Unique predual problem for Lipschitz spaces

Determine whether, for every complete pointed metric space (M,d,0)(M,d,0), the Banach space Lip⁡0(M)\operatorname{Lip}_0(M) of real-valued Lipschitz functions vanishing at 00 has a strongly unique predual; equivalently, whether its canonical Lipschitz-free predual is strongly unique.

References

Primary source

arXiv

Progress summary

Refreshed
Claimed progress

A recent preprint overturns part of the earlier argument and proves only a special finite-dimensional case, so the general question remains open.

The problem asks whether Lipschitz function spaces have unique, or strongly unique, preduals in broad classes of metric spaces. A 2016 paper claimed the result for all Lip⁡(X)\operatorname{Lip}(X) and for Lip⁡0(X)\operatorname{Lip}_0(X) when XX is finite-dimensional in the relevant convex setting, but part of that proof was later withdrawn.

Known results

  • Lip⁡(X)\operatorname{Lip}(X) has a unique predual for every metric space XX (Weaver, 2016).
  • The earlier finite-diameter and complete-convex claims for Lip⁡0(X)\operatorname{Lip}_0(X) relied on a faulty proof of Lemma 3.13.1.
  • For complete convex XX, strong uniqueness reduces conditionally to strong uniqueness on every closed ball XnX_n.

September 1, 2026 counterexample and special-case proof

On September 1, 2026, a preprint reported an explicit counterexample to inheritance of strong unique preduals by 11-codimensional weak-star-closed subspaces and claimed a new proof for the convex finite-dimensional case. This corrects a proof technique but does not settle the problem for all complete metric spaces.

Current status (as of September 2026): uniqueness for Lip⁡(X)\operatorname{Lip}(X) is retained, while the broad strongly unique-predual claim for Lip⁡0(X)\operatorname{Lip}_0(X) remains open; the finite-dimensional convex result is claimed but unverified.

Sources

Solutions 0

No solutions have been posted yet.