Bourgain's condition, sticky Kakeya, and new examples

Basic Information

• Paper ID: 2511.10918
• Title: Bourgain's condition, sticky Kakeya, and new examples
• Author: Arian Nadjimzadah (UCLA)
• Classification: math.CA (Classical Analysis and ODEs)
• Submission Date: November 14, 2024 to arXiv
• Paper Link: https://arxiv.org/abs/2511.10918

Abstract

This paper proves for all dimensions n≥3 that for any Hörmander-type oscillatory integral operator satisfying Bourgain's condition, the sticky case of the corresponding curved Kakeya conjecture can be reduced to the sticky case of the classical Kakeya conjecture. This supports the conjecture of Guo-Wang-Zhang that an operator satisfies the same $L^p$ bounds as in the restriction conjecture if and only if it satisfies Bourgain's condition.

The research is based on a new geometric characterization of Bourgain's condition, which is established on the structure of curved $\delta$-tubes within $\delta^{1/2}$-tubes. The author constructs examples showing that this property does not hold in larger tubes, particularly demonstrating that in each dimension n≥3, there exist operators satisfying Bourgain's condition for which no diffeomorphism maps the corresponding curve families to straight lines. This indicates that general-to-sticky reductions in the Wang-Zahl style require substantially new ideas.

Research Background and Motivation

Core Problem

This paper studies $L^p$ estimates for Hörmander-type oscillatory integral operators. Given a phase function $\phi: M \times \Sigma \to \mathbb{R}$, the corresponding oscillatory integral operator is defined as: $T_\phi^\lambda f(x) = \int_{\mathbb{R}^{n-1}} e^{i\lambda\phi(x/\lambda, \xi)}f(\xi)a(x/\lambda, \xi)d\xi$

The central problem is to determine the range of $p$ for which the following estimate holds: $\|T_\phi^\lambda f\|_p \leq C_{p,\epsilon}\lambda^\epsilon\|f\|_\infty$

Importance of the Problem

1. Unifying Classical Conjectures: This problem unifies several important conjectures:
   • When $\phi = \phi_{n,rest}(x,\xi) = x\cdot\xi + \frac{1}{2}t|\xi|^2$, it yields the Fourier restriction problem
   • When $\phi = \phi_{n,BR}$, it yields the Bochner-Riesz conjecture
   • The conjectured range is $p > \frac{2n}{n-1}$
2. Connection between Geometry and Analysis: This problem deeply links the geometric properties of Kakeya sets with the analytic properties of oscillatory integrals
3. Dimensional Differences: Hörmander proved that all phase functions satisfy the estimate when n=2, but Bourgain showed that general phase functions fail the estimate when n=3

Limitations of Existing Methods

1. Understanding Bourgain's Condition: Although Guo-Wang-Zhang introduced Bourgain's condition as the boundary between good and bad phase functions, its geometric meaning remains unclear
2. Curved Kakeya Conjecture: For phase functions satisfying Bourgain's condition, the corresponding curved Kakeya conjecture ($\phi$-Kakeya sets have Hausdorff dimension n) remains unsolved
3. Sticky Case: Wang-Zahl solved the classical sticky Kakeya conjecture in R³, but the reduction for the curved case is unclear

Research Motivation

The author aims to:

1. Provide a transparent geometric characterization of Bourgain's condition
2. Reduce the curved sticky Kakeya problem to the classical case
3. Explore the possibilities and obstacles for general-to-sticky reductions

Core Contributions

1. Geometric Characterization of Bourgain's Condition (Theorem 1.8): Proves that a phase function $\phi$ satisfies Bourgain's condition if and only if for each $(\delta^{1/2}, \phi)$-tube $T_0$, the family of $(\delta, \phi)$-tubes it contains is diffeomorphic to a family of straight $\delta$-tubes while preserving directions (with error $O(|(\xi,v)-(\xi_0,v_0)|^2)$)
2. Reduction of Sticky Case (Theorem 1.19): For phase functions satisfying Bourgain's condition, reduces the sticky curved Kakeya conjecture in dimension n to the classical sticky Kakeya conjecture in dimension n. Combined with Wang-Zahl's result, this proves the sticky curved Kakeya conjecture for n=3
3. Mild Reformulation of Bourgain's Condition (Proposition 1.10): Provides an equivalent form of Bourgain's condition: $\nabla_\xi^2\phi(x,\xi) = A(\nabla_\xi\phi(x,\xi), \xi) + c(x,\xi)B(\nabla_\xi\phi(x,\xi), \xi)$ where B is non-degenerate and $(G\cdot\nabla_x)c \neq 0$
4. Sharp Counterexamples (Examples 1.20 and Proposition 1.21): Constructs the "tan-example," the simplest example of a phase function satisfying Bourgain's condition whose curve family cannot be diffeomorphic to a family of straight lines. This shows:
   • The error in Theorem 1.8 cannot be improved to $O(|(\xi,v)-(\xi_0,v_0)|^4)$
   • No diffeomorphism to straight lines exists at scales between $\delta$ and $\delta^{1/4}$
   • General-to-sticky reductions require new ideas

Detailed Methods

Task Definition

$\phi$-curves and $\phi$-tubes:

• $\phi$-curves are defined as: $\ell_{\xi,v} = \{x \in M: \nabla_\xi\phi(x,\xi) = v\}$
• A $(\phi, \delta)$-tube $T_{\xi,v}^\delta$ is the $\delta$-neighborhood of $\ell_{\xi,v}$
• The parameter $\xi \in \Sigma$ is called the direction, and $(v,\xi) \in V \times \Sigma$ parametrizes the curve family

$\phi$-Kakeya Sets: A compact set $K \subset M_0$ is a $\phi$-Kakeya set if for each direction $\xi \in \Sigma_0$, there exists $v \in V_0$ such that $\ell_{\xi,v} \cap M_0 \subset K$

Sticky $\phi$-Kakeya Sets: A $\phi$-Kakeya set K is sticky if there exists a collection of $\phi$-curves $\mathcal{L} \subset \mathcal{C}(\phi)$ with packing dimension n-1 containing at least one curve from each direction, satisfying $\ell \cap M_0 \subset K$ for all $\ell \in \mathcal{L}$

Core Technical Route

1. Reformulation of Bourgain's Condition (Proposition 1.10)

Forward Implication (Bourgain's condition → reformulated form):

• Choose local coordinates so that the Gauss map $G(x,\xi_0) = e_n$
• In the new coordinates, $\partial_{x_n}\nabla_\xi\phi(x,\xi_0) = 0$
• Bourgain's condition becomes: $\partial_{x_n}^2\nabla_\xi^2\phi = \lambda\partial_{x_n}\nabla_\xi^2\phi$
• Solving this differential equation yields the form (1.19)

Reverse Implication (reformulated form → Bourgain's condition):

• Apply the operator $G\cdot\nabla_x$ to (1.19): $(G\cdot\nabla_x)\nabla_\xi^2\phi = (G\cdot\nabla_x)c \cdot B$
• Applying again yields: $(G\cdot\nabla_x)^2\nabla_\xi^2\phi = \frac{(G\cdot\nabla_x)^2c}{(G\cdot\nabla_x)c}(G\cdot\nabla_x)\nabla_\xi^2\phi$

2. Proof of Geometric Characterization (Theorem 1.8)

Forward Direction (Bourgain's condition → geometric property):

Set up parametrization: Choose coordinates so that $\phi$-curves are transverse to t-slices, parametrized as: $\ell_{\xi,v} = \{(X(\xi,v,t), t): |t| \leq 1\}$

Using the data A, B, c from Proposition 1.10, define:

• $\Xi(\xi) = B(v_0,\xi_0)(\xi-\xi_0)$
• $\mathcal{V}(\xi,v) = (v-v_0) - A(v_0,\xi_0)(\xi-\xi_0)$

Construct three diffeomorphisms:

Step 1 (Recentering): $F_1(x,t) = (x-X(\xi_0,v_0,t), t)$

Step 2 (Twist in each slice): $F_2(x,t) = (\nabla_vX(\xi_0,v_0,t)^{-1}x, t)$

Step 3 (Diffeomorphism of t-component): $F_3(x,t) = (x, \tilde{c}(t))$, where $\tilde{c}(t) = c(X(\xi_0,v_0,t), t, \xi_0)$

Finally obtain: $F_3 \circ F_2 \circ F_1(\ell_{\xi,v}) \subset \text{line}_{\Xi(\xi), \mathcal{V}(\xi,v)} + O(|(\xi,v)-(\xi_0,v_0)|^2)$

Reverse Direction (geometric property → Bourgain's condition):

Assuming a diffeomorphism F exists satisfying (1.14), through Taylor expansion and coefficient matching, derive: $\nabla_\xi^2\phi = A(\nabla_\xi\phi, \xi) + c(x,\xi)B(\nabla_\xi\phi, \xi)$

3. Proof of Sticky Reduction (Theorem 1.19)

Employ a scale-inductive strategy:

Base Case: By extracting individual tubes, for $\delta < \delta_0(\epsilon)$ we have $SK'(\phi, \epsilon, 1, \delta^{\epsilon/(10n)})$

Inductive Step (Lemma 5.3): Prove $SK'(\phi, \epsilon, \eta, \delta^{1/2}) \land SK'(\phi_{n,rest}, \eta/10, \tilde{\eta}, \delta^{1/2}) \Rightarrow SK'(\phi, \epsilon, \tilde{\eta}/10, \delta)$

Key ideas:

1. Cover the $\delta$-tube collection $\mathcal{T}$ with $\rho = \delta^{1/2}$-tubes $\mathcal{T}_\rho$
2. For each $T_\rho \in \mathcal{T}_\rho$, apply the diffeomorphism F from Theorem 1.8
3. Apply radial dilation $h(x,t) = (\rho^{-1}x, t)$ to obtain straight tubes at standard scale
4. Apply the classical sticky Kakeya hypothesis $SK'(\phi_{n,rest}, \eta/10, \tilde{\eta}, \delta^{1/2})$
5. Reverse transform and apply $SK'(\phi, \epsilon, \eta, \delta^{1/2})$

By iterating $N = O(\log(1/\epsilon))$ times, reduce from scale $\delta^{\epsilon/(10n)}$ to scale $\delta$

Technical Innovations

1. Geometric Transparency: The characterization in Theorem 1.8 is manifestly invariant under diffeomorphisms in x and $\xi$, explaining the invariance discovered by Guo-Wang-Zhang
2. Multi-scale Structure: Reveals the local linearization property of Bourgain's condition at the $\delta^{1/2}$-tube scale
3. Constructive Method: Proposition 1.10 provides a systematic method for constructing phase functions satisfying Bourgain's condition (by choosing A, B, c)
4. Sharpness: The tan-example shows Theorem 1.8 is optimal in the sense that:
   • The error cannot be improved to $O(|(\xi,v)-(\xi_0,v_0)|^4)$
   • The property fails at larger scales (e.g., $\delta^{1/4}$-tubes)

Construction of the tan-example

Construction Strategy

Using Proposition 1.10, choose a simple ansatz:

Phase Function Form: $\phi(x,\xi) = \sum_{j=1}^{n-1} \int f_j(x_j, t, \xi_j)d\xi_j$

Choice of A, B, c: $A(v,\xi) = \begin{pmatrix} 0 & 0 \\ 0 & v_{n-1}^2 \end{pmatrix}, \quad B(v,\xi) = I_{n-1}, \quad c(x,\xi) = t^2$

This leads to a decoupled ODE system:

• $\partial_{\xi_j}f_j = t^2$ ($1 \leq j \leq n-2$)
• $\partial_{\xi_{n-1}}f_{n-1} = f_{n-1}^2 + t^2$

Solutions:

• $f_j = t^2\xi_j + x_j$ ($1 \leq j \leq n-2$)
• $f_{n-1} = t\tan(t\xi_{n-1} + x_{n-1})$

Obtaining the phase function: $\phi_{n,tan}(x,\xi) = x' \cdot \xi' + \frac{1}{2}t^2|\xi'|^2 + \log(\sec(t\xi_{n-1} + x_{n-1}))$

Curve Family

$\ell_{\xi,v} = \{(v' - t^2\xi', \tan^{-1}(\frac{v_{n-1}}{t}) - t\xi_{n-1}, t): |t-1| \leq 1/10\}$

Proof of Non-linearizability (Proposition 1.21)

Core Idea: Prove that the family of curves passing through $\ell_{0,0}$ and a point $p \notin \ell_{0,0}$ is not contained in a surface, whereas a family of straight lines must be contained in a plane.

Technical Steps:

1. Simplify the curve family to third-order error
2. Solve for curve parameters $\xi_p(s), v_p(s)$ passing through $\ell_{0,0}(s)$ and $p$
3. Compute the tangent vector $\gamma(s)$ at point $p$ and its derivatives
4. Prove $|\gamma(1) \wedge \dot{\gamma}(1) \wedge \ddot{\gamma}(1)| \neq 0$

Explicit calculation shows: $|\gamma(1) \wedge \dot{\gamma}(1) \wedge \ddot{\gamma}(1)| = \frac{2|p_{n-1}|^3|p_{n-2}|}{3t_0^2(t_0-1)^6} + O(|p|^5) > 0$

This contradicts the assumed linearization.

Experimental Setup

This is a pure mathematical theory paper with no numerical experiments. All results are rigorous mathematical proofs.

Experimental Results

Verification of Main Theorems

Theorem 1.8 (Geometric Characterization):

• Complete proof of both forward and reverse directions
• Forward direction uses the structure from Proposition 1.10 through explicit diffeomorphism construction
• Reverse direction derives the form of Proposition 1.10 through Taylor expansion and coefficient matching

Theorem 1.19 (Sticky Reduction):

• Establishes the formal reduction: $(\forall\epsilon>0\exists\eta,\delta_0>0: SK(\phi_{n,rest},\epsilon,\eta,\delta_0)) \Rightarrow (\forall\epsilon>0\exists\eta,\delta_0>0: SK(\phi,\epsilon,\eta,\delta_0))$
• Combined with Wang-Zahl's Theorem 1.17 (n=3 classical sticky Kakeya), yields the corollary: For n=3, sticky curved Kakeya sets for phase functions satisfying Bourgain's condition have Hausdorff dimension 3

Proposition 1.21 (Non-linearizability of tan-example):

• Rigorously proves no linearization exists with error $O(|(\xi,v)|^4)$
• Shows the $O(|(\xi,v)|^2)$ error in Theorem 1.8 is nearly optimal

Theoretical Significance

1. Supporting Hörmander's Dichotomy Conjecture (Conjecture 1.3): By establishing connections in the sticky case, provides strong evidence for "T_\phi^\lambda satisfies (1.4) if and only if $\phi$ satisfies Bourgain's condition"
2. Partial Resolution of Curved Kakeya Conjecture: Completely resolves the n=3 sticky case
3. Revealing Obstacles: The tan-example shows that general-to-sticky reductions (like Wang-Zahl) require substantially new ideas, since curves cannot be organized into surfaces as effectively as straight lines

Related Work

Historical Background

1. Hörmander (1973): Proved all Hörmander-type phase functions satisfy the estimate when n=2
2. Bourgain (1991):
   • Proved general phase functions fail the estimate when n=3
   • Constructed the counterexample $\phi_{worst}$
3. Guo-Wang-Zhang (2024):
   • Introduced Bourgain's condition as the boundary between good and bad phase functions
   • Generalized Bourgain's counterexample to all n≥3
   • Proved $p > 2 + \frac{2.5921}{n} + O(n^{-2})$ when Bourgain's condition is satisfied
   • Proved $\phi$-tubes satisfy strong polynomial Wolff axioms

Progress on Kakeya Problems

1. Wolff (1995): Classical Kakeya sets have dimension ≥5/2 (n=3)
2. Wang-Zahl (2022, 2025a, 2025b):
   • Resolved sticky Kakeya conjecture in R³
   • Resolved Kakeya conjecture in R³
   • Developed general-to-sticky reductions
3. Wang-Wu (2024): Developed techniques from Kakeya-type estimates to restriction-type estimates

Related Geometric Research

1. DGGZ (2024): Carleson-Sjölin operators on manifolds; proved Bourgain's condition implies curve families are diffeomorphic to straight lines in translation-invariant and manifold settings
2. GLX (2025): Curved Kakeya and Nikodym problems on manifolds
3. Nadjimzadah (2025): Positive results for curved Kakeya using "coniness"

Positioning of This Paper

This paper advances the field by:

1. First providing a transparent geometric characterization of Bourgain's condition
2. First establishing reduction from curved sticky Kakeya to classical sticky Kakeya
3. First constructing examples satisfying Bourgain's condition that cannot be globally linearized
4. Pointing out obstacles and possible directions for general case research

Conclusions and Discussion

Main Conclusions

1. Geometric Essence of Bourgain's Condition: At the $\delta^{1/2}$-tube scale, satisfying Bourgain's condition is equivalent to local linearizability of $\delta$-tube families while preserving directions
2. Complete Reduction of Sticky Case: For phase functions satisfying Bourgain's condition, the n-dimensional sticky curved Kakeya problem is equivalent to the n-dimensional classical sticky Kakeya problem
3. Result for Dimension 3: Combined with Wang-Zahl's work, the sticky curved Kakeya conjecture holds in R³
4. Obstacles for General Case: The tan-example shows general-to-sticky reductions require new ideas beyond existing techniques

Limitations

1. Scale Restriction: Theorem 1.8 holds only at the $\delta^{1/2}$-tube scale, failing at larger scales (e.g., $\delta^{1/4}$-tubes)
2. Sticky Assumption: Theorem 1.19 only handles the sticky case; the general case remains open
3. Dimensional Restriction: Although the theory applies to all n≥3, concrete applications depend on the classical sticky Kakeya conjecture, currently fully resolved only for n=3
4. Hausdorff Dimension vs Maximal Functions: Only considers Hausdorff dimension; stronger maximal function estimates remain unresolved

Future Directions

The author proposes the following research directions:

1. General-to-Sticky Reduction: Attempt general-to-sticky reduction for the curved case in dimension 3, possibly requiring polynomial Wolff axioms rather than convex Wolff axioms
2. Study of Semi-linearized Families: Even when $\phi$ doesn't satisfy Bourgain's condition, sticky problems may reduce to semi-linearized families: $\ell_{\xi,v} = \{(v + A(t)\xi, t): |t| \leq 1\}, \quad A(t) = tI_{n-1} + O(t^2)$
3. Finer Classification: Explore whether the structure of semi-linearized families provides finer classification of Hörmander oscillatory integral problems
4. Global Information: Seek more global information about Bourgain's condition based on Proposition 1.10
5. Genericity Research: Study the genericity of tan-example type examples

In-Depth Evaluation

Strengths

1. Deep Geometric Insight: Theorem 1.8 transforms the algebraic condition (1.9) into a clear geometric statement, revealing the essence of Bourgain's condition
2. Technical Innovation:
   • Proposition 1.10 provides a constructive equivalent characterization
   • Multi-scale inductive argument elegantly exploits the structure of $\delta^{1/2}$-tubes
   • tan-example construction is concise and powerful
3. Important Results:
   • First resolution of n=3 sticky curved Kakeya conjecture
   • Provides strong support for Hörmander's dichotomy conjecture
   • Clearly identifies obstacles for general case
4. Clear Writing:
   • Sufficient motivation exposition
   • Clear technical roadmap
   • Well-placed examples (such as cartoon diagrams in Figures 1.1, 1.2)
5. Completeness: Forms a complete theoretical picture from positive results (Theorems 1.8, 1.19) to counterexamples (Proposition 1.21)

Weaknesses

1. Scale Gap: Theorem 1.8 holds between scales $\delta$ and $\delta^{1/2}$, but fails between $\delta$ and $\delta^{1/4}$. Whether an optimal scale exists between these is unexplored
2. Error Estimates: While proving the error cannot be improved to $O(|(\xi,v)|^4)$, the possibility of $O(|(\xi,v)|^3)$ is not discussed
3. General Case: The most important general-to-sticky reduction remains unresolved, though obstacles are identified
4. Computational Details: Some calculations in the tan-example (such as equation 6.49) are quite technical and might benefit from more geometric intuition
5. Application Scope: Does not discuss direct implications for specific problems (such as the Bochner-Riesz conjecture)

Impact

1. Theoretical Contribution:
   • Provides new perspective on core problems in harmonic analysis
   • Connects geometry (Kakeya sets) and analysis (oscillatory integrals)
   • Advances Hörmander's dichotomy conjecture
2. Methodological Value:
   • Proposition 1.10's constructive method may apply to other problems
   • Multi-scale inductive strategy has universal applicability
   • tan-example construction technique is generalizable
3. Follow-up Research:
   • Indicates directions for general-to-sticky reductions
   • tan-example provides benchmark for testing new methods
   • Framework for studying semi-linearized families
4. Reproducibility: All proofs are complete and rigorous; results are verifiable

Applicable Scenarios

1. Theoretical Research:
   • Study of $L^p$ estimates for Hörmander-type oscillatory integrals
   • Research on curved Kakeya and Nikodym problems
   • Development of new geometric measure theory techniques
2. Specific Problems:
   • Fourier restriction conjecture
   • Bochner-Riesz conjecture
   • Carleson-Sjölin operators on manifolds
3. Method Applications:
   • Problems requiring local linearization
   • Multi-scale geometric analysis
   • Tube geometry research
4. Counterexample Construction: Techniques from tan-example can be used to construct other phase functions with specific properties

Key References

1. Bourgain (1991): First discovered counterexamples for n=3
2. Guo-Wang-Zhang (2024): Introduced Bourgain's condition and Hörmander's dichotomy conjecture
3. Wang-Zahl (2022, 2025a, 2025b): Resolved classical sticky and general Kakeya conjectures (n=3)
4. Hörmander (1973): Foundational results for n=2 case
5. DGGZ (2024): Related work on manifolds
6. Wang-Wu (2024): Transformation techniques from Kakeya to restriction estimates

Overall Assessment: This is a high-quality theoretical mathematics paper achieving substantial progress on core problems in harmonic analysis. By revealing the geometric essence of Bourgain's condition and establishing complete reduction of the sticky case, it provides strong support for Hörmander's dichotomy conjecture. The tan-example construction both demonstrates the sharpness of the theory and points to obstacles and directions for future research. The paper features deep techniques, rigorous arguments, and clear exposition, and will have significant impact on the field.