On a summer weekend a few years ago my wife and I visited friends at their ranch in the Glass Mountains of West Texas. After dinner we sat outside on lawn chairs and looked up at the sky. Far from city lights, in clear, dry air, and with the moon down, we could see not only Altair and Vega and the other bright stars that you can see from anywhere on a cloudless summer night, but also an irregular swath of light running across the sky, the Milky Way, as I had not seen it in decades of living in cities.

The view is something of an illusion: the Milky Way is not something out there, far from us—rather, we are in it. It is our galaxy: a flat disk of about a hundred billion stars, almost a hundred thousand light years across, within which our own solar system is orbiting, two thirds of the way out from the center. What we see in the sky as the Milky Way is the combined light of the many stars that are in our line of sight when we look out along the plane of the disk, almost all of them too far away to be seen separately. Staring at the Milky Way and not being able to make out individual stars in it gave me a chilling sense of how big it is, and I found myself holding on tightly to the arms of my lawn chair.

Astronomers used to think that our galaxy was the whole universe, but now we know that it is one of many billions of galaxies, extending out billions of light years in all directions. The universe is expanding: any typical galaxy is rushing away from any other with a speed proportional to the distance between them. As the universe becomes less crowded it is also becoming colder. Observing what is happening now, and using what we know of the laws of physics, we can reconstruct what must have been happening in the past.

Here is the account that is now accepted by almost all working cosmologists. About 10 to 15 billion years ago, the contents of the universe were so crowded together that there could be no galaxies or stars or even atoms or atomic nuclei. There were only particles of matter and antimatter and light, uniformly filling all space. No definite starting temperature is known, but our calculations tell us that the contents of the universe must once have had a temperature of at least a thousand trillion degrees centigrade. At such temperatures, particles of matter and antimatter were continually converting into light, and being created again from light. Meanwhile, the particles were also rapidly rushing apart, just as the galaxies are now. This expansion caused a fast cooling of the particles, in the same way that a refrigerator is cooled by the expansion of the freon gas in its coils. After a few seconds, the temperature of the matter, antimatter, and light had dropped to about ten billion degrees. Light no longer had enough energy to turn into matter and antimatter. Almost all matter and antimatter particles annihilated each other, but (for reasons that are somewhat mysterious) there was a slight excess of matter particles—electrons, protons, and neutrons—which could find no antimatter particles to annihilate them, and they therefore survived this great extinction. After three more minutes of expansion the leftover matter became cold enough (about a billion degrees) for protons and neutrons to bind together into the nuclei of the lightest elements: hydrogen, helium, and lithium.

For three hundred thousand years the expanding matter and light remained too hot for nuclei and electrons to join together as atoms. Stars or galaxies could not begin to form because light exerts strong pressure on free electrons, so any clump of electrons and nuclei would have been blasted apart by light pressure before its gravity could begin to attract more matter. Then, when the temperature dropped to about three thousand degrees, almost all electrons and nuclei became bound into atoms, in what astronomers call the epoch of recombination. (The “re” in “recombination” is misleading. At the time of recombination electrons and nuclei had never before been combined into atoms.) After recombination, gravitation began to draw matter together into galaxies and then into stars. There it was cooked into all the heavier elements, including those like iron and oxygen from which, billions of years later, our earth was formed.

This account is what is commonly known as the big-bang cosmology. As the term is used by cosmologists, the big bang was not an explosion that occurred sometime in the past; it is an explosion involving all of the universe we can see, that has been going on for 10 to 15 billion years, since as far back in time as we can reliably trace the history of the universe, and it will doubtless continue for billions of years to come, and perhaps forever.


Though the big-bang cosmology is a great scientific achievement, it has not answered the most interesting question: What is the origin of the universe?—not what happened in the first three hundred thousand years of the big bang, or in its first three minutes, or in its first few seconds, but at the very start, if there was one, or even before. In 1992 I was among a group of physicists trying to sell Vice President Gore on the need for a large new elementary particle accelerator, the Superconducting Super Collider. Knowing that even people who couldn’t care less about the laws of physics are curious about cosmology, we described how this instrument might lead to the discovery of mysterious particles that are thought to inhabit the voids between the galaxies and to make up most of the matter of the universe. The Vice President listened politely and promised warm support from the Clinton administration.1 Then, just as he was leaving, he turned back into the room, and diffidently asked if we could tell him what happened before the big bang.

I don’t recall what we answered, but I am sure that it was not very enlightening. No one is certain what happened before the big bang, or even if the question has any meaning. When they thought about it at all, most physicists and astronomers supposed until recently that the universe started in an instant of infinite temperature and density at which time itself began, so that questions about what happened before the big bang are meaningless, like questions about what happens at temperatures below absolute zero. Some theologians welcome this view, presumably because it bears a resemblance to scriptural accounts of creation. Moses Maimonides taught that “the foundation of our faith is the belief that God created the Universe from nothing; that time did not exist previously, but was created….”2 Saint Augustine thought the same.

But opinions among cosmologists have been shifting lately, toward a more complicated and far-reaching picture of the origin of the universe. This new view is given prominence in the books under review by Timothy Ferris, Alan Guth, and Sir Martin Rees. All three books also give clear introductions to the standard big-bang theory and to the physical theories used by cosmologists.

Ferris is emeritus professor of journalism at the University of California at Berkeley, and an experienced writer on science for the general reader. He has an attractive, cool style, and his book is broad in scope, extending even into philosophical matters, but without the gushing or silliness that sometimes characterizes discussions of cosmology. Although he is not a working scientist, Ferris has a sure grasp of the science he describes. I do not think there is a better popular treatment of some of the topics that Ferris covers. (He tells some good stories too. Since the late 1960s, he writes, the physicist Ray Davis has been measuring the rate at which the sun emits the elusive particles known as neutrinos, by catching one now and then in a large tank of perchlorethylene, a common dry-cleaning fluid. After ordering 100,000 gallons of perchlorethylene from a chemical supply house, Davis began to receive advertisements from a supplier of coat hangers.)

Rees is an eminent astrophysicist, currently the fifteenth Astronomer Royal of England and professor at the University of Cambridge, and one of the best expositors of his subject. Because he has made important contributions to many problems of astrophysics, Rees is able to give the reader a better sense of how science is done than most philosophers of science, in the way that a good military memoir—by Grant, say, or Bradley—gives a better sense of the nature of war than the generalities of Sun Tzu or Clausewitz. He makes some telling points about science policy: he is, for example, dead right about the harm done to astronomical research by our commitment to the manned space program. He quotes Riccardo Giacconi, the first director of the Space Telescope Science Institute, as saying that if from the beginning it had been planned that an unmanned rocket instead of the space shuttle would put the Hubble Space Telescope into orbit, then seven similar telescopes could have been built and launched for what so far has been spent on just one. Rees’s book also evocatively describes scientists he has known—not only those, like Stephen Hawking and Roger Penrose, whose names are familiar to the public, but others who have also made major contributions to astrophysics, including Subrahmayan Chandrasekhar, Robert Dicke, John Wheeler, Yakov Zeldovich, and Fritz Zwicky.

Guth is a young professor of physics at MIT. Though this is his first book for the general reader, Guth writes engagingly and understandably. His book concentrates on research on the beginning of the universe to which he has made a major contribution, and so it is an important historical document as well as enjoyable to read.


In December 1979 Guth was a postdoctoral research associate visiting the Stanford Linear Accelerator Center, worrying about where he would get his next job. Together with a Cornell colleague, Henry Tye, he was studying the cosmological effects of certain physical fields. Fields are conditions of space itself, considered apart from any matter that may be in it. Fields can change from moment to moment and from point to point in space, something like the way that temperature and wind velocity are conditions of the air that can vary with time and position in the atmosphere. The most familiar example of a field is the gravitational field, which we all feel tugging us toward the center of the earth. Most people have also felt the magnetic field pulling a piece of iron in their hand toward the north or south pole of a bar magnet. In the modern theory of elementary particles known as the Standard Model, a theory that has been well verified experimentally, the fundamental components of nature are a few dozen different kinds of field.

The fields considered by Guth in this work are called scalar fields, which means that they are made up of quantities which like the air temperature are purely numerical, in contrast to gravitational and magnetic fields, which like wind velocity point in a definite direction. Scalar fields do not tug at anything, so we are not normally conscious of them, but physicists think that they pervade the present universe. In the simplest version of the Standard Model of elementary particles, it is the action of scalar fields on electrons and quarks and other elementary particles that gives these particles their masses.

All fields can carry energy, so these scalar fields can give an energy even to otherwise empty space. According to Einstein’s General Theory of Relativity, all forms of energy affect the rate of expansion of the universe. To judge from the measured rate of expansion of the universe, empty space now has almost no energy. But the energy of any field naturally depends on its strength. For instance, no one notices the energy in the weak magnetic field of an ordinary bar magnet, but the much stronger magnetic fields of modern electromagnets have energies that can physically wreck the magnets if they are not carefully designed. The strengths of the scalar fields considered by Guth were different under the conditions of the early universe from what they are now; they gave “empty space” an enormous energy, quite unlike the zero-energy space we live in now.

Guth calculated that in the early universe the energy of the scalar fields would have remained constant for a while as the universe expanded, which would produce a constant rate of expansion, in contrast to the situation in the present universe, where the rate of expansion decreases as the density of matter decreases. With a constant rate of expansion the universe would have grown exponentially, like the growth of a bank account with a constant rate of compound interest, but with the size of the universe doubling again and again in a tiny fraction of a second. Guth called this phenomenon “inflation.” The possibility of an exponential expansion had been realized by others, including Andrei Linde and Gennady Chibisov at the Lebedev Physical Institute in Moscow, and in itself this would have been a technical result of interest only to other physicists. But then it occurred to Guth that the existence of an era of inflation would solve one of the outstanding problems of cosmology, known as the “flatness problem.”

The problem is to understand why the curvature of space was so small in the early universe. General Relativity tells us that space can be curved, and that this curvature has an effect on the rate of expansion of the universe that is similar to the effect of the energy in matter and in the scalar fields. Other things being equal, the greater the curvature the faster the expansion. We don’t know precisely what the curvature of space is right now, but from measurements of the rate of expansion and the amount of matter in clusters of galaxies we do know that the energy of matter accounts for at least 10 percent of the rate of expansion, and maybe all of it. This leaves at most 90 percent of the expansion rate to be produced by curvature. But as the universe has been expanding, the density of matter has been falling, so the fraction of the expansion rate due to curvature has been steadily growing—if it is no more than 90 percent now, then in the first second of the big bang it must have been less than about one part in a thousand trillion. This is not a paradox—there is no reason why the curvature should not be very small—but it is the sort of thing physicists would like to explain if they could.

What Guth realized was that during inflation the fraction of the expansion rate caused by curvature would have been rapidly decreasing. (The reason is that the curvature was itself decreasing while the energy of the scalar fields remained roughly constant.) So to understand why space was so flat at the beginning of the present big bang it is not necessary to make any arbitrary assumptions; if the big bang was preceded by a sufficient period of inflation, it would necessarily have started with negligible curvature. Guth wrote in his diary that inflation “can explain why the universe [was] so incredibly flat,” and added a heading, “SPECTACULAR REALIZATION.” So it was.

Guth soon also discovered that inflation would solve other cosmological puzzles, some of which he had not even realized were puzzles. One of these is known as the “horizon problem.” Conditions in the universe at the time of recombination—when almost all electrons and nuclei became bound into atoms—seem to have been pretty uniform over distances of at least ninety million light years. This is revealed by observation of microwave radiation left over from that time. The problem is that at recombination there had not yet been enough time since the beginning of the big bang for light or anything else to have traveled more than a small fraction of this distance, so that no physical influence would have been able to stir things up to form the uniform texture we observe.

Like the flatness problem, this is not a logical contradiction—there is no reason why the universe could not have started out perfectly uniform—but, again, it is the sort of thing we hope to explain. Guth found that inflation provides an explanation: during the inflationary era the part of the universe that we can observe would have occupied a tiny space, and there would have been plenty of time for everything in this space to be homogenized. Guth gave a talk about his work a few weeks later, and the next day received two job offers and three invitations to give talks at other physics departments.

Then Guth did something that might seem odd: he went to work to find things wrong with his inflation theory. This is the way physics works; Guth must have known that if something were wrong with inflationary cosmology then this would soon become clear to many physicists, and Guth would naturally want to be the first one who found that something was wrong with his theory and would have a chance to fix it. It is thus that the existence of a common standard of judgment leads physicists, who are no more saintly than economists, to question their own best work.

As it turned out, there was something wrong with the original version of inflation theory. Andrei Linde in Moscow seems to have been the first to realize this, but it was discovered independently by a host of others: Guth and Erick Weinberg of Columbia University; Paul Steinhardt and Andreas Albrecht at the University of Pennsylvania; and Stephen Hawking, Ian Moss, and John Stewart at the University of Cambridge. The difficulty had to do with the end of inflation. Guth had originally assumed it ended with what is called a phase transition, like the freezing of water at zero degrees centigrade or its boiling at 100 degrees.

The phase transition at the end of inflation was of course different from the one in which water turns into ice or vapor; rather, when the temperature of the universe dropped to about a million trillion trillion degrees, the scalar fields jumped from their original values—the numbers that characterize the strengths of the fields—to the values they have now. During this transition bubbles of ordinary zero-energy empty space would have formed here and there, like bubbles of vapor formed by boiling water. The energy that had been in the scalar fields during inflation would have wound up on the surfaces of the bubbles.

Guth had thought at first that these bubbles would have merged and disappeared, spreading the energy on their surfaces evenly throughout space, after which the universe could be described by the conventional big-bang theory. But calculations showed that although the bubbles were expanding rapidly, the universe was expanding faster, so that the bubbles never could have merged with one another. These calculations posed an immediate difficulty for Guth: there would be no place in such a universe that would look like the big bang in which we find ourselves. Since we now live in zero-energy space, our own big bang could only be located inside one of these bubbles; but by the end of inflation the bubbles would have expanded so much that the density of matter and light inside them would be infinitesimal.

Linde and—independently—Albrecht and Steinhardt also found a way out of this difficulty. Making different assumptions about the physical forces responsible for the release of the space energy, they showed that this energy could have leaked into the interior of the bubbles of ordinary space during inflation, eventually turning into matter and light at temperatures around a million trillion trillion degrees. This matter and light would then have expanded and cooled as in the conventional big-bang theory. The part of the universe we observe, extending billions of light years and containing billions of galaxies, thus would be just a tiny part of the interior of one of these bubbles. There would be countless other bubbles of ordinary space, too far away to see, and presumably many of these would develop into big bangs like our own. Soon after this work, Guth gave a talk at the Harvard-MIT Joint Theoretical Seminar, with the subtitle “How Linde and Steinhardt Solved the Problems of Cosmology, While I Was Asleep.”

Loosely put, each bubble of ordinary space could be called a “universe”—this is what Ferris and Rees mean when they refer to “universe(s)” and “our universe and others” in their subtitles. If instead we stick to the usual definition of “universe” as everything, the whole shebang, then the idea of a multiplicity of big bangs, if correct, would represent an enormous enlargement of our notion of the universe. It would be the third step in a historical progression that started with the suggestion in 1584 by Giordano Bruno that the stars are suns like our own, and continued with the demonstration by Edwin Hubble in 1923 that many faint patches of light in the sky are galaxies like our own.

This “new inflationary cosmology” has its own internal problems. Since 1980 other inflation theories have proliferated. To me, the most interesting is the “chaotic inflation” theory of Linde (now at Stanford). He makes the reasonable assumption that the scalar fields did not start at the beginning of time with some definite value, uniform throughout the universe, but instead were fluctuating wildly, so that inflation began here and there at different times.

Chaotic inflation opens up the possibility I mentioned earlier, of a new view of what happened before our big bang. If the scalar fields don’t evolve in lock step everywhere in the universe, then very far away there may have been other big bangs before our own, and there may be others yet to come. Meanwhile the whole universe goes on expanding, so there is always plenty of room for more big bangs. Thus, although our own big bang had a definite beginning about 10 to 15 billion years ago, the bubbling up of new big bangs may have been going on forever in a universe that is infinitely old.

An even stranger idea has been gaining ground lately. Just as what we have been calling the universe may be only a tiny part of the whole, so also what we usually call the “constants of nature,” like the masses we ascribe to the elementary particles, may vary from one part of the universe to another. Inflationary cosmology offers a concrete realization of this idea. The evolution of the scalar fields within each expanding bubble may lead to final values of these fields that differ from one bubble to another, in which case each big bang would wind up with different values for physical constants. (The Harvard theorist Sidney Coleman has shown how something like this can happen even apart from inflation when quantum mechanics is applied to the whole universe.) In any case, if for some reason or other the constants of nature vary from one part of the universe to another, then it would be no mystery why these constants are observed to have values that allow for the appearance of intelligent life: why, for instance, the charge and mass of the electron are so small. Because they are small, the force between electrons and quarks is too weak to keep electrons inside atomic nuclei, so that electrons in atoms form clouds outside the nuclei, which hold atoms together in chemical molecules, including those necessary for life. Only in the parts of the universe where the constants have such values is there anyone to worry about it.

Arguments of this sort may explain why in the present universe there is almost no energy in empty space, despite the fact that according to quantum mechanics there are continual fluctuations in the gravitational and electromagnetic and other fields that by themselves would give empty space an enormous energy. Here and there in the universe there are regions where by chance the scalar fields happen to wind up at the end of inflation with negative energies that cancel almost all the energy of the field fluctuations. In the far more numerous regions where empty space winds up with large energy, there are forces that would prevent the formation of stars and galaxies; and so there would, in those regions, be no one who could raise the question of the energy of space.

This sort of reasoning is called anthropic, and it has a bad name among physicists. Although I have used such arguments myself in some of my own work on the problem of the vacuum energy,3 I am not that fond of anthropic reasoning. I would personally be much happier if the values of all the constants of nature could be precisely calculated on the basis of fundamental principles, rather than having to think about what values are favorable to life. But nature cares little about what physicists prefer.

How much of all this are we to believe? As Ferris, Guth, and Rees all make clear, in answering this question one must distinguish between the big-bang theory itself, which describes what happened once the temperature of the observable part of the universe dropped below a few trillion degrees, and the inflationary cosmologies, which try to account for what happened earlier.

About the big-bang theory we can be quite confident. Our understanding of the laws of physics is good enough to allow us to trace the history of the big bang back to a time when the temperature of the universe was a thousand trillion degrees. Also, until the formation of galaxies, conditions in the universe were much the same everywhere, so that in our calculations we don’t have to deal with complicated differences among local conditions of the sort that we find here on Earth, which make it so difficult to predict whether it will rain next week.

The big-bang theory is also confirmed by the discovery of various relics of the early universe. The most dramatic such relic is a whisper of microwave radiation that was produced in the epoch of recombination, radiation that has been cooled by the thousandfold expansion of the universe since then to a temperature of three (actually 2.73) degrees above absolute zero.

The most convincing quantitative evidence for the big-bang theory comes from another relic: five isotopes of the lightest elements, found spectroscopically in interstellar matter that has not yet been processed into stars. The measured abundances of these five isotopes generally agree nicely with calculations of the amounts of the same isotopes which, according to the theory, were produced by nuclear reactions at the end of the first three minutes.

The big-bang theory is not a temporary theoretical fashion, likely to be blown away by the next round of astronomical observation, but is almost certain to endure as part of any future theory of the universe. Ferris remarks that this conclusion “may seem curious to readers of the many newspaper and magazine articles that have appeared during the past decade proclaiming that this or that observational finding has put the big-bang theory in jeopardy.” He quotes Time magazine reporting that the big-bang theory is “unravelling.” Journalists generally have no bias toward one cosmological theory or another, but many have a natural preference for excitement. It is exciting to report that some new observation threatens to throw the big-bang theory into the dustbin of history. It is dull to report that although some detail of the theory has been put in question, the big-bang theory itself is doing well. It is like one of those dull headlines that journalism students are warned about, such as “Crime Rates Remain Low in Toronto.”

To be fair, I should add that overheated science journalism is occasionally abetted by scientists. Martin Rees has hard words for some astronomers, and justly points out that “journalists sometimes need to assess scientists’ claims with as much skepticism as they customarily bring tothose of politicians.”

Although the big-bang theory is overwhelmingly the consensus view of physicists and astronomers, you can still find dissidents among respected scientists who have a longstanding stake in other theories. One alternative is the “steady-state” theory, in which there is no evolution of the universe as a whole; rather, new matter is always being created to fill up the gaps between the receding galaxies. Sir Fred Hoyle, one of the authors of the steady-state theory, coined the term “big-bang theory” in order to poke fun at the consensus. The idea that the universe had no start appeals to many physicists philosophically, because it avoids a supernatural act of creation. And the idea that the universe does not evolve is attractive pragmatically, because if it is true at all, it could only be true under stringent conditions on the contents of the universe, and these conditions would give us an extra handle on the problem of explaining why things are the way they are. Chaotic inflation has in a sense revived the idea of a steady-state theory in a grander form; our own big bang may be just one episode in a much larger universe, which, on average, never changes. But the original form of the steady-state theory as an alternative to the big bang was convincingly ruled out by the discovery of the three-degree cosmic microwave radiation.

It is conceivable that some of the skeptics will turn out to be right about the big-bang theory, but this seems unlikely. Rees cautiously gives odds of only 10 to 1 in favor of the big bang, but he quotes Yakov Zeldovich as saying that the big bang is “as certain as that the Earth goes round the Sun.” At least within the past century, no other major theory that became the consensus view of physicists or astronomers—in the way that the big bang theory has—has ever turned out to be simply wrong. Our theories have often turned out to be valid only in a more limited context than we had thought, or valid for reasons that we had not understood. But they are not simply wrong—not in the way, for instance, that the cosmology of Ptolemy or Dante is wrong. Consensus is forced on us by nature itself, not by some orthodox scientific establishment; as Rees says, “Philosophers of science would be surprised at how many astronomers are eager rather than reluctant to join a revolutionary bandwagon.”

Among skeptics outside the sciences, there are those multiculturalists who don’t so much disagree with the standard cosmological theory as avoid the question of its objective truth. They see modern science as an expression of our “Western” civilization; it works for us, but the belief that the Milky Way is a river in the sky worked for the Mayans, and the belief that the Milky Way is a great canoe rowed by a one-legged paddler worked for the early peoples of the Amazon basin, so who can say that one belief is better than another? I can.

For one thing, modern cosmology is not confined within Western culture. European astronomy received important contributions from Egypt, Babylon, Persia, and the Arabs, and today astronomy and physics are pursued in the same way throughout the world. The West is not so unanimous about science; it has no shortage of believers in astrology, in “Heaven’s Gate,” and similar nonsense.

Apart from its inaccuracy, there is a certain risk in the attempt to tie modern cosmology to Western civilization. Whether or not the Mayans and early Amazonians and other ancient peoples believed in the objective truth of their theories of the Milky Way—they may have just used these theories to suggest convenient markers in the sky, like our own Big Dipper—they certainly did not know that the Milky Way is actually a disk containing billions of stars like our sun, much like billions of other galaxies throughout the universe. Anyone who became convinced that modern cosmology was peculiarly Western, and who did care about objective truth, might reasonably conclude that Western civilization is superior to all others in at least one respect, that in trying to understand what you can see in the sky on a starry night, Western astronomers got it right.

In contrast to the big-bang theory, the theory of inflation is a good idea that explains a lot, but we can’t yet be confident that it is right. There is no consensus here; Rees quotes Roger Penrose as saying that inflation is a “fashion the high-energy physicists have visited on the cosmologists” and that “even aardvarks think their offspring are beautiful.” I don’t agree with Penrose, but it is certainly true that it will be difficult to settle on a specific inflationary cosmology and decide if it is correct, for reasons both of astronomy and of physics.

The only relic we know that would have survived from the era of inflation and that would allow a quantitative astronomical test of the theory is the pattern of nonuniformities in matter and light. Quantum mechanics tells us that during inflation there must have been small fluctuations in the scalar fields. At recombination three hundred thousand years later, these fluctuations would show up as tiny nonuniformities in the temperature of matter and light. Those nonuniformities could be observed in the microwave radiation that comes to us from that time. So far, observations from the Cosmic Background Explorer satellite have given results that agree with the predictions of inflationary cosmology, but the observed nonuniformity is pretty much what had been widely expected before anyone had thought of inflationary cosmology.

A critical test of these new theories will have to wait until new microwave telescopes can study finer details in the cosmic microwave radiation. Even then, it may not be possible to decide for or against inflation, both because these observations are clouded by radio noise from our own galaxy and also because by now there are so many versions of inflationary cosmology. As we make progress in understanding the expanding universe, the problem itself expands, so that the solution seems always to recede from us.

I doubt if it will be possible to decide which version if any of inflationary cosmology is correct on the basis of astronomical observation alone, without fundamental advances in physics. Our present theories of elementary particles are only approximations, which don’t adequately describe conditions at the time of inflation. We don’t even know for sure whether the scalar fields really exist or, if they do, what different types there may be. This question will be settled at least in part by experiments at the next big accelerator, the Large Hadron Collider under construction near Geneva.

We shall also have to solve the old problem of reconciling the theory of gravitation and the principles of quantum mechanics. Rees make a good point that so far we have generally been able to dodge this problem because gravitation and quantum mechanics are almost never both important in the same context. Gravitation governs the motions of planets and stars, but it is too weak to matter much in atoms, while quantum mechanics, though essential in understanding the behavior of electrons in atoms, has negligible effects on the motions of stars or planets. It is only in the very early universe that gravitation and quantum mechanics were both important. Famous theorems of Roger Penrose and Stephen Hawking use General Relativity to show that there must have been a definite beginning to the universe, but their proofs do not take quantum mechanics into account and are therefore inconclusive.

Our best hope for a quantum theory of gravitation lies in the speculative class of theories known as superstring theories.4 Ferris says eloquently that superstring theory has become “a repository of the highest hopes of the finest minds in physics.” Unfortunately, these theories have not yet settled down to a final form, and so far are not confirmed by the success of any quantitative predictions. It may be a long time before we can use these theories to decide whether the universe had a definite beginning, or whether the bubbling up of new big bangs has been going on forever.

About one thing I am sure. Those who think that an infinitely old universe is absurd, so there must have been a first moment in time, and those who think that a first moment is absurd, so the universe must be infinitely old, have one thing in common: whichever side happens to be right about the origin of the universe, the reasoning of both sides is wrong. We don’t know if the universe is infinitely old or if there was a first moment; but neither view is absurd, and the choice between them will not be made by intuition, or by philosophy or theology, but by the ordinary methods of science.

This Issue

June 12, 1997