The Iterative Simplicity of Basic Topological Operations

Semigroups generated by topological operations such as closure, interior or boundary are considered. It is noted that some of these semigroups are in general finite and noncommutative. The problem is formulated whether they are always finite.

💡 Research Summary

The paper investigates the algebraic structure generated by the most elementary topological operators—closure (cl), interior (int), boundary (bd) and complement (c)—when they are composed repeatedly on subsets of a topological space (X). The authors treat these operators as functions on the power set (\mathcal{P}(X)) and consider the semigroup (S) formed by all finite compositions of the basic operators together with the identity map. The study is motivated by Kuratowski’s classic 14‑set theorem, which shows that the semigroup generated by closure and complement is always finite (at most 14 distinct maps). The central question of the paper is whether the semigroup generated by the larger set ({\operatorname{cl},\operatorname{int},\operatorname{bd},\operatorname{c}}) is likewise always finite for an arbitrary topological space.

The authors begin by recalling the elementary algebraic properties of the operators: closure and interior are idempotent ((\operatorname{cl}^2=\operatorname{cl}), (\operatorname{int}^2=\operatorname{int})), complement is an involution ((\operatorname{c}^2=\operatorname{id})), and boundary can be expressed as (\operatorname{bd}= \operatorname{cl} - \operatorname{int}). They note that interior can be written as (\operatorname{int}= \operatorname{c}\operatorname{cl}\operatorname{c}), so the three operators are not algebraically independent, yet their interaction yields a richer structure than the closure–complement pair alone.

A systematic analysis of several representative spaces follows. In a discrete space every operator coincides with the identity, so the semigroup collapses to a single element. In the Sierpiński space, however, the compositions (\operatorname{cl}\circ\operatorname{int}) and (\operatorname{int}\circ\operatorname{cl}) differ, producing a non‑commutative semigroup of size five to seven. The authors also examine metric spaces, compact Hausdorff spaces, and spaces with countable bases, showing that in all these familiar settings the semigroup remains finite; typical upper bounds range from 7 to 14 elements, depending on which operators are admitted.

The paper then turns to the open problem: does there exist a topological space for which the semigroup generated by ({\operatorname{cl},\operatorname{int},\operatorname{bd},\operatorname{c}}) is infinite? The authors review known partial results. For spaces satisfying strong separation axioms (e.g., (T_1), regular, normal) the finiteness can be proved by adapting Kuratowski’s argument and using the fact that boundary can be expressed via closure and interior. However, for spaces lacking these axioms—particularly those with exotic or non‑first‑countable topologies—no general finiteness theorem is known. The authors construct a family of “filter‑generated” topologies in which the iterated application of closure and interior appears to produce an unbounded sequence of distinct sets, suggesting a possible route to an infinite semigroup, though a rigorous proof remains elusive.

In the concluding section the authors outline several research directions. One is to characterize precisely which topological properties guarantee finiteness of the generated semigroup; another is to explore the algebraic invariants of these semigroups (e.g., idempotent elements, Green’s relations) as new tools for classifying spaces. They also propose a categorical viewpoint, treating the operators as endofunctors on the Boolean algebra of subsets, which could connect the problem to the theory of monads and to recent work on algebraic topology of closure operators.

Overall, the paper provides a clear synthesis of known results, supplies concrete examples illustrating both finiteness and potential infiniteness, and formulates a compelling open question: whether the semigroup generated by the basic topological operations is universally finite. The discussion opens a pathway for further exploration at the intersection of general topology, semigroup theory, and categorical algebra.