Trading linearity for ellipticity: a nonsmooth approach to Einstein's theory of gravity and the Lorentzian splitting theorems

Basic Information

• Paper ID: 2501.00702
• Title: Trading linearity for ellipticity: a nonsmooth approach to Einstein's theory of gravity and the Lorentzian splitting theorems
• Author: Robert McCann (University of Toronto)
• Classification: math-ph math.AP math.DG math.MG math.MP
• Publication Date: January 3, 2025
• Paper Link: https://arxiv.org/abs/2501.00702

Abstract

Although Einstein's theory of gravity is established in a smooth setting, the singularity theorems of Hawking and Penrose describe how this smoothness must necessarily break down in many physical situations. In the positive definite signature case, there exists a highly successful theory of metric and metric-measure geometry that takes Riemannian manifolds as special cases but allows extraction of nonsmooth limits under dimension and curvature bounds analogous to relativistic energy conditions. This lecture highlights recent progress in developing analogous theory under Lorentzian signature, with the ultimate goal of providing a nonsmooth theory of gravity. In particular, by trading the linearity of the d'Alembertian for ellipticity, low-regularity splitting theorems are obtained, utilizing negatively homogeneous p-d'Alembert operators for this purpose.

Research Background and Motivation

Core Problem

This research addresses a fundamental contradiction in Einstein's theory of gravity: although general relativity is founded on smooth Lorentzian manifolds, the Hawking-Penrose singularity theorem demonstrates that in many physical situations (such as black holes and the Big Bang), this smoothness must necessarily break down.

Problem Significance

1. Theoretical Completeness: A nonsmooth theory of gravity capable of handling singularities is needed
2. Mathematical Unification: A successful nonsmooth theory exists in the positive definite metric case (Ricci curvature reformulated through entropy convexity of probability measures along geodesics), but the Lorentzian case lacks corresponding theory
3. Physical Realism: Black holes and cosmological singularities are fundamental predictions of general relativity, requiring mathematical tools for rigorous treatment

Limitations of Existing Approaches

Traditional proofs of Lorentzian splitting theorems face three fundamental difficulties:

1. Absence of Ellipticity: The d'Alembertian (□²) is not an elliptic operator
2. Failure of Maximum Principle: Cannot deduce function equality from ordering relations
3. Failure of Bochner Identity: The left-hand side is no longer non-negative definite in the Lorentzian setting

Core Contributions

1. Innovative Nonsmooth Lorentzian Geometric Method: Obtaining ellipticity by trading the linearity of the d'Alembertian
2. Theory of Negatively Homogeneous p-d'Alembert Operators: Establishing non-uniform ellipticity in the range p < 1
3. Unified Proof of Splitting Theorem: Incorporating the Lorentzian splitting theorem into a framework closer to the Cheeger-Gromoll Riemannian splitting theorem
4. Simplified Proofs of Classical Results: Providing simplified proofs for the confirmation of Yau's (1982) conjecture by Eschenburg (1988), Galloway (1989), and Newman (1990)

Methodology Details

Problem Formulation

Investigation of geometric splitting properties of Lorentzian manifolds containing timelike lines under the strong energy condition (SEC), particularly under timelike geodesic completeness or global hyperbolicity conditions.

Core Technical Framework

p-d'Alembert Operator

Define the negatively homogeneous p-d'Alembert operator:

□ₚu := -∇ · (|∇u|ₚ₋₂_F ∇u) = -δE/δu

where p < 1, u is a future-pointing function, and the operator arises from the variational derivative of the energy functional:

E(u) = ∫_M H(du)dvol_g

where H(w) = -1/p |w|ᵖ_F* is the Hamiltonian.

Ellipticity Recovery Mechanism

Ellipticity is achieved through:

1. Convexity Foundation: Convexity of the Hamiltonian H when p < 1 (McCann 31, Mondino-Suhr 33)
2. Non-uniform Ellipticity: The operator □ₚ is non-uniformly elliptic on the set of future-pointing functions
3. Uniform Ellipticity: Achieved locally through equi-Lipschitz estimates

Technical Innovations

1. p-Harmonicity of Busemann Functions

For a timelike geodesic γ : ℝ → Mⁿ, define the Busemann function:

b±_r(x) := ∓ℓ(x,γ(r)) ± ℓ(γ(0),γ(r))
b± := lim_{r→±∞} b±_r

Key properties:

• |∇b_r|_F = 1 = |∇b±|_F
• Satisfy reverse triangle inequality ordering
• Become p-harmonic functions when p < 1

2. Comparison Theorem

Theorem 4 (Nonsmooth p-d'Alembert Comparison): For p < 1, the operator □ₚ is non-uniformly elliptic on future-pointing functions, and (SEC) implies:

□ₚb⁺_r ≤ (n-1)/ℓ(·,γ(r))  (in the distributional sense)

3. Uniform Ellipticity

Theorem 5 (Equi-Lipschitz Estimates): Under conditions (a) and/or (b), there exist a neighborhood X and constants R, C such that for r ≥ R:

• Maximizing geodesics connect each point in X to γ(r)
• {b⁺_r} is timelike and uniformly equi-Lipschitz on X

Main Results

New Proof of Lorentzian Splitting Theorem

Theorem 3: Let (Mⁿ,g_) be a connected smooth spacetime satisfying the strong energy condition and containing a timelike line. If M is timelike geodesically complete, then M is a geometric product of ℝ with a (Ricci non-negative, complete) Riemannian submanifold Σⁿ⁻¹.

Proof Strategy

1. p-Superharmonicity: ±b± are p-superharmonic in the distributional sense: □ₚb⁺ ≤ 0 ≤ □ₚb⁻
2. Strong Maximum Principle: Improving b⁺ ≥ b⁻ to b⁺ = b⁻ ∈ C^{1,1}(X)
3. Modified Bochner Identity: Deformation with homogeneity 2p-2 < 0:

Tr[(√D²H∇²b√D²H)²] + Ric(DH,DH) = 0

4. Geometric Splitting: Deducing orthogonal splitting of the metric from Hess b = 0

Technical Details

Equi-Semiconcavity Control

Lemma 6: For some constant C̃, all u ∈ {b⁺r}{r≥R} satisfy:

lim_{t→0} [u(exp^g̃_x tv) + u(exp^g̃_x -tv) - 2u(x)]/g̃(v,v) ≤ C̃

This ensures that ∇b⁺_r → ∇b⁺ converges almost everywhere, allowing comparison results to extend to r = ∞.

Linearization Analysis

The linearization of the operator in non-divergence form involves the Hessian:

H^{ij} := ∂²H/∂w^i∂w^j = |w|^{p-2}[(2-p)g^{ik}g^{jl}w_kw_l/|w|² - g^{ij}]

which becomes positive definite for p < 1 and w along the time axis in normal coordinates.

Related Work

Development of Riemannian Splitting Theorems

• Cheeger-Gromoll (1971): Splitting theorem for complete Ricci non-negative Riemannian manifolds
• Gigli (2021): Nonsmooth version in infinitesimal Hilbert metric-measure spaces
• Sturm, Lott-Villani: Curvature-dimension conditions CD(0,N) defined through entropy displacement convexity

History of Lorentzian Splitting Theorems

• Yau (1982): Proposed the Lorentzian splitting conjecture
• Eschenburg (1988): Proved under global hyperbolicity + completeness
• Galloway (1989): Requires only global hyperbolicity
• Newman (1990): Requires only timelike geodesic completeness

Nonsmooth Lorentzian Geometry

• Cavalletti-Mondino (2024): Timelike curvature-dimension condition TCD(0,N) in Lorentzian length spaces
• Beran et al. (2023): Splitting theorems under non-negative timelike curvature

Conclusions and Discussion

Main Conclusions

This paper successfully establishes a nonsmooth theory framework in Lorentzian geometry. Through the innovative p-d'Alembert operator method, the proof of the splitting theorem is unified into a framework closer to the Riemannian case.

Theoretical Significance

1. Methodological Breakthrough: First time obtaining ellipticity in the Lorentzian setting by trading linearity
2. Unification: Incorporating splitting theorems under different conditions into a unified framework
3. Simplification: Providing a more direct proof path than traditional methods

Limitations

1. Smoothness Requirements: Current results still require gᵢⱼ ∈ C^∞(Mⁿ), though subsequent work will extend to gᵢⱼ ∉ C²(Mⁿ)
2. Locality: Uniform ellipticity holds only locally, requiring additional arguments for globality
3. Technical Complexity: Treatment in the range p < 1 involves complex variational analysis

Future Directions

1. Low-Regularity Extension: Extending the theory to metrics below C² regularity
2. Penrose-Type Theorems: Establishing analogues of Penrose singularity theorems in nonsmooth settings
3. Numerical Applications: Developing corresponding numerical methods for practical physical problems

In-Depth Evaluation

Strengths

1. Strong Originality: The p-d'Alembert operator method is an original contribution
2. Theoretical Depth: Transforms complex geometric analysis problems into elliptic theory problems
3. Unified Framework: First time incorporating Lorentzian and Riemannian splitting theorems into similar proof strategies
4. Mathematical Rigor: Rigorous mathematical arguments addressing technical obstacles insurmountable by traditional methods

Weaknesses

1. Readability: High technical threshold for non-specialists
2. Applicability: Currently primarily theoretical contributions; practical applications remain to be developed
3. Completeness: Some technical details (such as globalization arguments) are not fully expanded in this paper

Impact

1. Disciplinary Impact: May open new research directions in nonsmooth Lorentzian geometry
2. Theoretical Value: Provides new tools for the mathematical foundations of general relativity
3. Methodological Value: The idea of trading linearity for ellipticity may be applicable to other fields

Application Scenarios

1. Theoretical Physics: Black hole geometry, cosmological singularity research
2. Mathematical Physics: Mathematical description of nonsmooth spacetimes
3. Geometric Analysis: PDE theory on Lorentzian manifolds

References

The paper cites 44 important references spanning from Einstein field equations to modern metric geometry theory, particularly the pioneering work of Cheeger-Gromoll, Hawking-Penrose, Yau and others, as well as recent advances by Gigli, Cavalletti-Mondino and others in nonsmooth geometry.