Could a small imperfection save a collapsing star?
Imagine squeezing a star so tightly that its gravity overwhelms the pressure supporting it. In an ideal calculation, every part can fall inward together and the density can grow without limit. Real stars rotate, have uneven interiors and receive disturbances from their surroundings. For years, that difference left a serious question: was the disastrous ending a feature of nature, or an accident of an overly perfect mathematical star?
Roger Penrose’s 1965 paper changed the question. He found a way to recognize a decisive stage of collapse without following every parcel of matter or demanding spherical symmetry. The key was a closed surface from which even the outward-directed light initially converges. Once such a surface forms, suitable conditions on gravity and the large-scale structure of spacetime force the classical description to become incomplete.
Sources: [1][2]
Light gives the surface a physical meaning
Picture a flash emitted from a spherical shell in ordinary space. Some light heads inward and some heads outward. The inward flash shrinks toward the centre; the outward flash spreads across an ever larger area. That spreading is a local property of the light bundle. You can ask how its cross-sectional area changes just after the flash, without waiting to see where every ray eventually goes.
In a sufficiently collapsing region of curved spacetime, both future-directed families of light can initially shrink. The family we would call outward-pointing also converges. Penrose called a closed surface with this property a trapped surface. The surface can be distorted: its definition concerns how the light expands, so it does not depend on the shell being an exact sphere. This gave the problem a test that survives irregular shape.
Sources: [1][2]
The theorem follows the light and the causal structure
Light travels along the straightest possible paths in spacetime, called null geodesics. “Null” here means lightlike; a ray always travels at light speed locally. Matter and energy curve those paths, and under the theorem’s focusing condition they encourage nearby light rays to converge. Rays leaving a trapped surface already start with that convergence in both future directions.
Following the convergence is only part of the proof. Light rays can cross, much as sunlight crosses after passing a lens, without producing a spacetime singularity. Penrose combined the local focusing with a global argument about which events can influence which others. He assumed a well-behaved causal development from a noncompact initial spatial slice. Roughly, the universe has an extended initial space, and every causal history in the region is determined from it.
Suppose every future light path could continue indefinitely. The focusing from a closed trapped surface then gives the edge of its causal future properties incompatible with that global spacetime structure. At least one assumption must fail. With the stated geometric and causal assumptions retained, some future-directed null geodesic ends after a finite value of its natural path parameter. That precise conclusion is called null geodesic incompleteness.
Sources: [1]
What it means for the theory to become incomplete
A geodesic ending in this way signals that the smooth classical spacetime cannot describe the whole future of the collapse. Physicists call this a singularity theorem. The theorem does not provide a photograph of a tiny central point, specify the matter inside it or prove that a particular curvature quantity diverges along every possible path. Its strength is that it reaches an unavoidable breakdown without first constructing an exact solution for the entire collapsing star.
The conditions carry physical content. The focusing assumption is commonly related, through Einstein’s equations, to a nonnegative lightlike energy density. The initial-space and causality assumptions describe the wider spacetime in which collapse occurs. Changing those conditions can change the conclusion. Quantum effects also lie beyond the classical argument. The result tells us where general relativity’s smooth description demands a deeper account.
Sources: [1][2]
A general argument changed the place of black holes in physics
Earlier exact solutions had made extreme collapse imaginable. Penrose made it robust against the loss of a perfect sphere. A trapped surface could be recognized geometrically in a distorted collapse, and the global argument could then proceed. The theorem transformed black holes from a curiosity tied to special solutions into a serious expectation of relativistic astrophysics.
The 2020 Nobel Prize recognized this contribution alongside a very different kind of work: Reinhard Genzel’s and Andrea Ghez’s observations of stars around the Milky Way’s central compact object. The proof concerns what a theory requires under its assumptions. The orbit measurements concern what a particular astronomical system contains. Together they explain why black holes became both a firm theoretical subject and an observable part of the universe.
Sources: [2][3]
Penrose had already asked where light cannot go
Seven years before the theorem, Penrose posed a smaller puzzle about where light can and cannot reach. In a set of puzzles for Christmas 1958 in New Scientist, he and his father, the geneticist Lionel Penrose, described a room with perfectly mirrored walls and a candle inside. With the candle in the right place, parts of the room stay dark however many times the light bounces.
The ellipse does the work. Light leaving one focus of an ellipse reflects off the curve and passes through the other focus. It follows that a ray crossing the segment between the two foci crosses it again after it reflects, and a ray crossing outside the foci comes back outside them. The room has a half-ellipse at each end of a middle section. Small alcoves open off each half-ellipse only beyond its foci, behind umbrella-shaped walls whose tips sit on the foci. Light from a candle in the middle can enter a half-ellipse only between its foci, so it always returns to the middle and never reaches an alcove.
The puzzle is not part of the 1965 proof, but it asks the same kind of question: taking every possible light path into account at once, which places can light reach? Its exactness has conditions too. The walls must be perfect mirrors and light must travel as rays. Slightly matte walls, or foci moved off the tips, let light leak into the alcoves. Real light is a wave and bends around the tips, so, as Kim and colleagues noted in 2022, no region stays completely dark.
Sources: [4][5]