific article; zbMATH DE number 1007358 (Why is no real title available?)
- Les beaux automorphismes. (The beautiful automorphisms)
- Model companions of \(T_{\Aut}\) for stable \(T\)
- Model companions of theories with an automorphism
- Model Theory
- Model theory of difference fields
- On notions of genericity and mutual genericity
- On the Reconstruction of ℵ0 -Categorical Structures from their Automorphism Groups
- Reconstruction of homogeneous relational structures
- The Small Index Property for ω‐Stable ω‐Categorical Structures and for the Random Graph
- The strict order property and generic automorphisms
- Turbulence, amalgamation, and generic automorphisms of homogeneous structures
This page was built for publication: Generic expansions of countable models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q691128)
