Genericity, Data Classes, and Matter Models

“Generic” is meaningful only after the space of admissible initial data and its topology have been fixed. In cosmic censorship, this choice decides which perturbations count as small, which outcomes are stable, and whether a symmetry-reduced counterexample says anything about unrestricted gravity. Open, dense, residual, finite-codimension, prevalent, and probability-one claims are different mathematical statements; none is the unqualified meaning of genericity.

Required background. The Einstein Equation supplies the field equations and their constraint content. Helpful background. Pseudo-Riemannian Metrics and Signature supplies metric regularity; Affine and Variational Geodesics supplies completeness; Causality and Global Structure supplies visibility and Cauchy horizons; Formulations of General Relativity separates constraints, evolution, and gauge; Exact Solutions of Einstein’s Equations supplies model families; PDEs and Green Functions for Gravitation supplies norms, well-posedness, and stability.

The initial-data space

For vacuum gravity on a three-manifold $\Sigma$, a candidate datum is a pair $(h_{ij},K_{ij})$, but admissible data lie on the nonlinear constraint set. With

$$K_{ij}=-h_i{}^a h_j{}^b\nabla_a n_b =-\frac12\mathcal L_n h_{ij},$$

the vacuum constraints are

$${}^{(3)}R+K^2-K_{ij}K^{ij}=2\Lambda, \qquad D_jK^j{}_i-D_iK=0.$$

Matter adds field data and constraint equations. Maxwell theory adds electric and magnetic Gauss constraints; a scalar field adds its position and normal derivative; fluids add an equation of state and positivity restrictions. The data space is therefore not a common vector space shared by all models.

Diffeomorphisms of $\Sigma$ act on the data without changing the physical development. One can form a quotient, work in a local slice, or impose a gauge, but a topology on raw coordinate components can mistake gauge motion for physical perturbation. A precise genericity theorem says how this redundancy is handled.

Topology and asymptotics are also inseparable. Compact Cauchy data might be measured in $H^s$ or $C^k$. Asymptotically flat data use weighted norms controlling decay and a chosen number of derivatives. Asymptotically de Sitter data are naturally compared at a spacelike conformal boundary. Anti-de Sitter evolution additionally fixes timelike-boundary conditions. Changing a weight can admit slower tails, exclude conserved charges, or alter the continuity of the evolution map.

Topological meanings of genericity

Let $\mathcal D$ be a topological data space and $\mathcal P\subset\mathcal D$ the data whose developments have the property under study.

Open. If $\mathcal P$ is open, each datum in it has a neighborhood with the same property. This is a stability statement. It does not say that most of $\mathcal D$ has the property.

Dense. If $\mathcal P$ is dense, every datum can be approximated by data in $\mathcal P$. Density alone permits $\mathcal P$ to be fragile and to have empty interior.

Open dense. This combines local stability with approximation from every part of the data space. Many model SCC theorems aim for this form.

Residual. A residual set contains a countable intersection of open dense sets. In a Baire space its complement is meagre, so residual is a powerful topological notion of typicality. A residual set can nevertheless have empty interior and need not have probability one for a chosen measure.

Finite codimension. A smooth submanifold of codimension $m$ requires $m$ independent conditions locally. In nonlinear or nonsmooth infinite-dimensional settings, an author may instead construct transverse finite-dimensional families meeting an exceptional set sparsely. That precise construction must be quoted; the word codimension alone does not supply a manifold theorem.

Prevalent. Prevalence is a measure-like notion often defined in Banach spaces by declaring a complement shy under finite-dimensional probe measures and translations. A gravitational constraint set or its diffeomorphism quotient is not globally a canonical Banach vector space, so the charts, probes, and invariance properties must be stated.

Probability one. A probabilistic claim requires a probability measure or random-data ensemble. Infinite-dimensional data spaces have no canonical analogue of finite-dimensional Lebesgue measure. “Positive probability” is undefined until a distribution, covariance, cutoff, and limiting procedure are supplied.

Open dense, residual, prevalent, and probability-one sets can disagree. A censorship conjecture should select the notion needed for its physical and mathematical purpose instead of moving among them after a counterexample appears.

Regularity changes which perturbations are small

A topology determines both amplitude and roughness. A perturbation can be small in a weak norm while introducing high-frequency or low-regularity structure that is large in a stronger norm. Conversely, a ball in a strong localized Hölder norm may forbid precisely the rough directions that trigger an instability.

The local Cauchy theorem imposes a lower regularity threshold: below it, a unique development may not be available. The censorship property imposes another threshold: for strong censorship it specifies allowed extensions, while for weak censorship it may govern the regularity of null infinity or singular data. These thresholds should not be conflated.

Smooth genericity has several possible meanings. The Fréchet $C^\infty$ topology is not the same as choosing one large finite $C^k$ or Sobolev index. A set open in $C^1\times C^0$ and dense in $C^1$ is a mixed-topology statement, not simply “open dense in smooth data.” Ringström’s general $T^3$ Gowdy theorem gives exactly this distinction: the set $\mathcal G_{i,c}$ is open in a $C^1\times C^0$ topology and dense in the $C^1$ topology, and its MGHDs are $C^2$-inextendible Ringström 2009, Theorem 2 and Corollary 1, pp. 1190–1191. The earlier polarized $T^3$ theorem instead obtains smooth inextendibility on a subset open and dense in the $C^\infty$ topology of its smaller symmetry class Chruściel, Isenberg, and Moncrief 1990, theorem statement, p. 1673.

Symmetry is a restriction of the data space

Spherical symmetry reduces Einstein–matter evolution to a $1+1$ system and eliminates gravitational-wave degrees of freedom. Gowdy symmetry retains wave dynamics but imposes two commuting spacelike Killing fields. Stationary Kerr–Newman–de Sitter metrics form a finite-dimensional parameter family. None is an open set of unrestricted initial data.

Consequently:

• open among spherical data does not mean open after nonspherical perturbations are admitted;
• residual in a Gowdy constraint set does not mean residual in the unrestricted vacuum constraint set;
• a transition surface in the three-parameter KNdS family is not a codimension-one surface in the full space of collapse data; and
• “generic linear modes are present” on a fixed background is not nonlinear genericity of the background itself.

Symmetry-reduced theorems remain valuable: they isolate mechanisms and supply exact statements. Their conclusions must retain the symmetry label.

Matter models define different phase spaces

Vacuum, scalar, electromagnetic, and fluid models cannot be pooled into a single evidence count. Their constraint sets, conserved quantities, characteristic speeds, and continuation criteria differ.

A massless scalar field in spherical symmetry retains a nonlinear focusing mechanism and has a well-developed characteristic problem. Electromagnetic charge supports Reissner–Nordström-type inner horizons. Fluids can form shocks while the spacetime metric remains regular, so fluid breakdown and geometric singularity require separate continuation criteria. Charged or massive scalar perturbations have different spectra from neutral massless ones. An energy condition restricts stress energy but does not establish hyperbolic well-posedness or select “reasonable matter” uniquely.

For SCC, matter regularity also enters the extension. A scalar in $H^1_{\mathrm{loc}}$ has locally integrable energy, whereas a discontinuous fluid or non-$L^2$ connection may not define the desired weak Einstein–matter solution. For WCC, matter influences trapped-surface formation and asymptotic decay. Every formulation should name the system rather than append “with reasonable matter” as an undefined clause.

Comparing theorems across data classes

The following examples are deliberately not collapsed into one verdict. Evidence is current through 2026-08-12.

Result	Data and dynamics	Genericity statement	What it does not establish
Christodoulou 1999	nonlinear spherical Einstein–massless-scalar data represented in $BV\cap L^1$	through each exceptional datum, a two-dimensional affine plane meets the exceptional set only at that datum	open-dense or unrestricted WCC
Gowdy SCC results	nonlinear vacuum, compact symmetry-reduced data	polarized theorem: open dense in the $C^\infty$ topology, with smooth inextendibility; general $T^3$ theorem: open in $C^1\times C^0$ and dense in $C^1$, with $C^2$-inextendibility	unrestricted vacuum SCC
Davey–Dias–Sola Gil 2024	linear neutral massless scalar on fixed stationary KNdS backgrounds	spectral-mode criterion over a finite-dimensional black-hole parameter scan	nonlinear Einstein–Maxwell genericity or arbitrary collapse
Zheng 2026	nonlinear spherical Einstein–scalar characteristic data near finite-regularity self-similar backgrounds	a small open ball in a localized Hölder topology produces incomplete $\mathscr I^+$	nonspherical or arbitrary smooth-data WCC

Christodoulou’s exact conclusion is not merely “unstable”: for the exceptional set $E\subset BV\cap L^1$, Theorem 4.1 constructs planes $\Pi_\vartheta$ with $(\Pi_\vartheta\setminus\{\vartheta\})\cap E=\varnothing$ Christodoulou 1999, pp. 214–216. It does not identify $E$ as a smooth submanifold or assign it a probability.

Zheng’s theorem uses $p_k=(1-k^2)^{-1}$ and a global localized Hölder space $\overline C_N^{p_k,\delta}$ with $N\ge6$ and $0<\delta\le\delta_0$. For sufficiently small nonzero $k$, a norm-small ball around the self-similar characteristic data generates asymptotically flat, trap-free spacetimes with incomplete $\mathscr I^+$ Zheng 2026, Theorem 3.1, pp. 23–25. The space controls threshold Hölder behavior at one sphere and asymptotic decay Zheng 2026, Definitions 2.7 and 2.11, pp. 18–20. These perturbations are not the same directions as arbitrary $BV$ perturbations.

The KNdS scan similarly illustrates the distinction between parameter and data genericity. It determines which quasinormal-mode decay rates cross the $\beta=1/2$ finite-energy threshold for a linear scalar on a fixed background Davey, Dias, and Sola Gil 2024, §§2.3.3 and 5, pp. 13–14 and 34–40. It does not put a measure or topology on nonlinear collapse data.

How to formulate a stable claim

A complete censorship claim can be recorded as a tuple

$$(\text{equations},\ \Sigma,\ \text{asymptotics},\ \mathcal D,\ \mathcal T, \ \text{genericity},\ \text{symmetry},\ \text{outcome},\ \text{regularity}).$$

For WCC, the outcome might be completeness of a named component of asymptotically flat $\mathscr I^+$. For SCC, it might be $C^2$-inextendibility or inextendibility as a weak solution with locally square-integrable connection. This tuple prevents a theorem from migrating silently between vacuum and matter, spherical and unrestricted, linear and nonlinear, or asymptotically flat and de Sitter/AdS settings.

Formulation diagnostic

Problem. Claim A says that naked outcomes form an open ball in a localized Hölder topology on spherical Einstein–scalar characteristic data. Claim B says that, in a $BV\cap L^1$ representation of spherical Einstein–scalar data, an exceptional set has the two-dimensional-plane property above. Do the claims contradict? Does either settle unrestricted smooth WCC?

Solution

They do not contradict. The topologies admit different perturbation directions and measure roughness differently; an open ball in one topology can be unstable under perturbations allowed by another. Claim B is a precise positive-codimension construction, not automatically an open-dense or measure-one theorem. Both statements impose spherical symmetry and a massless scalar field. Neither includes nonspherical gravitational-wave perturbations or supplies a topology on unrestricted smooth constraint data, so neither settles unrestricted smooth WCC.

The next page uses this tuple to compare formulations without manufacturing implications: Logical Relations and Open Formulation Choices.

References

• 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.
• 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.