General approach to characterizations of \(C(X)\) (Q408571)

A distributive lattice is said to be a real lattice if it contains a fixed countable dense-in-itself chain that is conditionally complete and has neither upper nor lower bound. Besides, singletons are also considered to be real lattices. This definition is motivated by the fact that for a non-empty space \(X\), the set of all constant functions is a sublattice of \(C(X)\) isomorphic to \(\mathbb R\); if \(X=\emptyset\) then \(C(X)\) is a singleton. A real non-trivial lattice contains the real line \(\mathbb R\) as a sublattice and the morphisms between real lattices are the lattice homomorphisms identical on \(\mathbb R\). The respective category is denoted by \({\mathcal L}_0\). The family of all function spaces \(C(X)\) for a realcompact space \(X\) is also considered a category \(\mathcal F\) whose morphisms are the maps \(C(f): C(X)\to C(Y)\); here \(f:Y\to X\) is a continuous map and \(C(f)(g)=g\circ f\) for any \(g\in C(X)\). Then \(\mathcal F\) is a subcategory of \({\mathcal L}_0\). The authors define a class of categories lying between \(\mathcal F\) and \({\mathcal L}_0\) and call it \(\mathcal L\)-convenient. They also introduce several categorical properties to guarantee that a subcategory of an \(\mathcal L\)-convenient category is naturally isomorphic to \(\mathcal F\) whenever it has those properties. Namely, they prove the following result. \textbf{Theorem.} Suppose that \(\mathcal C\) is an \(\mathcal L\)-convenient category and \(\mathcal E\) is its full subcategory consisting of semi-simple objects which are pointwise complete real lattices and their sublattices of bounded elements are normal, separating and ideal-complete. Then \(\mathcal E\) is naturally isomorphic to \(\mathcal F\).