Cosmic Censorship and Determinism

Cosmic censorship begins only after the Cauchy problem has supplied a unique maximal globally hyperbolic development. Weak censorship asks whether singular behavior is generically hidden from a specified infinity; strong censorship asks whether that development is generically inextendible at a specified regularity. Neither conjecture has a formulation-independent verdict. This chapter builds the stable definitions, records rigorous model theorems, and uses dated results only within their actual data spaces.

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; Causality and Global Structure supplies infinity, domains of dependence, and horizons; Formulations of General Relativity separates constraints, evolution, and gauge; Exact Solutions of Einstein’s Equations supplies model extensions; and PDEs and Green Functions for Gravitation supplies well-posedness, weak solutions, and stability.

Enter this chapter

The fundamental deterministic theorem is not itself a censorship conjecture. Given admissible initial data for a locally geometrically well-posed Einstein or Einstein–matter system, there is a development into which every other globally hyperbolic development embeds, unique up to an isometry fixing the data. Choquet-Bruhat and Geroch state this as Theorem 3 and prove it by ordering developments, constructing maximal common regions, and resolving the Hausdorff gluing obstruction Choquet-Bruhat and Geroch 1969, pp. 332–334.

The result determines the solution while the initial hypersurface remains Cauchy. It does not say that the MGHD is the largest Lorentzian spacetime containing that region. A proper extension within the same solution and regularity class must lose the original Cauchy property. Arbitrary metric extensions may instead fail to satisfy the equations at the required regularity, so they require a separate analysis. Strong censorship specifies which extensions should generically be absent. Weak censorship instead asks whether a collapse singularity can influence an asymptotic receiver.

Every usable censorship statement therefore has coordinates:

$$(\text{target},\ \text{asymptotics},\ \text{equations},\ \text{data space}, \ \text{topology},\ \text{genericity},\ \text{symmetry}, \ \text{linear/nonlinear},\ \text{regularity}).$$

Suppressing any entry can make two compatible theorems look contradictory.

The chapter sequence

Read the six pages in order; each supplies a missing quantifier for the next.

1. Maximal Cauchy Development and Determinism defines admissible data, developments, data-fixing isometry, and the precise Choquet-Bruhat–Geroch maximality category. It separates geometric uniqueness from coordinate gauge and from spacetime inextendibility.
2. What Counts as a Naked Singularity? separates incomplete curves, absent ideal endpoints, curvature blowup, local visibility, and visibility from a named conformal-boundary component.
3. Weak Cosmic Censorship states the asymptotically flat completeness and predictability formulations and shows why the data topology belongs in the conjecture.
4. Strong Cosmic Censorship indexes inextendibility by $C^0$, locally Lipschitz, Christodoulou weak-solution, $C^2$, and smooth thresholds, then bounds the meaning of inner-horizon evidence.
5. Genericity, Data Classes, and Matter Models distinguishes open, dense, open dense, residual, finite-codimension, prevalent, and probabilistic notions on constraint and gauge-reduced data spaces.
6. Logical Relations and Open Formulation Choices explains why WCC and SCC have no general implication ordering, maps their formulation axes, and routes changing verdicts to dated Research assessments.

The sequence prevents the phrase “generic naked singularity” from appearing before “generic,” “naked,” and “singularity” have separate definitions.

Definitions, theorems, and evidence

Three levels of claim appear in this chapter.

Definitions. A schematic asymptotically flat WCC statement says that a specified generic subset of complete regular asymptotically flat data has future-complete $\mathscr I^+$. A schematic SCC statement says that a specified generic subset has MGHDs inextendible in an explicit regularity and weak-equation class. These templates are stable; their truth values depend on the remaining entries.

Theorems. The maximal-development theorem is established under the local existence and geometric-uniqueness hypotheses of the chosen system. Nonlinear model censorship theorems also exist. Christodoulou’s spherical Einstein–scalar result gives a precise two-dimensional transverse-plane property for the exceptional set in a $BV\cap L^1$ data representation Christodoulou 1999, Theorem 4.1, pp. 214–216. In polarized $T^3$ Gowdy vacuum spacetimes, a subset open and dense in the $C^\infty$ topology has smoothly inextendible MGHDs Chruściel, Isenberg, and Moncrief 1990, theorem statement, p. 1673. For the general $T^3$ Gowdy class, Ringström proves $C^2$-inextendibility on a set open in $C^1\times C^0$ and dense in $C^1$ Ringström 2009, Theorem 2 and Corollary 1, pp. 1190–1191.

Dated evidence. Linear spectra, finite-dimensional black-hole scans, and new preprints can sharpen or challenge a particular formulation without settling the unrestricted conjecture. Evidence on these pages is current through 2026-08-12. Later results, corrections, or withdrawals are outside this snapshot.

Weak versus strong censorship

The two conjectures test different boundaries.

Axis	Weak cosmic censorship	Strong cosmic censorship
Primary target	exterior visibility	deterministic continuation
Natural boundary	asymptotically flat $\mathscr I^+$	boundary of the MGHD, often a Cauchy horizon
Typical good property	completeness or future predictability of $\mathscr I^+$	inextendibility in class $\mathcal R$
Typical bad property	singular behavior visible from infinity	proper extension beyond the Cauchy horizon
Essential extra choice	asymptotic end and visibility definition	extension and weak-solution regularity

A complete $\mathscr I^+$ can coexist with a continuously extendible inner Cauchy horizon: WCC-type exterior behavior is good while $C^0$ SCC fails. Conversely, a globally naked singularity can be $C^0$-inextendible: WCC fails while the solution-level SCC property holds. Isenberg states the resulting logical independence in Isenberg 2015, §3, pp. 6–7.

The adjectives do not order the conjectures. Nor do singularity theorems settle either one. Geodesic incompleteness does not say whether the incomplete approach lies in $J^-(\mathscr I^+)$ or whether a metric extension exists.

The regularity ladder

For ordinary metric extensions,

$$\text{smooth}\subset C^2\subset C^{0,1}\subset C^0,$$

so, for the same extension definition,

$$C^0\text{-inextendible} \Longrightarrow C^{0,1}\text{-inextendible} \Longrightarrow C^2\text{-inextendible} \Longrightarrow C^\infty\text{-inextendible}.$$

The reverse implications fail. Curvature blowup can block $C^2$ continuation without blocking a continuous metric. A Christodoulou-type class commonly asks for enough weak first-derivative control—often locally square-integrable connection coefficients—to formulate the Einstein equations. It is not identical to $C^{0,1}$, and its precise location depends on gauge, nondegeneracy, and matter regularity.

This hierarchy changes the deterministic interpretation. A $C^0$ extension continues causal geometry. If its connection is not locally $L^2$, it may not continue the Einstein equations as a weak evolution. A $C^2$ extension continues classical curvature. The phrase “SCC holds” is incomplete unless one threshold is selected.

The formulation matrix

The main axes interact rather than merely add detail.

Asymptotics. Asymptotically flat WCC uses null $\mathscr I^+$. De Sitter future infinity is spacelike and cosmological horizons are observer dependent. AdS infinity is timelike, so boundary conditions are part of evolution. Results do not migrate across these settings unchanged.

Vacuum and matter. Vacuum gravity, Einstein–scalar, Einstein–Maxwell, and fluids have different constraints, conserved quantities, characteristic structures, and weak continuation criteria. “Reasonable matter” is not a mathematical class.

Symmetry. Spherical and Gowdy data spaces are not open subsets of unrestricted data. A stationary KNdS family is finite-dimensional. Openness within one of these spaces is not openness after missing gravitational degrees of freedom are restored.

Linear and nonlinear. A linear test field can diagnose decay versus blue shift on a fixed metric. It cannot establish nonlinear metric backreaction or genericity of collapse. A coupled nonlinear theorem can do more, but only in its stated data and symmetry class.

Topology and genericity. A ball in localized Hölder data, transverse planes in $BV\cap L^1$, an open-dense smooth theorem, and a probability-one ensemble are mutually different claims. Infinite-dimensional constraint spaces have no canonical Lebesgue measure.

Cosmic-censorship results through 2026-08-12

Two recent examples make formulation dependence unusually concrete.

Zheng’s 2026 preprint considers continuously self-similar naked singularities of the spherical Einstein–massless-scalar equations. With $p_k=(1-k^2)^{-1}$, sufficiently small nonzero $k$, $N\ge6$, and $0<\delta\le\delta_0$, a small ball in the localized Hölder space $\overline C_N^{p_k,\delta}$ generates nonlinear asymptotically flat MGHDs with no trapped surface and incomplete $\mathscr I^+$ Zheng 2026, Theorem 3.1, pp. 23–25. The norm localizes the background’s $C^{1,k^2/(1-k^2)}$ threshold at one sphere and includes asymptotic decay Zheng 2026, Definitions 2.7 and 2.11, pp. 18–20. This is nonlinear open stability in spherical characteristic finite-regularity data, not unrestricted smooth WCC.

Davey, Dias, and Sola Gil study a neutral massless scalar linearly on fixed KNdS. If

$$\beta=-\frac{\operatorname{Im}\omega}{\kappa_-},$$

then the nonsmooth Cauchy-horizon branch has locally square-integrable radial derivative exactly for $\beta>1/2$ Davey, Dias, and Sola Gil 2024, §2.3.3, pp. 13–14. Their parameter scan finds a transition surface around small near-extremal Reissner–Nordström–de Sitter, with the explicit $r_+/r_c\ge0.1$ qualification and sparser coverage extremely near extremality Davey, Dias, and Sola Gil 2024, §§5–6, pp. 34–42. This is linear de Sitter test-field evidence, not a nonlinear Einstein–Maxwell SCC theorem.

The 1999 and 2026 spherical scalar results are not contradictory: they allow different perturbation roughness. The 2024 KNdS result does not adjudicate either because it changes asymptotics, target, system, and linearity.

Common invalid inferences

Maximal means inextendible. It means maximal while the original surface remains Cauchy. A non-globally-hyperbolic extension can still exist.

Incomplete means naked. Incompleteness may be removable, hidden inside a black hole, or unrelated to the chosen infinity.

$C^2$-inextendible means $C^0$-inextendible. The implication runs the other way.

Open in spherical data means open in all data. Nonspherical gravitational-wave perturbations are absent from the spherical space.

A linear mode settles nonlinear SCC. A finite-energy test field neither constructs a metric extension nor includes backreaction.

A de Sitter inner-horizon result settles asymptotically flat WCC. The conformal boundary and target property are different.

Chapter diagnostic

Problem. A report makes four statements: (i) constraint-satisfying initial data have a unique MGHD up to data-fixing isometry; (ii) an open ball of spherical Einstein–scalar characteristic data in a localized Hölder topology produces incomplete $\mathscr I^+$; (iii) each resulting spacetime also has an internal boundary across which the metric extends in $C^0$, but no extension has locally $L^2$ Christoffel symbols and curvature diverges; (iv) a separate fixed KNdS background has only neutral scalar QNMs with $\beta>1/2$. Classify the deterministic, WCC, and SCC content. May (iv) change the verdicts about (ii)–(iii)?

Solution

(i) is the established maximal-development statement: the data determine the globally hyperbolic region, not every possible extension. (ii) is an open-set failure of the future-complete-$\mathscr I^+$ WCC formulation only in the named spherical Einstein–scalar localized Hölder space; it is not an unrestricted verdict. In (iii), $C^0$ inextendibility fails for each datum, while the stated non-$L^2$ connection and curvature blowup obstruct the stipulated Christodoulou weak-solution and $C^2$ extensions. To call those facts SCC results still requires genericity in the named data topology. Statement (iv) is a linear neutral test-field fact on a different de Sitter background. It suggests finite-energy scalar continuation there, but it changes none of (ii)–(iii): the asymptotics, equations, backgrounds, and linear/nonlinear status do not match. WCC and SCC properties in (ii)–(iii) can coexist without contradiction because they test different boundaries.

For changing evidence beyond this cutoff, use the dated Research records for Weak Cosmic Censorship and Strong Cosmic Censorship and Regularity.

References

• Choquet-Bruhat, Yvonne, and Robert Geroch. “Global Aspects of the Cauchy Problem in General Relativity.” Communications in Mathematical Physics 14 (1969): 329–335. https://doi.org/10.1007/BF01645389.
• 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.
• Chruściel, Piotr T., James Isenberg, and Vincent Moncrief. “Strong Cosmic Censorship in Polarised Gowdy Spacetimes.” Classical and Quantum Gravity 7 (1990): 1671–1680. DOI.
• Davey, Alex, Óscar J. C. Dias, and David Sola Gil. “Strong Cosmic Censorship in Kerr–Newman–de Sitter.” Journal of High Energy Physics 2024, no. 7 (2024): 113. https://doi.org/10.1007/JHEP07(2024)113; arXiv:2404.03724.
• Isenberg, James. “On Strong Cosmic Censorship.” Surveys in Differential Geometry 20 (2015): 17–36. arXiv:1505.06390.
• Ringström, Hans. “Strong Cosmic Censorship in $T^3$-Gowdy Spacetimes.” Annals of Mathematics 170, no. 3 (2009): 1181–1240. DOI.
• 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.