Unique nearness structures (Q328632): Difference between revisions

If \((X, \mu)\) is a nearness space in the sense of \textit{H. Herrlich} [General Topology Appl. 4, 191--212 (1974; Zbl 0288.54004)], where \(\mu\) is the collection of uniform covers of \(X\), then \(\tau_{\mu} = \{ U \subset X \mid U = \text{int}_{\mu}U \}\) is a topology on \(X\), where \(\text{int}_{\mu}U = \{x \in U\mid \text{there exists a }\mathcal{U} \in \mu\text{ such that }\text{St}(x, \mathcal{U}) \subset U \}\) and \(\text{St}(x, \mathcal{U}) = \bigcup \{U \in \mathcal{U}\mid x \in U\}\). It is well known that a topological space \((X, \tau)\) has a compatible nearness structure, i.e., a nearness structure \(\mu\) on \(X\) such that \(\tau_{\mu} = \tau\), if and only if \((X, \tau)\) is an \(\text{R}_{0}\)-space. This class of topological spaces, sometimes also called symmetric spaces, was introduced by \textit{N. A. Shanin} in [C. R. (Dokl.) Acad. Sci. URSS, n. Ser. 38, 110--113 (1943; Zbl 0061.39607)] and rediscovered by \textit{A. S. Davis} in [Am. Math. Mon. 68, 886--894 (1961; Zbl 0106.15504)]. It contains all \(\text{T}_{1}\)-spaces. The authors prove that a \(\text{T}_{1}\)-space has precisely one compatible nearness structure if and only if it contains a point for which all of its neighbourhoods are cofinite. It follows that a \(\text{T}_{2}\)-space has precisely one compatible nearness structure if and only if it is finite or if it is a one-point-compactification of an infinite discrete topological space.