← All problems

Seeger–Wright weak-type (1,1) question for lacunary maximal operators

For every d≥1d\ge 1, let μ\mu be a compactly supported Borel probability measure on Rd\mathbb{R}^d satisfying polynomial Fourier decay: there exist constants β>0\beta>0 and C<∞C<\infty such that ∣μ^(ξ)∣≤C∣ξ∣−β|\widehat{\mu}(\xi)|\le C|\xi|^{-\beta} for all ∣ξ∣≥1|\xi|\ge 1. For the dyadic dilates μj\mu_j of μ\mu, define the lacunary maximal operator Mμf(x)=sup⁡j∈Z∣(f∗μj)(x)∣M_\mu f(x)=\sup_{j\in\mathbb{Z}}|(f*\mu_j)(x)|. Must there exist a constant Cμ<∞C_\mu<\infty such that ∥Mμf∥L1,∞(Rd)≤Cμ∥f∥L1(Rd)\|M_\mu f\|_{L^{1,\infty}(\mathbb{R}^d)}\le C_\mu\|f\|_{L^1(\mathbb{R}^d)} for every f∈L1(Rd)f\in L^1(\mathbb{R}^d)?

References

Progress summary

Refreshed
Open · solution reported

A specialist preprint claims a counterexample, but the result has not been independently checked.

The Seeger–Wright question asks whether lacunary maximal operators satisfy a weak-type (1,1)(1,1) bound under polynomial Fourier decay. Before the new preprint, no example or counterexample was known for singular measures in this class.

Known results

  • Seeger–Wright (2012) proved weak-type bounds from H1(Rd)H^1(\mathbb{R}^d) to L1,∞(Rd)L^{1,\infty}(\mathbb{R}^d) under dimensional and regularity assumptions.
  • They also recorded an Llog⁡log⁡LL\log\log L endpoint estimate, which does not imply weak type (1,1)(1,1).
  • Related positive H1H^1 results are known for spherical surface measure and other measures with dimensional and smoothing hypotheses.

October 2026 counterexample

Xiaojing Zhou’s specialist preprint Fourier decay and endpoint estimates for lacunary maximal operators claims a construction disproving the weak-type (1,1)(1,1) statement and two related conference problems. The claim is mathematically substantive but remains unverified by independent assessment.

Current status (as of October 2026): A preprint claims the general question has a negative answer, but the counterexample has not been independently verified.

Sources

Solutions 0

No solutions have been posted yet.