Demolishing stars, powering blasts of high-energy radiation, rending the fabric of spacetime: it is not hard to see the allure of black holes. They light up the same parts of the brain as monster trucks and battlebots do. They explain violent celestial phenomena that no other body can. They are so extreme, in fact, that no one really knows what they are.

Most researchers think of them as microscopic pinpricks, the remnants of stars that have collapsed under their own weight. But over the past couple of years, a number of mavericks have proposed that black holes are actually extended bodies, made up of an exotic state of matter that congeals, like a liquid turning to ice, during the collapse. The idea offers a provocative way of thinking about quantum gravity, which would unify Einstein's general theory of relativity with quantum mechanics.

In the textbook picture, the pinprick (or singularity) is surrounded by an event horizon. The horizon is not a physical surface, merely a conceptual one, and although it marks the point of no return for material plummeting toward the singularity, relativity says that nothing special happens there; the laws of physics are the same everywhere. For quantum mechanics, though, the event horizon is deeply paradoxical. It allows information to be lost from our world, an act that quantum theory forbids. "What you have been taught in school is almost certainly wrong, because classical black hole spacetimes are inconsistent with quantum mechanics," says physicist George Chapline of Lawrence Livermore National Laboratory.

The new conceptions of black holes eliminate the event horizon altogether. The basic idea is that there does, in fact, exist a force that could halt the collapse of a star when all else fails. That force is gravity itself. In matter with certain properties, gravity switches from being an attractive force to a repulsive force. Such a material, going by the name "dark energy," is thought to be driving the acceleration of cosmic expansion.

Last year physicists Pawel O. Mazur of the University of South Carolina and Emil Mottola of Los Alamos National Laboratory reasoned that a pocket of the stuff might freeze out, like ice crystals, during the collapse of a star. The result, which they call a gravastar, would look like fried ice cream: a crust of dense but otherwise ordinary matter stabilized by a curious interior. The crust replaces what would have been the event horizon.

Another proposal goes further. It conjectures not only that dark energy would freeze out but that relativity would break down altogether. The idea comes from a dark-horse contender for quantum gravity, the proponents of which are struck by the resemblance between the basic laws of physics and the behavior of fluids and solids (also known as condensed matter). In many ways, the equations of sound propagation through a moving fluid are a dead ringer for general relativity; sound waves can get trapped in the fluid much as light gets trapped in a black hole. Maybe spacetime is literally a kind of fluid.

What makes this approach so interesting is that the behavior of condensed matter is collective. The details of individual molecules hardly matter; the system's properties emerge from the act of aggregation. When water freezes, the molecules do not change, but the collective behavior does, and the laws that apply to liquids no longer do. Under the right conditions, a fluid can turn into a superfluid, governed by quantum mechanics even on macroscopic scales. Chapline, along with physicists Evan Hohlfeld, Robert B. Laughlin and David I. Santiago of Stanford University, has proposed that a similar process happens at event horizons. The equations of relativity fail, and new laws emerge. "If one thinks of spacetime as a superfluid, then it is very natural that in fact something physical does happen at the event horizon--that is, the classical event horizon is replaced by a quantum phase transition," Chapline says.?

For now, these ideas are barely more than scribbles on the back of an envelope, and critics have myriad complaints about their plausibility. For example, how exactly would matter or spacetime change state during the collapse of a star? Physicist Scott A. Hughes of the Massachusetts Institute of Technology says, "I don't see how something like a massive star--an object made out of normal fluid, with fairly simple density and pressure relations--can make a transition into something with as bizarre a structure as a gravastar." Mainstream theories of quantum gravity are far better developed. String theory, for one, appears to explain away the paradoxes of black holes without abandoning either event horizons or relativity.

Observationally, the new conceptions of black holes could be hard to distinguish from the classical picture--but not impossible. Gravitational waves should reveal the shape of spacetime around putative black holes. A classical hole, being a simple object without a true surface, has only a couple of possible shapes. If one of the gravitational-wave observatories now going into operation finds a different shape, then the current theories of physics would be yet another thing in the universe to get torn to shreds by a black hole.?