Scientific American astronomy editor George Musser explains.

This question really has two parts. First, how was matter able to get out of the big-bang singularity? After all, physicists describe a black hole singularity as a pit into which material flows but from which it cannot escape. Let us leave aside the fact that singularities are an idealization. The basic point is that the universe was born with a tendency to expand, which overcame the tendency of matter to collapse. According to relativity theory, space does not like to remain static; for all but the most special cases, it either expands or contracts. But why it initially chose the former is still a mystery.

In some ways, you can think of the universe as a black hole turned inside out. A black hole is a singularity into which material flows. The universe is a singularity out of which material has flowed. A black hole is surrounded by an event horizon, a surface inside which we cannot see. The universe is surrounded by a cosmological horizon, a surface outside of which we cannot see. (A crucial difference, though, is that the event horizon is fixed whereas the cosmological horizon varies from observer to observer.)

The second part of the question is: Why didnt matter in the early universe collapse into black holes? After all, physicists say that if you squeeze matter to a high enough density, it will collapse into a black hole, and the density of matter in the early universe was extremely high. The answer is that black-hole formation actually depends on the variation in density from one place to another--and there was very little variation back then. Matter was spread out almost perfectly smoothly.

In fact, cosmologists usually turn the question around. The fact that the universe did not recollapse into a swarm of black holes is evidence that sharp density variations did not exist (or were extremely rare). This lack of sharp variations, in turn, is evidence for the inflationary model that most cosmologists today accept.

Answer posted September 22, 2003. Below explanations posted October 21, 1999.

Robert J. Nemiroff, assistant professor of physics at Michigan Technological University, responds.

First of all, it is not really known whether or not the universe started from a singularity. Our measurements can take us back only so far; ideas about the nature of the cosmos at the start of the big bang are mostly unproved conjecture.

Second of all, the concept of a black hole is only one type of solution to Einstein's General Theory of Relativity, our best current theory of gravity. This reading of general relativity--known as the Schwarzschild solution--is thought to give an accurate description of the gravity near an isolated, nonrotating black hole, as well as the 'normal' gravity near the earth and throughout our solar system.

But other solutions to general relativity are known to exist, including ones that apply to a whole universe. These alternative solutions typically assume that the early universe was perfectly uniform so that there were no places for black holes to form, even if the density were so great that particles were "cheek by jowl." The most popular class of general relativity solutions applying to the entire cosmos are known as Friedmann-Robertson-Walker solutions. These formulations appear to describe correctly our expanding universe; that is, they demonstrate how objects not held together by local forces (such as the electromagnetism that bonds atoms in molecules or the gravity that keeps the earth intact) stream away from one another in a predictable manner.

Still, there is room in the theories for some of the matter in the universe to be hidden in black holes that might have formed from local, unusually dense regions in the very early universe. These black holes could conceivably contribute to the large amount of dark matter that exists in the universe. Astronomers are therefore diligently searching for these objects. In one scenario discussed by Jeremiah Ostriker of Princeton University and his collaborators, black holes as massive as one million times the mass of our sun might be common throughout the universe and still be nearly invisible. Although other black holes might come out of some big bang models involving quantum mechanics, a common expectation by cosmologists is that only elementary particles survived these early epochs of our universe.

Christ Ftaclas is an associate professor of physics, also at Michigan Tech. He adds the following:

The space-time singularity associated with the big bang differs in two important ways from the singularity associated with a black hole. First of all, a black hole has an "outside." That is, we assume that at large distances from the black hole space-time is essentially flat and defines a background against which we observe the black hole. This is not true in the case of the big bang, because we are all participants.

The second difference is critical to this question: one of the initial conditions of the big bang is expansion of the matter, whereas a Schwarzschild black hole is associated with a static gravitational field. One might think motion would not make a difference, because no velocity is great enough to escape from a black hole, but that is only true for a particle whose motion is measured relative to the stationary black hole. In the case of the big bang, everything is moving, with the result that the solution to the gravitational-field equations is fundamentally altered.

Edward L. ("Ned") Wright is the vice chair for astronomy at the University of California at Los Angeles; he also maintains a thorough on-line Cosmology Tutorial. Wright offers a somewhat different approach to this question:

A black hole is a local region from which light cannot escape. It has a boundary called the event horizon. Inside the event horizon, light cannot escape to infinity, whereas outside the event horizon, light can escape to infinity if it is traveling in the right direction. Even outside the event horizon, however, light that travels straight in toward the black hole will not escape.

In contrast, the universe is thought to be homogeneous and isotropic. Isotropic means that all directions appear the same; this property of the universe is well established by observations that show the effective temperature of the cosmic microwave background is identical in all directions to one part in 100,000. Homogeneous means that any place in the universe is equivalent to any other place. We can observe the universe from only one position, of course, but it does appear to be homogeneous on very large scales, after smoothing over the stars, galaxies, clusters of galaxies and superclusters.

A homogeneous space cannot have a boundary, so there can be no event horizon. And the future behavior of light rays cannot depend on their directions in an isotropic space. Thus, a homogeneous and isotropic universe is not a black hole. The universe does have one similarity to a black hole: light cannot escape from it. But this is true for any place in the universe and for light traveling in any direction, unlike the case for a black hole.

If this proof by contradiction seems unsatisfying, the advanced reader could look at the technical solution to the question. Solving the field equations will show that the space metric in a Friedmann-Robertson-Walker universe expands from infinite density without forming a black hole.

Answer originally posted October 21, 1999.