The properties of \(\models\)-filters of a topological system (Q780198): Difference between revisions

There exists the concept of topological system of [\textit{S. Vickers}, Topology via logic. Cambridge etc.: Cambridge University Press (1989; Zbl 0668.54001)], which provides a common framework for both point-set and point-free topology. Recall that a \textit{topological system} is a triple \(D=(\text{pt} D, \Omega D, \models)\), where \(\text{pt} D\) is a set, \(\Omega D\) is a frame (see, e.g., [\textit{P. T. Johnstone}, Stone spaces. Cambridge etc.: Cambridge University Press (1982; Zbl 0499.54001)] for a convenient introduction into the theory of frames), and \(\models\) is a relation on \((\text{pt} D, \Omega D)\) (called \textit{satisfaction relation}) such that: (1) for every \(S\subseteq\Omega D\), \(x\models\bigvee S\) iff \(x\models s\) for some \(s\in S\); (2) for every finite \(S\subseteq\Omega D\), \(x\models\bigwedge S\) iff \(x\models s\) for every \(s\in S\). If \(x\models a\), then \(x\) \emph{satisfies} \(a\). Every topological space \((X,\tau)\) gives a topological system \(D=(X,\tau,\in)\), where ``\(\in\)'' is the standard ``element of'' relation between points and sets. Given a subset \(S\subseteq\text{pt} D\), the present paper considers the set \(F(S)=\{a\in\Omega D\mid s\models a\text{ for every }s\in S\}\) (in case of a topological system arising from a topological space as above, \(F(S)\) is the set of all open sets containing \(S\)). \(F(S)\) is shown to be a lattice filter (Lemma~3.1 on page~12952), which is thus called \textit{\(\models\)-filter}. Further, the authors prove that the set \(L=\{F(S)\mid S\subseteq\text{pt}D\}\), equipped with a partial order defined by \(F(S_1)\leqslant F(S_2)\) if and only \(F(S_1)\subseteq F(S_2)\), is a coframe (Theorem~3.8 on page~12954) and show an explicit construction of joins and meets in \(L\) (Theorem~3.11 on page~12955). The paper finally presents two topologies on the set \(\mathrm{MF}(L)\) of maximal \(\models\)-filters of \(L\). The first (resp. the second) one makes the set \(\mathrm{MF}(L)\) into a \(T_2\) (resp. \(T_1\)) topological space. In particular, the first topology is given by \(\{\mu(\{F(A_i)\}_{i\in I})\mid \{F(A_i)\}_{i\in I}\subseteq L\}\), where \(\mu(\{F(A_i)\}_{i\in I})=\{F(S)\in\mathrm{MF}(L)\mid F(A_i)\not\subseteq F(S)\text{ for some }i\in I\}\), and the second one is given by \(\{\nu(\{F(A_i)\}_{i\in I})\mid \{F(A_i)\}_{i\in I}\subseteq L\}\), where \(\nu(\{F(A_i)\}_{i\in I})=\{F(S)\in MF(L)\mid F(A_i)\subseteq F(S)\text{ for some }i\in I\}\). The respective results can be found in Theorem~4.6 on page~12958 and Theorem~4.11 on page~12959. The paper is not quite well written (there is a number of typos), but contains most of its required preliminaries, is easy to read, and will be of interest to all those researchers, who study topological systems and point-free topology.