Preservation Theorems Through the Lens of Topology
Aliaume Lopez
TL;DR
This paper develops a unifying topological framework for preservation theorems by introducing logically presented pre-spectral spaces (lpps) and diagram bases, which connect semantic preservation to syntactic fragments via topological openness. A generic preservation theorem shows that preservation holds exactly when the definable opens form compact sets in a topology generated by FO-definable sets, enabling construction of new preservation results by composing spaces through morphisms, subspaces, finite sums, and finite products. The authors relate the framework to Noetherian and spectral spaces, prove stability under closures and FO-interpretations, and show that Rossman’s finite-homomorphism preservation result extends to classes with the finite model property via projective limits. They further demonstrate that all pre-spectral spaces arise as limits of Noetherian spaces, providing a comprehensive, modular approach to classical and novel preservation theorems with applications to finite structure classes and beyond.
Abstract
In this paper, we introduce a family of topological spaces that captures the existence of preservation theorems. The structure of those spaces allows us to study the relativisation of preservation theorems under suitable definitions of surjective morphisms, subclasses, sums, products, topological closures, and projective limits. Throughout the paper, we also integrate already known results into this new framework and show how it captures th essence of their proofs.
