When Do Black Holes Hide Their Singularities? New Study Pinpoints the Tipping Point

One of the deepest unresolved puzzles in Einstein’s general relativity has just been sharpened considerably. In a paper published in the journal General Relativity and Gravitation, Koushiki and Pankaj S. Joshi of the International Centre for Space and Cosmology at Ahmedabad University have laid out precise mathematical conditions that determine whether the singularity formed at the end of a star’s gravitational collapse remains hidden behind a black hole’s event horizon, or instead becomes visible to the outside universe as a so-called naked singularity. The work, which appeared online on 7 October 2026, addresses a question that has haunted relativists for more than half a century: does nature always conspire to cloak its most extreme objects from view?

The stakes of this question are enormous. In 1969, Roger Penrose proposed the cosmic censorship conjecture, arguing that singularities arising from gravitational collapse should always be concealed within event horizons, forever shielded from any external observer. If that conjecture holds, the predictive power of general relativity is preserved, because a naked singularity would be a region where the theory’s equations break down in plain sight, spewing causally connected unpredictability into the observable universe. If it fails, physics as we know it could lose its ability to forecast the future in certain extreme scenarios. Despite decades of effort, no general proof or disproof of cosmic censorship has ever been achieved, and the new study adds a crucial piece to this ongoing investigation.

The central insight of the new research is that the fate of a collapsing matter cloud is governed by the dynamics of the apparent horizon, a surface that marks the boundary of a region where even light cannot escape. Unlike the event horizon, which is defined globally and requires knowledge of the entire future evolution of the spacetime, the apparent horizon is a quasi-local construct: it can be located by examining the geometry of the spacetime at a given moment. Within the apparent horizon lie trapped surfaces, two-dimensional surfaces from which both outgoing and ingoing light rays converge. The authors show that whether the final singularity is visible or hidden depends critically on the relative timing of when these trapped surfaces first form and when the central singularity comes into existence.

Technically, the analysis is carried out for spherically symmetric collapse of what relativists call a type-I matter field, a broad class that includes most physically reasonable forms of matter, such as dust, perfect fluids, and scalar fields, where the energy-momentum tensor has a well-behaved algebraic structure. The authors work with initial data satisfying C-squared regularity, meaning the relevant functions are twice continuously differentiable, and the weak energy condition, which demands that matter has non-negative energy density as measured by any observer. These are the same broad physical assumptions that underpin most rigorous work on collapse in classical general relativity, giving the results considerable generality within the spherical model.

The key diagnostic tool in the study is the behavior of outgoing null geodesics, the trajectories that light rays follow as they emerge from the region near the forming singularity. If families of future-directed outgoing null geodesics can escape from arbitrarily close to the central singularity, then the singularity is locally visible: light signals generated at or near it can reach distant observers, at least for a finite period before the apparent horizon eventually swallows the region. If, on the other hand, the apparent horizon forms early enough in the collapse, before the singularity has a chance to communicate with the outside, then every outgoing light ray is trapped, and the singularity is destined to be hidden within a black hole. The relative timing of these two events, the birth of trapped surfaces and the birth of the singularity, therefore acts as the decisive switch between the two possible end states.

This timing criterion connects the new work to a long lineage of collapse studies stretching back to the foundational models of the 1930s and the numerical analyses of Eardley and Smarr in 1979, which first made explicit how the apparent horizon curve in a collapsing dust cloud determines what external observers can see. Subsequent work by Christodoulou in the 1980s and 1990s demonstrated both violations of cosmic censorship in dust collapse and the instability of certain naked singularities in scalar field collapse, while Joshi and collaborators showed through the 1990s and 2000s that naked singularities arise in inhomogeneous dust, perfect fluid, and scalar field models under wide classes of initial conditions. The new paper synthesizes this tradition by characterizing, in terms of the apparent horizon’s dynamics, precisely when visibility occurs for a general type-I matter field rather than for one specialized equation of state at a time.

Why does the distinction between apparent horizon and event horizon matter so much? The event horizon of a black hole is teleological: its location at any instant depends on the entire future history of the spacetime, including whether collapse will ever end. This makes it useless for real-time diagnostics, both analytically and in numerical simulations. The apparent horizon, by contrast, is determined by the local geometry, specifically by the vanishing of the expansions of both future-directed null congruences emanating from a closed two-surface. Because of this quasi-local character, apparent horizons can be located in numerical relativity codes during dynamical simulations, and they have become the workhorse of modern black hole physics, from modeling binary mergers detected by gravitational-wave observatories to testing general relativity with future detectors such as LISA. Establishing that the apparent horizon’s evolution controls singularity visibility therefore ties an abstract question about censorship to quantities that observers and simulators can actually compute.

The results also carry weight for observational astrophysics. If naked singularities can form in realistic collapse, they would not be mere mathematical curiosities but potential astrophysical objects with distinct observational signatures. Recent theoretical work has explored how accretion flows around naked singularities differ from those around black holes: hydrodynamic and general-relativistic magnetohydrodynamic simulations of accretion onto Reissner-Nordström-type naked singularities and timelike naked singularities reveal outflows and disc properties that differ qualitatively from the black hole case. Earlier studies by Joshi, Malafarina, and Narayan showed that accretion disc profiles can in principle distinguish black holes from naked singularities. A rigorous characterization of when collapse produces visible singularities, grounded in the apparent horizon dynamics, helps identify the parameter regimes in which such exotic objects might plausibly exist and be searched for with X-ray and gravitational-wave facilities.

There are also connections to quantum gravity and to the critical phenomena discovered by Choptuik in 1993, where scalar field collapse exhibits universal scaling behavior near the threshold of black hole formation, with naked singularities appearing exactly at the critical point. The authors’ own recent research program has extended this picture, examining universality and criticality in massless scalar field collapse and the possible formation of primordial naked singularities in the early universe. If tiny naked singularities were produced in the primordial cosmos, their subsequent quantum evaporation, a possibility explored by Goswami, Joshi, and Singh in 2006, could have left observable imprints. The new timing criterion provides a framework for assessing whether such scenarios are dynamically permitted under generic, physically reasonable initial data.

Of course, important caveats remain. The analysis is restricted to spherical symmetry, and the cosmic censorship question in full generality, with rotation, gravitational radiation, and arbitrary asymmetries, remains open. Numerical studies of nonspherical collapse, from Shapiro and Teukolsky’s rotating spheroids to modern binary black hole simulations, continue to probe whether censorship holds beyond the symmetric idealization. Nevertheless, by reducing the visibility question to the concrete, computable dynamics of the apparent horizon and trapped surfaces, and by establishing the result for general type-I matter under standard energy and regularity assumptions, Koushiki and Joshi have converted a famously intractable conjecture into a sharply posed dynamical problem. The race between trapped surfaces and the singularity, it turns out, is the whole story, and knowing the rules of that race brings physicists one step closer to deciding whether the universe’s most violent endings are always hidden from view.

Subject of Research: Visibility of spacetime singularities and apparent horizon dynamics in spherical gravitational collapse

Article Title: Apparent horizon and causal structure of spacetime singularities

Article References: Koushiki, & Joshi, P. S. (2026). Apparent horizon and causal structure of spacetime singularities. General Relativity and Gravitation, 58(10), Article 114. https://doi.org/10.1007/s10714-026-03618-y

Image Credits: AI Generated

DOI: 10.1007/s10714-026-03618-y

Keywords: general relativity, gravitational collapse, spacetime singularities, apparent horizon, trapped surfaces, naked singularity, cosmic censorship, black holes, null geodesics, event horizon, type-I matter field, weak energy condition