The Singularity That Wasn’t a Black Hole: Penrose’s 1965 Theorem Revisited
19 May 2026, Yanjiang
Penrose’s 1965 theorem proves geodesic incompleteness, not necessarily black holes; cosmic censorship conjectures are required to bridge the gap.
We think of black holes as the universe’s most extreme objects — regions where gravity has crushed matter into a singularity concealed behind an event horizon. But what if the mathematical argument that first convinced physicists such things must exist — Roger Penrose’s famous 1965 singularity theorem — doesn’t, in fact, prove the existence of a black hole? Only a few months ago we celebrated the Nobel Prize for Penrose’s black‑hole work, and now a historical and conceptual analysis by Klaas Landsman, appearing as a preprint (arXiv:2205.01680), contends that the theorem’s content has long been oversold. The paper argues that even if the theorem’s assumptions are all met, as Landsman bluntly writes, “the theorem fails to prove the existence or formation of black holes.”
Every year, hundreds of physics students learn that Penrose’s theorem proves that black holes must form from gravitational collapse. Every year, they could — if they scrutinised the original paper — notice that the theorem actually proves something much more abstract: the presence of an incomplete geodesic, a path through spacetime that simply cannot be extended. The theorem guarantees an “edge” — a geodesic path that cannot be continued — but it says nothing about whether this signals a curvature singularity, let alone whether it is hidden behind an event horizon. It provides a smoking gun, not a confirmed kill.
So what does the theorem actually say? It begins with a trapped surface — a closed two‑dimensional surface such that outward‑directed light rays emanating from it are forced to converge. (Inward rays converge as well, but the defining feature is the convergence of the outward‑directed ones.) This is the mathematical signature of a region from which even light cannot escape. Add standard energy conditions that prevent exotic negative masses, and Penrose’s argument shows that the spacetime containing this trapped surface cannot be both geodesically complete and free of singularities. The only resolution is that some world‑line — a path traced by a photon or a freely falling particle — must hit an edge. That edge is a singularity.
But the theorem never names the culprit. It doesn’t say whether that edge is cloaked by an event horizon, nor whether it is a genuine boundary of existence rather than an artifact of a bad coordinate system. To transform this bland logical outcome into a black hole, physicists need additional hypotheses — ones that Penrose himself introduced soon after: the cosmic censorship conjectures.
Now, Landsman’s study — drawing on decades of earlier critical literature by figures such as Tipler, Clarke, Earman, and Curiel — provides the most meticulous historical dissection of this gap yet. He traces how each assumption was motivated by astrophysical intuition, yet the final theorem stopped short of its astrophysical goal. Penrose, the paper shows, was well aware of the shortfall and spent the subsequent decades trying to bridge it through censorship.
Weak cosmic censorship is the first bridge. The conjecture holds that any singularity that forms from generic, non‑singular initial conditions must be invisible from far away: an event horizon must wrap around it. This converts the abstract “something bad happened” into a “black” object. But the theorem alone cannot enforce this; it could just as happily describe a naked singularity that broadcasts its breakdown to the cosmos at large. If weak censorship fails, determinism itself is threatened, because unpredictable influences could emerge from the singularity to scramble the external world.
Strong cosmic censorship is the second bridge. Even if the singularity is hidden behind a horizon, the theorem still doesn’t guarantee that the interior is a genuine hole. It might only mark the boundary of a region that could be extended — a coordinate patch that runs out. Strong censorship asserts that spacetime is inextendible beyond the horizon: the interior is truly singular, not a portal to another sector. Together, the two censorships would turn the theorem’s geodesic incompleteness into a full‑fledged astrophysical black hole. Without them, the theorem’s edge is ambiguous.
Think of a detective who finds a locked room and deduces that a crime must have occurred because the room was sealed from inside. The detective has proved that a sealed room implies something happened — but not whether it was a crime, an accident, or a door merely painted shut. Penrose’s theorem is that detective: it proves a catastrophic event in spacetime, but does not identify the mechanism. (Unlike a crime scene, though, the theorem’s inference is mathematically inevitable: given the premises, the geodesic must end.)
The insufficiency of the theorem has long been noted by the more philosophically minded. Scholars such as John Earman and Chris Clarke argued already in the 1980s that the theorem’s conclusion was far weaker than the folklore. Landsman’s paper synthesises these earlier critiques and uncovers fresh historical detail, demonstrating that Penrose himself understood the gap: the cosmic censorship conjectures were his explicit response. The 1965 work, in short, was never meant to be the last word; it was a provocation.
The gap was never hidden, yet it is rarely discussed in standard textbooks. Most physicists are taught that Penrose’s theorem proves black holes exist, period. The delicacy of the logical chain — and the dependence on unproven censorship — tends to evaporate in the retelling.
Today, the status of cosmic censorship remains unsettled. Weak censorship has fared well in numerical simulations of realistic collapsing matter, but a general proof remains elusive. Strong censorship, on the other hand, has been challenged by studies of rotating black holes, where quantum effects near the inner horizon might paradoxically smooth the singularity. The theorem, far from closing a chapter, opened a research program that is still in progress.
This state of affairs carries a philosophical lesson about the relation between mathematics and physical reality. General relativity is a geometric theory, and its theorems are statements about differentiable manifolds. To translate those statements into claims about black holes in the sky requires physical hypotheses that go beyond pure geometry — and those hypotheses remain unproven. The gap Landsman documents is not merely a historical curiosity; it reflects the ongoing difficulty of connecting rigorous theorems to astrophysical observation.
Yet Landsman’s analysis does not diminish Penrose’s achievement. Instead, it restores the theorem to its proper role: not as the final word, but as the opening gambit in a chess game that is still being played. The theorem says: under these conditions, spacetime cannot continue smoothly to infinity — there is an edge. The cosmic censorship conjectures are our attempts to characterise that edge as a black hole. Whether every possible edge must be clothed in an event horizon, and whether every internal horizon must conceal a true singularity, remain among the deepest unsettled questions in theoretical physics.
Perhaps the greatest legacy of the 1965 paper is not that it “proved” black holes are real, but that it taught us that proving anything in general relativity is a far more subtle art than we once imagined. The theorem points to an edge; it is up to us, with censorship and observation, to decide whether that edge is a black hole — or something stranger that we have yet to name.
— Yanjiang
Yanjiang is an online editor of LoomSci.com.
References
- Klaas Landsman, Penrose’s 1965 singularity theorem: From geodesic incompleteness to cosmic censorship, arXiv:2205.01680