Pith. sign in

REVIEW 3 major objections 4 minor 14 references

Commutators of Hilbert transforms along monomial curves

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read This paper proves that commutators of Hilbert transforms along monomial curves are bounded on L^p whenever the symbol belongs to the corresponding non-isotropic BMO space, and gives a partial converse through a new testing BMO space.

desk verdict Genuinely new commutator bounds for parabolic Hilbert transforms; upper bound clean, lower bound has a repairable but real geometric gap. read the letter →

arxiv 1909.02118 v1 pith:RFG5ITHE submitted 2019-09-04 math.CA

classification math.CA MSC 42B2042B2542B35
keywords commutatorsHilberttransformalongcurvesparabolicBMOtestingsparsedominationMuckenhouptweightsreverseHoldermonomial
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper establishes that the commutator of a Hilbert transform taken along a curved monomial path, such as a parabola (t, $t^{2}$), is controlled by a natural parabolic version of BMO: if the symbol b has bounded mean oscillation over parabolic cubes, then [b,H_gamma] is bounded on L^p($R^{2}$) for every 1

What carries the argument

The mechanism is the pair: (i) the Cauchy integral trick, writing [b,H_gamma]f = 2 d/dz|z=0 $e^{{zb/2}}$H_gamma($e^{{-zb/2}}$f) and then applying Cauchy's formula to bound the commutator by weighted operator norms of H_gamma with weight w=$e^{{Re z b}}$; and (ii) sparse domination of H_gamma by parabolic sparse forms (Cladek-Ou), through which weighted bounds are expressed in terms of mixed Muckenhoupt and reverse-Holder characteristics of w, which in turn are controlled by ||b||_{BMO_gamma} through quantitative exponentiation lemmas. The lower bound uses a geometric flow set E_Q = {x-gamma(t): x in Q, 9 ell(Q) <= t <= 10 ell(Q)} and the identity that the commutator applied to chi_{E_Q}, averaged over Q, computes exactly the testing-BMO oscillation of b; the key geometric fact is |E_Q| ~ |Q| with uniform constants. A parabolic cube is a rectangle Q=I x J with |J|=|I|^2.

What would settle it

Find a parabolic-cube weight w with [w]_{A_{2/r}} bounded by a fixed constant but [w]_{RH_{1+$\sigma$}} arbitrarily large for a fixed small $\sigma$; that would falsify the transferred reverse-Holder estimate (2.5) used in the proof. Alternatively, construct b in testing BMO for the parabola for which [b,H_gamma] fails to be bounded on $L^{2}$, which would show the lower-bound theorem cannot be reversed.

Watch

Extended reading notes

Core claim

The central claim is that the commutator [b,H_gamma] with the parabolic Hilbert transform H_gamma f(x)=p.v. integral f(x-(t,$t^{2}$)) dt/t is bounded on L^p($R^{2}$), 1<p<infinity, whenever b lies in parabolic BMO, the space defined by the supremum over parabolic cubes Q=I x J with |J|=|I|^2 of the mean oscillation of b over Q. The upper bound is obtained through the Cauchy integral trick of Coifman-Rochberg-Weiss, which reduces the commutator to weighted $L^{2}$ estimates for H_gamma; those weighted estimates come from Cladek-Ou sparse domination and the sharp weighted bounds of Bernicot-Frey-Petermichl, with the weight characteristics controlled by exponentiation lemmas transferred to the parabolic grid. The converse direction introduces a testing BMO norm measuring, for each parabolic cube Q, how much b(x) deviates from its average along the portion of the curve x-gamma(t) that flows into a shifted copy of Q; boundedness of the commutator on $L^{2}$ implies this testing norm is finite and bounded by the commutator norm. The same two-step argument works for monomial curves and torsion-free curves, and fails to give containment for lines, where curvature is absent.

Load-bearing premise

The proof of the upper bound assumes that the classical dyadic exponentiation lemma for BMO and the sharp reverse-Holder theorem for A_infinity weights transfer to the grid of parabolic cubes with only a change of constants; if that transfer fails, the weighted estimates (2.4)-(2.5) are not justified and the L^p bound does not follow.

Editorial extensions

If this is right

  • For every monomial curve eta and every 1<p<infinity, the commutator [b,H_eta] is bounded on L^p(R^n) with norm controlled by BMO_eta, and the same holds for local Hilbert transforms along torsion-free curves.
  • Higher-order commutators T^k_b satisfy ||T^k_b|| <= (C k * k!) ||b||_{BMO_gamma}, so the Cauchy-integral method degrades only by a factorial factor.
  • Any symbol whose commutator is L^2 bounded lies in testing BMO, giving the quantitative chain ||b||_{test} <= ||[b,H_gamma]|| <= ||b||_{BMO_gamma}.
  • For the line Hilbert transform, the two natural BMO-type spaces overlap but neither contains the other, so the parabolic result genuinely requires curvature.
  • The paper leaves open whether the inclusions BMO_gamma into bounded-commutator symbols into testing BMO are proper, so a full characterization is not yet available.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The testing BMO space may be strictly larger than BMO_gamma; one testable route is to search for a symbol satisfying the oscillation-along-flow condition but with unbounded parabolic mean oscillation.
  • The uniform estimate |E_Q| ~ |Q| for monomial curves in higher dimensions is delicate, and extending the lower bound to curves whose torsion vanishes at isolated points could break the testing-BMO conclusion.
  • The sparse-domination exponent range suggests that the weighted method may be sharp near the boundary of the allowed triangle; a weighted counterexample near that boundary would refine the parameter selection.
  • The same Cauchy-integral and sparse-domination route could plausibly yield two-weight or multilinear commutator estimates, since the parameter choices are independent of the symbol b.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper studies commutators [b,H_γ] of the parabolic Hilbert transform along γ(t)=(t,t^2), proving an upper bound when the symbol b lies in a parabolic BMO space (Theorem 1.1) and a lower bound in terms of a newly introduced 'testing' BMO space adapted to the curve (Theorem 1.2). The upper bound is obtained via the Coifman–Rochberg–Weiss Cauchy integral trick combined with the sparse domination theorem of Cladek and Ou and weighted estimates for H_γ. The lower bound is derived by testing the commutator on characteristic functions of parabolic cubes and their flow sets E_Q. The results are then extended to monomial curves in R^n (Theorems 4.1 and 4.2) and local torsion-free curves (Theorem 4.3), and the final section contrasts the parabolic setting with Hilbert transforms along straight lines.

Significance. If correct, the paper provides a natural BMO-type characterization for commutators of a genuinely non-Calderón–Zygmund singular integral, which is a valuable contribution to the harmonic analysis of Hilbert transforms along curves. The use of the Cauchy integral trick and Cladek–Ou sparse domination is methodologically sound, and the introduction of a testing BMO space is a useful new idea that gives a concrete necessary condition. The paper is clearly written and the main structural claims are plausible. However, two load-bearing points in the written proofs need repair: the geometric computation of the flow set E_Q in Section 3 is incorrect as stated, and the transfer of two classical weighted/BMO results to the parabolic-cube setting in Section 2 is asserted but not proved. These issues are likely fixable, but they currently prevent the central theorems from being fully established.

major comments (3)
  1. [Section 3, Eq. (3.3)] The assertion that E_Q is a rectangle is not correct, and the displayed area computation has wrong units. For Q=[0,ℓ]×[0,ℓ^2], the set E_Q={x−γ(t): x∈Q, 9ℓ≤t≤10ℓ} is a swept region whose vertical fiber over a fixed first coordinate u is an interval of length ℓ^2+(t_max^2−t_min^2), where t_min and t_max are the endpoints of the admissible t-range; the region has triangular end fibers and is not a rectangle. In particular, the claimed height 20ℓ should be of order 20ℓ^2. The comparison |E_Q|∼|Q| is nevertheless true — for ℓ=1, integrating the fiber lengths gives area 21 — so Theorem 1.2 is likely repairable, but the proof of (3.3) as written must be replaced by a correct computation.
  2. [Section 2, Eqs. (2.4)–(2.5)] The proof of Theorem 1.1 depends on two unproved assertions: that the exponentiation lemma of Bényi, Martell, Moen, Stachura, and Torres [1, Lemma 3.5] holds for parabolic cubes with the same constant 4, and that the reverse Hölder theorem of Hytönen, Pérez, and Rela [6, Theorem 2.3] 'immediately translates' to the parabolic setting. These transfers are used to obtain the uniform bounds on [w]_{A_{2/r}} and [w]_{RH_{(s'/2)'}} that feed into the weighted estimate (2.3); without them, the upper bound in Theorem 1.1 is not established. Please provide either a proof of the parabolic versions or a precise reference where they appear.
  3. [Section 2, L^p case, around Eq. (2.7)] In the passage leading to (2.7), the mixed characteristic is written as [w]_{A_{p/r}}[w]_{A_{(s'/p)'}}, but the surrounding text describes a reverse Hölder characteristic, and the L^2 estimate in (2.3) uses [w]_{A_{2/r}}[w]_{RH_{(s'/2)'}}. As written, the displayed formula is inconsistent with the argument, and the parameter selection cannot be checked. Please state the correct Cladek–Ou L^p estimate and make the notation align with the reverse Hölder characteristic used in the text.
minor comments (4)
  1. [Section 2, parameter triangle] The vertices of the acceptable parameter triangle are listed inconsistently: the L^2 case states (0,0), (1,0), (2/3,1/3), while the L^p case states (0,0), (1,1), (2/3,1/3). Please clarify which set of vertices is correct.
  2. [Section 3, Proposition 3.2] In the proof of Proposition 3.2, the assertion that one can find a small parabolic cube Q̃⊆Q with E_{Q̃}⊆R is stated without justification. Since this is the key geometric input of the proof, please provide a short argument or a reference.
  3. [Section 2, Cauchy integral trick] The parameter ǫ is introduced first as the radius of the integration contour and later redefined as 2/r−1 or p/r−1. This reuse of notation is confusing; consider using different symbols for the contour radius and the exponent gap.
  4. [References] Reference [1] is cited as an arXiv preprint; if a published version now exists, it would be helpful to update the citation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the main estimates are derived from external sparse-domination and weighted results, and the lower bound is a direct testing argument rather than a tautology.

full rationale

The paper's derivation chain is not circular. For the upper bound (Theorem 1.1), the Cauchy integral trick reduces the commutator norm to a weighted bound for H_gamma. The weighted bound (2.3) is imported from Cladek--Ou and Bernicot--Frey--Petermichl, and the Ap/reverse-Holder estimates (2.4)--(2.5) are imported from Benyi--Martell--Moen--Stachura--Torres and Hytonen--Perez--Rela; none of these are authored by the present authors, and none assume the commutator bound being proved. The BMO_gamma hypothesis is defined independently of the commutator, and the conclusion is not assumed in the hypothesis. For the lower bound (Theorem 1.2), the testing-BMO condition is defined via the flow set E_Q, but the proof derives the testing inequality from the assumed L^2 boundedness of [b,H_gamma] applied to characteristic functions, together with Cauchy--Schwarz. The geometric facts used, namely that the time interval I_{x,E_Q} has uniformly bounded Haar measure and that |E_Q| ~ |Q|, are supporting lemmas and not restatements of the conclusion. Even if the explicit rectangle computation for E_Q has a units or shape issue, that would be a correctness gap rather than circularity: the theorem is not true by definition. Proposition 3.1 likewise uses standard oscillations of BMO functions on nearby parabolic cubes and does not smuggle in the target result. There are no load-bearing self-citations, no imported uniqueness theorems, and no fitted quantity renamed as a prediction. The paper is self-contained against external benchmarks for the key analytic inputs.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The central claims rest on external theorems (sparse domination, weighted bounds, BMO weight exponentiation) that the paper cites; no free parameters are fitted to data, and no new physical or mathematical entities are postulated beyond the testing-BMO normed space, which is defined in the paper.

assumptions (3)
  • domain assumption Cladek-Ou sparse domination and weighted estimate for H_γ (Ref. [3], Corollary 1)
    Used in Section 2 as the black box that produces equation (2.3), the key weighted norm bound for the Cauchy integral trick. If this theorem fails for some curves, the proof breaks.
  • domain assumption Bényi et al. exponentiation lemma for BMO-adapted Ap weights (Ref. [1], Lemma 3.5)
    Gives [e^{λb}]_{A_{2/r}} ≤ 4 for small λ; assumed to hold for parabolic cubes even though the lemma is proved for standard dyadic cubes.
  • domain assumption Hytönen-Pérez-Rela reverse Hölder estimate (Ref. [6], Theorem 2.3)
    Gives [w]_{RH} ≤ 2 for σ small; stated for spaces of homogeneous type and assumed to apply to parabolic cubes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Commutators of Hilbert transforms along monomial curves." pith.science (2026). https://pith.science/paper/RFG5ITHE

@misc{pith2026190902118,
  author       = {Pith},
  title        = {Pith review of: Commutators of Hilbert transforms along monomial curves},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RFG5ITHE}},
  note         = {Machine review of arXiv:1909.02118}
}
abstract

The Hilbert transforms associated with monomial curves have a natural non-isotropic structure. We study the commutator of such Hilbert transforms and a symbol $b$ and prove the upper bound of this commutator when $b$ is in the corresponding non-isotropic BMO space by using the Cauchy integral trick. We also consider the lower bound of this commutator by introducing a new testing BMO space associated with the given monomial curve, which shows that the classical non-isotropic BMO space is contained in the testing BMO space. We also show that the non-zero curvature of such monomial curves are important, since when considering Hilbert transforms associated with lines, the parallel version of non-isotropic BMO space and testing BMO space have overlaps but do not have containment.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

  1. [1]

    If T is an operator which is bounded between, e.g

    Introduction Given an operator T and a function b, the commutator [ b,T ] is defined formally by the equality [b,T ]f =bT (f ) −T (bf ) for all functions in an appropriate function space. If T is an operator which is bounded between, e.g. L2 and itself, then there is a natural bound ‖[b,T ] : L2 →L2‖ ≤ 2‖b‖∞‖T :L2 →L2‖. However, this bound is generally not...

  2. [2]

    Hence forth we write Lp :=Lp(R2) to shorten notation

    Sufficient condition for boundedness: Proof of Theorem 1.1 In this section, we will show that if b ∈ BMOγ, then there is a universal constant C depending only on p such that ‖[b,H γ] : Lp(R2) →Lp(R2)‖ ≤ C‖b‖BMOγ. Hence forth we write Lp :=Lp(R2) to shorten notation. Proof. We will begin with p = 2. Since both sides of the claimed inequality scale linearly...

  3. [3]

    Recall that for a parabolic cube Q with dimensions ℓ(Q) ×ℓ(Q)2, we have defined a set EQ by EQ = {x −γ(t) : x ∈Q, 9ℓ(Q) ≤t ≤ 10ℓ(Q)}

    Necessary condition for boundedness: Proof of Theorem 1.2 We will now give a BMO-type condition forb which is necessary for theL2 bounds of the commutator [ b,H γ]. Recall that for a parabolic cube Q with dimensions ℓ(Q) ×ℓ(Q)2, we have defined a set EQ by EQ = {x −γ(t) : x ∈Q, 9ℓ(Q) ≤t ≤ 10ℓ(Q)}. For each x ∈ Q, let Ix,EQ = {t : x −γ(t) ∈ EQ}. The set EQ ...

  4. [4]

    The non-parabolic setting The techniques used in Sections 2 and 3 are not specialized to the par abolic setting; rather, they can be extended to a more general class of smooth curves. We will call a function η : R → Rn a monomial curve if η is given by η(t) = { (ǫ1|t|α1,...,ǫ n|t|αn ) if t ≥ 0 (ǫ′ 1|t|α1,...,ǫ ′ n|t|αn ) if t< 0 } where the coefficients ǫi ...

  5. [5]

    Definition 5.1

    Curves with torsion For comparison to the parabolic case, we will also consider the Hilbert transform along a line; this will show how it is important to have curvature in order to have compatibility between BMO and the commutator. Definition 5.1. The Hilbert transform along a line is defined by Hτf (x,y ) = ˆ∞ −∞ f (x −t,y )dt t . Alternatively, if f y den...

  6. [6]

    Boundedness results for commutators with BMO functions via weighted estimates: a comprehensive approach

    ´Arpad B´ enyi, Jos´ e Mar ´ ıa Martell, Kabe Moen, Eric Stachura, and Rodolfo Torres. Bounded- ness results for commutators with BMO functions via weighte d estimates: a comprehensive approach. arXiv:1710.08515, 2017

  7. [7]

    Sharp weighted norm estimates beyond Calder´ on-Zygmund theory.Anal

    Fr´ ed´ eric Bernicot, Dorothee Frey, and Stefanie Peter michl. Sharp weighted norm estimates beyond Calder´ on-Zygmund theory.Anal. PDE , 9(5):1079–1113, 2016

  8. [8]

    Sparse domination of Hilbert transforms along curves

    Laura Cladek and Yumeng Ou. Sparse domination of Hilbert transforms along curves. Math. Res. Lett., 25(2):415–436, 2018

Show all 14 references
  1. [9]

    R. R. Coifman, R. Rochberg, and Guido W eiss. Factorizati on theorems for Hardy spaces in several variables. Ann. of Math. (2) , 103(3):611–635, 1976

  2. [10]

    Fourier analysis , volume 29 of Graduate Studies in Mathematics

    Javier Duoandikoetxea. Fourier analysis , volume 29 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2001. Tran slated and revised from the 1995 Spanish original by David Cruz-Uribe

  3. [11]

    Shar p reverse H¨ older property for A∞ weights on spaces of homogeneous type

    Tuomas Hyt¨ onen, Carlos P´ erez, and Ezequiel Rela. Shar p reverse H¨ older property for A∞ weights on spaces of homogeneous type. Journal of Functional Analysis , 263, 07 2012

  4. [12]

    Hyt¨ onen

    Tuomas P. Hyt¨ onen. The Lp-to-Lq boundedness of commutators with applications to the Jacobian operator. arXiv:1804.11167, 2018

  5. [13]

    E. Laeng. On the Lp norms of the hilbert transform of a characteristic function . Journal of Functional Analysis, 262(10):4534 – 4539, 2012

  6. [14]

    E. M. Stein and S. W ainger. Problems in harmonic analysis related to curvature. Bull. Amer. Math. Soc. , 84:1239–1295, 1978. Tyler Bongers, Department of Mathematics & Statistics, W as hington University in Saint Louis, One Brookings Drive Saint Louis, MO 63130–4899 E-mail ad...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.