Independence of the separable quotient problem

The explanation

Start in ordinary three-dimensional space. Take the map that sends (x, y, z) to (x, y). It flattens space onto a plane. Every point on the z-axis goes to (0, 0). So the plane is a quotient of space by that axis. Its dimension is 3 − 1 = 2. A quotient is what remains after you collapse a subspace to zero.

A Banach space is a vector space with a length for each vector, where limits of good sequences always exist. It may be infinite-dimensional. It is separable if a countable set of vectors comes arbitrarily close to every vector. Subspaces are easy. Every infinite-dimensional Banach space contains a separable infinite-dimensional subspace. Take the closed span of a countable list of independent vectors. Quotients are the hard direction. The separable quotient problem asks whether every infinite-dimensional Banach space has a separable infinite-dimensional quotient. It is traditionally attributed to Banach and MazurSource 1. Brech identifies Rosenthal's 1969 paper as an early written statement of itSource 2.

Many cases were known. Johnson and Rosenthal showed every separable one has an infinite-dimensional quotient with a basisSource 3. Argyros, Dodos and Kanellopoulos proved the positive answer for every dual spaceSource 4. Saxon and Sánchez Ruiz proved it for every space of density below a set-theoretic number called 𝔟Source 5. Density is the size of the smallest dense set. Dodos, Lopez-Abad and Todorcevic gave a consistent positive answer for very large densities, assuming large cardinalsSource 6. But no one had a counterexample, or a proof for all spaces.

The paper claims the problem is independent of ZFC, the standard axioms of mathematicsSource 7. Independent means ZFC can neither prove nor refute it. This rests on two implications. First, if the continuum hypothesis holds, the statement failsSource 8. The continuum hypothesis, CH, says no set is strictly between the integers and the real numbers in size. Second, if the size of the continuum is real-valued measurable, the statement holdsSource 9. That means there is a measure on every set of real numbers with a strong additivity property. It is a large-cardinal assumption. The negative side needs only the consistency of ZFC itselfSource 10. For spaces of density ℵ₁, the smallest uncountable size, the paper claims independence relative to ZFC aloneSource 11.

The key is one criterion. A separable infinite-dimensional quotient exists exactly when some infinite-dimensional subspace of functionals has a separable evaluation imageSource 12. A functional is a bounded linear map from the space to the scalars. For the counterexample, the paper builds a dual space with coordinates indexed by ω₁, the first uncountable ordinal. It forms paths of functionals that carry integer weights. Distinct paths share only finitely many weightsSource 13. On every infinite-dimensional subspace, this gives uncountably many path functionals that stay a fixed distance apartSource 14. CH is then used to defeat every candidate quotient.

The obstacle on the positive side is to use a measure on all subsets of the continuum. The paper turns the measure into an unconditional basic sequence of functionals, a well-behaved infinite list. A result of Hagler and Johnson then gives the quotient.

How sure is this? Lean, a computer proof checker, proves the negative direction under CH, over both real and complex scalars. The positive direction and the full independence statement are outside that check. No specialist review has been published.

What it doesn't show. This does not decide the problem outright, and the positive half needs the consistency of a measurable cardinal.