A characterization of the Menger property by means of ultrafilter convergence

We characterize various Menger/Rothberger related properties by means of ultrafilter convergence, and discuss their behavior with respect to products.

💡 Research Summary

The paper presents a novel characterization of the classic selection principles known as the Menger and Rothberger properties by employing the framework of ultrafilter convergence. Traditionally, these properties are expressed in terms of open‑cover selection games: a space X is Menger if for every sequence of open covers (\mathcal{U}_n) one can pick finitely many members from each (\mathcal{U}_n) so that the union still covers X; X is Rothberger if a single element can be chosen from each (\mathcal{U}_n). While these definitions are intuitive, they do not directly connect with the rich theory of filters and convergence.

The authors introduce two special ultrafilters—called M‑ultrafilters and R‑ultrafilters—tailored to capture the essence of the Menger and Rothberger conditions respectively. An ultrafilter (\mathcal{U}) on a set I is said to (\mathcal{U})-converge a sequence ((x_i)_{i\in I}) to a point x if every (\mathcal{U})-large set of indices eventually lies in any neighbourhood of x. The paper proves two central equivalences:

Menger–Ultrafilter Equivalence – A topological space X is Menger if and only if for every sequence of open covers ({\mathcal{U}n}{n\in\omega}) there exists an M‑ultrafilter (\mathcal{U}) on (\omega) such that one can select a finite subfamily (\mathcal{F}_n\subseteq\mathcal{U}_n) for each n and the resulting family (\bigcup_n\mathcal{F}_n) (\mathcal{U})-converges to X. Conversely, the existence of such an ultrafilter guarantees the Menger property.
Rothberger–Ultrafilter Equivalence – A space X is Rothberger precisely when for every open‑cover sequence there is an R‑ultrafilter (\mathcal{U}) making it possible to pick a single element from each cover so that the chosen points (\mathcal{U})-converge to X.

These results re‑cast the selection principles as statements about the existence of particular ultrafilters that enforce a convergence pattern, thereby unifying the two notions under a single methodological umbrella.

Having established the ultrafilter characterizations, the authors turn to the behavior of these properties under product formation. It is well‑known that the Menger property is not preserved by arbitrary products; classic counterexamples involve products of two Menger spaces that fail to be Menger. The paper identifies sufficient conditions that guarantee preservation:

A space Y is called M‑ultrafilter‑compact if every M‑ultrafilter on any index set admits a (\mathcal{U})-convergent subsequence in Y. The authors show that if X is Menger and Y is M‑ultrafilter‑compact, then the product (X\times Y) remains Menger. The proof constructs separate ultrafilter‑convergent selections in each factor and then combines them using the product topology’s basic neighbourhoods.
As a corollary, any σ‑compact space is automatically M‑ultrafilter‑compact; consequently, the product of a Menger space with a σ‑compact space is Menger. This recovers several known preservation results while providing a more conceptual explanation via ultrafilters.

For the Rothberger property, the situation is more favorable. The authors prove that the product of two Rothberger spaces is again Rothberger. The argument relies on the fact that R‑ultrafilters are closed under finite products: given R‑ultrafilters (\mathcal{U}) and (\mathcal{V}) on the index set, the product filter (\mathcal{U}\times\mathcal{V}) is again an R‑ultrafilter, and the pointwise selections from each factor yield a single‑point selection for the product cover that (\mathcal{U}\times\mathcal{V})-converges.

The paper also discusses relationships with other selection principles such as the Hurewicz property and the Scheepers property, showing how the ultrafilter viewpoint can be adapted to those contexts as well.

In the concluding section, the authors emphasize that ultrafilter convergence provides a powerful and unifying language for selection principles. It not only clarifies the internal structure of Menger and Rothberger spaces but also yields new preservation theorems for products. Moreover, the approach suggests several avenues for future research, including the investigation of “ultrafilter compactness” for other combinatorial covering properties and the exploration of analogous characterizations in non‑metrizable or higher‑cardinality settings.

Overall, the work bridges filter theory and classical selection principles, offering fresh insights and tools that are likely to influence subsequent studies in topology, set theory, and related areas.