Weak Cosmic Censorship

Weak cosmic censorship (WCC) is the claim that gravitational collapse does not generically expose singular behavior to distant observers. Its natural setting is an asymptotically flat initial-value problem, and its cleanest formulations use either the completeness of future null infinity $\mathscr I^+$ or future asymptotic predictability. “Black holes hide singularities” is an intuition, not a complete conjecture, until the data class, topology, matter model, asymptotic end, and meaning of singularity are named.

Required background. The Einstein Equation supplies the field equations and their constraint content. Helpful background. Pseudo-Riemannian Metrics and Signature supplies Lorentzian geometry; Affine and Variational Geodesics supplies completeness language; Causality and Global Structure supplies causal and conformal boundaries; Formulations of General Relativity separates constraints, evolution, and gauge; Exact Solutions of Einstein’s Equations supplies comparison geometries; PDEs and Green Functions for Gravitation supplies well-posedness and data topology.

Definition: the asymptotically flat conjecture

Choose an Einstein–matter system and a class $\mathcal D_{\mathrm{AF}}$ of complete, regular, constraint-satisfying asymptotically flat initial data. Give this constraint set a topology $\mathcal T$, normally a weighted Sobolev, weighted Hölder, or bounded-variation topology appropriate to the local evolution theorem. For each datum $d\in\mathcal D_{\mathrm{AF}}$, let $\operatorname{MGHD}(d)$ be its maximal globally hyperbolic development.

A schematic completeness formulation is:

There is a specified generic subset $\mathcal G\subset\mathcal D_{\mathrm{AF}}$ such that, for every $d\in\mathcal G$, every designated component of future null infinity in $\operatorname{MGHD}(d)$ is future complete.

Completeness here concerns the future extent of null infinity, not just the physical affine length of individual outgoing light rays that reach it. In the standard sufficiently regular vacuum conformal setting, choose a Bondi conformal gauge in which the null generators of $\mathscr I^+$ are affinely parametrized by retarded time $u$. Future completeness requires those generators to continue to arbitrarily large $u$ Wald 1984, §11.1, pp. 279–282. An exact matter or low-regularity formulation may instead use an equivalent physical normalization and limiting criterion, which must be supplied by its theorem.

The distinction is easy to visualize: each outgoing ray at an allowed $u$ can travel an infinite physical affine distance to infinity even if the constructed spacetime supplies only a finite interval of retarded times. Infinite length toward $\mathscr I^+$ does not prove completeness along $\mathscr I^+$. The conjecture also needs an existence clause: the relevant infinity must exist with the stated asymptotic regularity. This combination of existence and completeness is explicit in Wald 1997, §2.

A future-asymptotic-predictability formulation instead asks that the entire exterior region relevant to $\mathscr I^+$ lie in the future domain of dependence of the initial surface, expressed in the conformal completion. A visibility formulation excludes singular ideal endpoints from the causal past of $\mathscr I^+$. Under strong hypotheses these capture the same intended exterior regularity, but they are not definitions that can be substituted without proof. Completeness, causal visibility, and regularity of the conformal boundary can fail in different ways.

For a chosen end, the black-hole region is

$$\mathcal B=M\setminus J^-(\mathscr I^+).$$

The slogan says singular behavior should lie in $\mathcal B$. The mathematical statement is usually phrased positively through $\mathscr I^+$ because a “singular set” is not an ordinary subset of the smooth manifold. The alternatives are compared on What Counts as a Naked Singularity?.

What “generic” must specify

The quantifier is not cosmetic. Exact negative-mass or overextremal solutions with visible singularities do not by themselves refute WCC, because the conjecture concerns developments of admissible regular data and deliberately allows exceptional outcomes. Conversely, one stable family inside a restricted topology does not settle every WCC formulation.

At minimum, a WCC statement fixes:

• whether $\mathcal G$ is open dense, residual, the complement of a finite-codimension set, prevalent, or of probability one under a stated measure;
• whether the topology controls finitely many derivatives, smooth convergence, weighted decay, bounded variation, or localized Hölder behavior;
• whether the constraint equations have already been solved and diffeomorphism redundancy has been quotiented or gauge-fixed;
• whether spherical, axial, or other symmetry is imposed; and
• whether the system is vacuum, Einstein–scalar, Einstein–Maxwell, a fluid model, or another hyperbolic matter coupling.

These notions are not equivalent. A residual set can have empty interior. “Codimension two” in a constructed affine family does not automatically mean measure zero in an infinite-dimensional space. There is no canonical translation from bounded variation to a smooth weighted Sobolev topology. The detailed function-space treatment belongs to Function-Space Formulations of Weak Cosmic Censorship; the present page fixes the global causal content.

Theorem results in a spherical matter model

The strongest classical theorem discussed here concerns the spherically symmetric Einstein equations coupled to a massless scalar field. It is a nonlinear coupled evolution, but it lives in a symmetry-reduced data space.

Christodoulou describes initial data through a reduced function $\vartheta\in BV\cap L^1(\mathbb R)$. His 1999 analysis identifies an exceptional set $E$ left when two trapped-surface criteria fail. For each $\vartheta\in E$, he constructs a two-dimensional affine plane $\Pi_\vartheta$ such that

$$(\Pi_\vartheta\setminus\{\vartheta\})\cap E=\varnothing,$$

and planes attached to distinct exceptional points do not meet. This is Theorem 4.1 in Christodoulou 1999, pp. 214–216. The result justifies his phrase “positive codimension” in that precise geometric sense. It shows instability of the exceptional naked behavior under particular bounded-variation perturbations that trigger trapped-surface formation.

The theorem is major support for a spherical Einstein–scalar WCC formulation in the $BV\cap L^1$ framework. It does not prove an open-dense theorem for unrestricted vacuum data, does not cover nonspherical gravitational degrees of freedom, and does not turn “positive codimension” into a canonical probability statement. Its hypotheses and perturbation directions are part of the conclusion, not technical decoration.

Localized Hölder stability

Evidence available through 2026-08-12 demonstrates why the topology cannot be suppressed. Zheng’s 2026 preprint studies Christodoulou’s continuously self-similar spherical Einstein–scalar naked singularities. Their characteristic data are smooth away from one distinguished sphere but have localized $C^{1,k^2/(1-k^2)}$ behavior there.

Let

$$p_k=\frac{1}{1-k^2}.$$

For sufficiently small nonzero $k$, initial scalar perturbations in a sufficiently small ball of $\overline C_N^{p_k,\delta}$, with $N\ge6$ and $0<\delta\le\delta_0$, have nonlinear MGHDs that are asymptotically flat, contain no trapped surface, and have incomplete future null infinity Zheng 2026, Theorem 3.1, pp. 23–25. The global norm combines weighted localized Hölder control at the singular sphere with $C^N$ control away from it and asymptotically flat decay; its precise definition is given in Zheng 2026, Definitions 2.7 and 2.11, pp. 18–20.

This is an open stability result at the background’s threshold regularity, within spherical symmetry and a characteristic data formulation. It is nonlinear, not merely a linear-mode calculation. It neither treats unrestricted nonspherical perturbations nor smooth Cauchy data. Zheng also notes an intermediate class, with tangential derivatives in $C^{0,k^2/(1-k^2)}\cap BV$ without the same localization, for which the WCC question remains open in that model Zheng 2026, §1, pp. 2–3.

There is no logical contradiction with Christodoulou’s theorem. The bounded-variation perturbations producing trapped surfaces may be rougher than the self-similar background at the distinguished sphere, while Zheng perturbs within the background’s localized Hölder threshold. The two results answer different topological formulations. They collectively rule out the sentence “spherical scalar collapse has a topology-independent censorship verdict.”

Other characteristic-data theorems

A 2024 v1 preprint claims a censorship theorem for the spherically symmetric Einstein–Maxwell–charged-complex-scalar system. For characteristic profiles in $BV([0,\infty),dv)\cap L^1((0,\infty),dv/v)$, data outside an exceptional set have complete $\mathscr I^+$; through each exceptional datum passes a complex two-dimensional transverse family containing no other exceptional datum An–Tan 2024, Theorem 1.1 and Remark 1, pp. 2–3, and Theorem 9.1, p. 136. This is an unrefereed, symmetry-reduced theorem claim, not an unrestricted measure statement.

Existence results occupy different cells from stability. Shlapentokh-Rothman glues a companion vacuum interior to the earlier exterior construction, producing a maximal development from a complete asymptotically flat characteristic cone with incomplete $\mathscr I^+$ Shlapentokh-Rothman 2022, Theorem 1.3, pp. 5–6. The initial data have arbitrarily high angular regularity but are only $C^{1,s}$ in the null direction Shlapentokh-Rothman 2022, Remark 1.2, pp. 5–6; this is neither smooth Cauchy formation nor stability.

An and Wu construct nonspherical Einstein–scalar solutions from structured $\kappa$-self-similar incoming data and small outgoing null perturbations. Their spacetimes have incomplete $\mathscr I^+$ and a singular inner Cauchy horizon An–Wu 2026, Theorem 1.3, pp. 3–4. The theorem removes spherical symmetry for a designed double-null family, but it does not establish an open neighborhood or genericity in an unrestricted data topology.

Vacuum, matter, symmetry, and nonlinearity

WCC is often voiced as a conjecture about generic vacuum or reasonable-matter gravitational collapse. Each qualifier changes the phase space.

In vacuum, gravitational radiation supplies the dynamical degrees of freedom, and singularity formation from large general data remains a difficult unrestricted problem. With a scalar field, spherical symmetry still leaves a radiative matter degree of freedom and yields a tractable $1+1$ system. Electromagnetic charge can support inner horizons and changes the admissible conserved quantities. Fluids may form shocks, so “breakdown” of a matter variable must be distinguished from a spacetime singularity. Energy conditions constrain focusing but do not select a data topology or prove visibility.

A result open among spherical data need not be open in the unrestricted constraint set: the spherical subspace has no nonspherical tensor perturbations and is typically infinite-codimensional in the full space. Similarly, linear instability of a fixed naked solution is evidence about a tangent equation, while WCC is a nonlinear statement about an open neighborhood or another named generic set. Christodoulou 1999 and Zheng 2026 are both nonlinear but symmetry reduced. An–Wu 2026 is nonlinear and nonspherical, yet its structured existence construction still does not supply that open-neighborhood quantifier.

Beyond asymptotic flatness

The familiar WCC formulation does not transfer unchanged across cosmological constant or boundary conditions.

de Sitter asymptotics. Future infinity is spacelike, and cosmological horizons divide observers. A formulation must select a component or observer class and replace completeness of asymptotically flat null infinity with an appropriate conformal or causal condition. The 2024 Kerr–Newman–de Sitter spectral work concerns regularity at an inner Cauchy horizon, not visibility of a collapse singularity from asymptotically flat $\mathscr I^+$; it is evidence for a strong-censorship model, not a WCC theorem.

Anti-de Sitter asymptotics. The conformal boundary is timelike. Boundary conditions are part of the evolution, and signals can reach and return from infinity. A claim of boundary visibility must name reflecting, dissipative, or other conditions. A theorem with one boundary condition need not survive another.

Thus WCC should normally be read as an asymptotically flat conjecture unless a different receiver and evolution problem are written explicitly.

Limitations and counterexample discipline

Three checks prevent false verdicts.

First, a visible singular exact metric is not necessarily produced by complete regular data. Second, a naked outcome from one datum does not refute a genericity claim. Third, instability or stability in a symmetry-reduced norm does not establish the unrestricted theory.

WCC also does not say that black-hole interiors are deterministic. An exterior can possess complete $\mathscr I^+$ while its MGHD admits a continuous extension across an internal Cauchy horizon. That possibility is the domain of Strong Cosmic Censorship. Isenberg states explicitly that neither weak nor strong censorship implies the other Isenberg 2015, §3, pp. 6–7.

Formulation diagnostic

Problem. Evaluate the claim: “Weak cosmic censorship is false because a small open ball of initial data produces naked singularities.” You are told only that the ball lies in a localized Hölder topology on spherically symmetric characteristic data for the Einstein–scalar system, and that the resulting $\mathscr I^+$ is incomplete.

Solution

The claim is valid only for the corresponding spherical Einstein–scalar WCC formulation in that localized Hölder data space, assuming the data are complete and asymptotically flat as stated. It is a nonlinear open-set counterexample to a generic-completeness claim in that topology. It says nothing by itself about unrestricted nonspherical data, vacuum evolution, smooth weighted Sobolev data, de Sitter or AdS boundaries, or other meanings of “generic.” A defensible conclusion names all of those axes rather than declaring WCC universally false.

The topology and data-space vocabulary is developed next on Genericity, Data Classes, and Matter Models, after the parallel regularity-dependent definition of strong censorship.

References

• An, Xinliang, and Hong Kiat Tan. “A Proof of Weak Cosmic Censorship Conjecture for the Spherically Symmetric Einstein–Maxwell–Charged Scalar Field System.” arXiv:2402.16250v1 (2024). arXiv:2402.16250v1.
• An, Xinliang, and Shengrong Wu. “Naked Singularities beyond Spherical Symmetry: Singular Inner Cauchy Horizons for the Einstein–Scalar Field System.” arXiv:2607.07134v1 (2026). arXiv:2607.07134v1.
• Christodoulou, Demetrios. “The Instability of Naked Singularities in the Gravitational Collapse of a Scalar Field.” Annals of Mathematics 149 (1999): 183–217. arXiv:math/9901147.
• Isenberg, James. “On Strong Cosmic Censorship.” Surveys in Differential Geometry 20 (2015): 17–36. arXiv:1505.06390.
• Shlapentokh-Rothman, Yakov. “Naked Singularities for the Einstein Vacuum Equations: The Interior Solution.” arXiv:2204.09891v1 (2022). arXiv:2204.09891v1.
• Wald, Robert M. 1984. General Relativity. University of Chicago Press, Chicago.
• Wald, Robert M. 1997. “Gravitational Collapse and Cosmic Censorship.” arXiv:gr-qc/9710068v3, §2. Open PDF.
• Zheng, Weihao. “Nonlinear Stability of Continuously Self-Similar Naked Singularities for the Einstein-Scalar Field Equations I: Main Results.” Preprint, 15 May 2026. arXiv:2605.16235.