A Sieve That Demands a Giant Fulcrum ✛ ✎
Nothing extraneous slips through an infinitely fine sieve. That's how a Q-space works: every set here is an intersection of countably many open regions. But can such a sieve be whole, with no isolated point-holes and a guarantee that the intersection of dense parts remains dense? This property, in the categorical sense, is called Baire completeness.
As it turns out, the sieve's wholeness hinges on the existence of a monstrously large number — a measurable cardinal. This isn't just any number; it’s a special point that can be assigned an exotic measure: it gives every subset either 0 or 1, and it does so nontrivially — the measure doesn't get 'stuck' on individual points. Such a cardinal was first described by John von Neumann. Using forcing (forcibly expanding the mathematical universe), the authors took a model with a measurable cardinal and 'tossed in' random numbers via methods from stochastic algebra — like a numerical simulation of a new probabilistic world. In the resulting construction, an ideal transforms the cardinal into a Q-space, and one with strictly positive measure on open sets, which wards off topological 'scrawniness'.
Thus, the optimal 'all-sifting' sieve requires us to postulate a measurable cardinal. Without one, these spaces don't exist. It's a reminder of how the branching of mathematical possibilities depends on seemingly distant axioms. Even the information-theoretic complexity of the space — how many open sets are needed to describe it — is bounded by the continuum, which imposes hard constraints. And the probabilistic logic of large cardinals sifts infinity ever more finely through the sieve of topology.
Most previously known Q-spaces were 'thin' — they can be broken into meager parts. But those built from a measurable cardinal are, by contrast, 'thick' and robust.