Sooner or later, even supercooled water freezes. Similarly, the metastable region would eventually experience its own phase transition. Once the forces were no longer unified, the rapid expansion would stop. Linde imagined this happening first in bubbles of space that would abruptly appear in the metastable region like the ice crystals that appear in supercooled water. The bubbles would then quickly trigger the phase transition throughout the metastable region and slow its growth.

Unfortunately, this stunning version of genesis seemed to have fatal flaws. Linde realized that when these bubbles collided, the energy in their walls would create vast sheets of superdense matter, or black holes. He calculated that the odds were good that at least one of the collisions would have taken place within the boundaries of the observable universe. But astronomers don’t see such structures. Moreover, it didn’t seem possible for the bubbles to appear and expand fast enough to engulf the entire metastable region, which would have continued growing at unimaginable speed. Since the Soviet scientists couldn’t solve these problems--and since they saw no compelling reason for trying--they dropped the idea. We thought it was garbage, Linde recalls.

Then in 1980 a young American physicist named Alan Guth rediscovered the expansion mechanism. He called it inflation and showed how it could solve some puzzles posed by conventional Big Bang theory.

One concern is the smoothness of the Big Bang’s faint afterglow, the cosmic background radiation. This microwave radiation, which was discovered in 1965, is one of the strongest pieces of evidence for the Big Bang. In the 1970s, however, some cosmologists began to wonder why the radiation seemed to have precisely the same intensity in all directions. Guth’s solution was that the region from which the observable universe sprang was initially so tiny that it had time to reach thermal equilibrium- -which would smooth out any inhomogeneities--before it began inflating.

Inflation could also account for the flatness of space in our universe. According to Einstein’s theory of relativity, space is warped by matter and energy in a process we perceive as gravity. If the whole universe had enough mass and energy, it would completely warp in on itself; it would be closed. If, on the other hand, the universe had too little mass and energy, it would be open-ended. In either case, the universe would be a weird place--parallel lines would either converge or split apart. Neither seems to be the case, and in fact astrophysicists who have added up the mass and energy in the universe and factored in its size believe that the universe is almost perfectly flat. Some theorists find that a remarkable coincidence: How could the universe achieve that kind of perfection on its own?

Inflation, Guth said, provides an explanation. Astrophysicists have limited themselves to the observable universe, but the theory of inflation says it’s actually an unremarkable fraction of a much larger universe we can never hope to see. Space probably is curved, but the piece of it we know is too small to demonstrate that curvature. The apparent flatness astronomers observe isn’t a coincidence--it is simply a product of the limits of our vision.

Linde immediately recognized the importance of Guth’s paper. He also recognized that Guth hadn’t solved the colliding-bubble problem. Linde renewed his attacks on the problem with such intensity, he says, that he developed an ulcer. The solution finally came to him during a late-night telephone conversation with a colleague. To avoid waking his wife and sons, Linde had been talking in the bathroom. He became so excited by his idea that he woke up his wife anyway. It seems that I know how the universe originated, he told her. (Fortunately for Linde, he has an understanding wife--noted theoretical physicist Renata Kallosh.)

Linde’s answer made the observable universe seem even tinier than before. The entire universe as we know it, Linde decided, is nestled inside a single expanding bubble. Astronomers don’t see the black holes created by colliding bubbles because the walls of superdense matter are pushed beyond the observable borders of our little cosmos. Linde wrote up his work in 1981, and it was published early in 1982. Soon after, Paul Steinhardt, a physicist at the University of Pennsylvania, and Andreas Albrecht, one of Steinhardt’s graduate students, set forth essentially the same idea, which came to be called new inflation.

No sooner had new inflation caught on--and catch on it did, with amazing speed--than Linde discovered another flaw. When he tried to make his scenario jibe with the latest theories of particle physics, he found it wouldn’t work. In effect, a phase transition in the early universe would give rise to inflation only if some special conditions were met, involving a lot of extreme, unwarranted assumptions about how subatomic particles behave. Linde considered it highly unlikely that those conditions could be met by chance.

Then another strange possibility occurred to Linde. According to certain interpretations of quantum physics, classical notions of space and time completely break down on extremely small scales. In the universe’s infancy, Linde thought, when it was trillions of times smaller than the smallest subatomic particle, it might have existed as a chaotic foam of many possible configurations of space and time. Such a space-time foam, he asserted, would inevitably toss up islands of space and time packed with the energy required to trigger inflation. No special circumstances were necessary to create the islands--they would be born out of the turmoil of quantum variability.

Linde’s first chance to tell the world about his chaotic inflation hypothesis came in 1983, when he was invited to a conference on Shelter Island, New York. A number of the world’s top physicists and cosmologists, including Guth, Steinhardt, and British cosmologist Stephen Hawking, were attending. But Linde’s speech was met for the most part with blank stares. Only Hawking understood, Linde says.

Indeed, in 1988 Hawking praised chaotic inflation in his best- selling book A Brief History of Time. Chaotic inflation, Hawking wrote, has all the advantages of the earlier inflationary models, but it does not depend on a dubious phase transition.

In 1986, however, chaotic inflation was far from widely accepted and Linde, the idea machine, was out of ideas. The Soviet government had forbidden scientists from publishing abroad for a year. Also, Linde felt mired in a book he was writing about his ideas on inflation. He fell into a depression so deep that he could not get out of bed. Then on extremely short notice the Soviet Academy of Sciences asked him--ordered him, actually--to deliver a speech at a meeting in Italy. The academy suggested- -demanded, actually--that he present some new theories, rather than hash over his old stuff. Linde realized that this was a rare chance to get something printed abroad. In half an hour of fevered thought, Linde conceived yet another strange possibility--the notion of the eternally self-reproducing universe.

Linde looked back at the moment when part of the space-time foam inflated. The energy that triggered the inflation, he realized, would slowly dwindle as it got diluted in the expanding space--but it would not dwindle everywhere. Thanks to quantum variations, the energy level would vary from point to point. In some regions of the universe the variation would lower the energy level, and soon inflation would peter out completely. Our visible universe is deep within one such region. Meanwhile, though, other areas would experience a quantum power surge. Instead of winding down, they would suddenly inflate even more furiously, expanding into their own enormous universes. Of course, the energy level would fluctuate inside them too: some regions would soon slow down their expansion, while others would continue the inflationary cycle, breeding universes without end.

In a self-reproducing universe, there’s nothing that says that our visible universe isn’t just an offshoot from another, older universe. In fact, in the tangle of reproducing universes Linde sees, he isn’t even sure he needs to rely on a primordial space-time foam anymore--or, for that matter, a traditional Big Bang. The evolution of the universe as a whole has no end, Linde says, and it may have had no beginning.

Linde is not the first physicist to posit the existence of other universes. In the late 1950s, for example, Hugh Everett III of Princeton proposed that, even though subatomic particles seem to follow only a single path, they actually follow all the paths mathematically allowed them by quantum physics--in separate universes. But most theorists treat other universes as mathematical abstractions, and somewhat embarrassing ones at that. Linde, on the other hand, delights in imagining what these alien worlds might be like. To do so, he borrows liberally from the language of genetics. Some of these offshoot universes retain the genes of their predecessors, he says, and evolve into universes with similar physical laws--and perhaps similar inhabitants. The fact that somewhere else there is life like ours is to me almost certain, says Linde. But perhaps we can never know this.

Other offshoot uni-verses may undergo mutations and evolve with constants of nature, physical laws, and even dimensions unlike ours. Of course, we could never hope to enter these exotic realms, but Linde has speculated that if our universe becomes uninhabitable in the future, we might be able to travel to a nearby one with conditions like our own.

Yet even now our universe might at some level be influenced by others. This thought occurred to Linde in 1987, while he was in the United States presenting his ideas on a tour of universities. Between lectures Linde wrote a little paper explaining his new idea. The paper’s primary aim was to explore one of the deepest mysteries of modern physics: the vacuum-energy problem. Vacuum energy refers to the amount of energy that inhabits empty space. You might suspect that empty space should be, well, empty. But according to quantum mechanics, the vacuum is never completely empty. It is pervaded by fluctuating fields of energy whose ultimate influence on space should be enormous. Yet, by what seems to be an incredible coincidence, the universe as a whole shows no trace of either positive energy, which would push space outward, or negative energy, which would make space contract. Why is the cosmic vacuum energy--as far as anyone can tell--precisely zero?

Linde came up with a typically far-out explanation. Some linkage between our universe and an unseen mirror universe, one with opposite energy values, cancels out any vacuum energy in either cosmos. It was absolutely crazy, Linde admits. Nevertheless he presented the hypothesis to various physicists he encountered on his tour, including Harvard’s Sidney Coleman. Most of them agreed with Linde: his idea was completely crazy. Linde went back to Moscow somewhat abashed.

Then a year later Coleman sent Linde a paper. Coleman had taken a different route than Linde did but arrived at the same conclusion: infinitesimal channels between our universe and others--what Coleman called wormholes--serve to cancel out the vacuum energy. Coleman’s paper, which cited Linde’s inspiring contribution, is still a hotly debated issue. The episode enhanced Linde’s already considerable reputation as a seminal thinker, and in 1988 CERN invited him and his wife to visit for a year. They left Moscow at the end of 1988 and have returned only for brief visits since then.

Even as Linde’s ideas have gained acceptance and respect, inflation itself is increasingly under attack. As Guth pointed out in 1980, one of the strengths of the theory is its ability to explain the exceptional smoothness of the cosmic background radiation. But recent surveys of galaxies show them huddled together in gigantic clusters surrounded by gigantic voids. If inflation made the universe so smooth early on, critics ask, how did it get to be so clumpy? In explaining why space appears to be flat, inflation also makes predictions about the density of the universe, but skeptics have pointed out that measurements of the total amount of matter in the universe have fallen short of those predictions.

Linde is fighting back, pointing out that inflation does explain how the universe can get clumpy. When an inflationary patch of space is expanding, little quantum fluctuations in its mass and energy get magnified as the space grows. True, the recent estimates of the clumpiness and mass density of the universe are beyond what inflation predicts. But, Linde says, the surveys are tentative, and chaotic inflation, with some modifications, can accommodate most of them. So far, there are nothing but words, Linde says of those who have tried to declare inflation dead.

Ever the showman, Linde is also trying to find ways to pitch his ideas. Taking advantage of his new milieu, for example, he recently persuaded a Silicon Valley company to lend him a $200,000 state-of-the-art computer-graphics system that can illustrate his chaotic, eternally self- reproducing inflation theory. The day after the company dropped off the computer at Linde’s house, his son Dmitri, a budding hacker, had it running a program illustrating chaotic inflation.

He is the computer expert, says Linde, proudly patting Dmitri’s shoulder. Dmitri generates an image of a Day-Glo mountain range. Through quantum fluctuations these peaks have risen up from a flat surface representing a patch of space. Each corresponds to an island of space-time where the conditions are right for inflation and where a new universe is being hatched. People still think of only a single Big Bang, Linde says. The only way to show them is to make pictures.

Fortunately for physics, defending inflation isn’t taking up all Linde’s time. Lately he’s been thinking about whether it is possible to manufacture another universe. Several physicists, including Guth, have toyed with this idea, calculating the materials and conditions required to trigger inflation in a laboratory. (You’d need only about 20 pounds of matter, Guth decided, but you’d have to squash it down to subatomic size. Linde sniffs at that weight. In his model of the universe, you need less than a millionth of an ounce.)

However, Linde much prefers to ask what he thinks is a more pertinent question: Why would someone want to create another universe? There would be no way for the creator to enter the other universe or even communicate with it; once it began inflating, it would almost instantaneously branch from its parent at faster-than-light speeds.

As usual, Linde is the first to arrive at an answer to his own question. Perhaps, he says, you could manipulate the seed of preinflation stuff in such a way that it evolved into a universe with particular dimensions, physical laws, and constants of nature. In that way the creator could impress a message onto the very structure of that universe.

In fact, Linde suggests, our own universe might have been created by beings in another universe, and physicists, in their fumbling attempts to unravel the physical structure of our world, may be on the path to decoding a message from our cosmic parents. After presenting this idea, Linde grants himself a tiny smile, as if he has just plucked a match from thin air. But the smile fades when he is asked to speculate on what the message might be. It seems, he replies wistfully, that we are still not quite grown up enough to know.