The science of eternity
This century may be a defining moment for the cosmos. If humans do not destroy themselves they may spread beyond the earth into a universe that could last almost forever. Life would have tunnelled through its moment of maximum jeopardy

Martin Rees
Over the past few centuries, the earth has aged spectacularly. Its creation has been moved back from 6pm on Saturday, 22nd October, 4004 BC, as calculated by the 17th-century scholar and Archbishop of Armagh, James Ussher, to a time and date some 4.5 billion years earlier. The story of life has been stretched back almost as far, and the story of complex, multicellular life-forms-relative newcomers-has itself been almost a billion years in the making. As a result, the way we see the world has changed profoundly. Not only can we now have some sense of the millions of years it takes to raise and then level mountains, or to open and then close oceans, we also have the clearest evidence of humanity's absence throughout those ages. To Ussher's mind, the creation of the world and the creation of humanity were within a week of one another; to our modern minds, the two events are unimaginably far apart. There was a vast absence before us, a physical and biological world untouched by introspection, and its record stares out at us from every rock.

If the earth's past has been stretched, what of its future? To those of Ussher's faith, the end of the world was a certainty and to some of his contemporaries history was already nearing its close. Sir Thomas Browne wrote, "the world itself seems in the wane. A greater part of Time is spun than is to come."

To look forward, we must turn from geology to cosmology. Current cosmology suggests a future that, if not infinite, dwarfs the past as much as the depths of time we now accept dwarf Ussher's exquisite estimates. What it cannot tell us, though, is whether these vast expanses of time will be filled with life, or as empty as the earth's first sterile seas. In the aeons that lie ahead, life could spread through the entire galaxy, even beyond it-and outlast it too. But life could also snuff itself out, leaving an eternity as empty as the space between the stars.

getting very hot

To begin at the beginning, we can, with some confidence, trace cosmic history back to its first few seconds, some 12 billion years ago. But in response to the fundamental question, "what happened before the beginning?" we cannot do much better than St Augustine in the 5th century, who sidestepped the issue by arguing that time itself was created with the universe. The origin of the "big bang" is, in some ways, as mysterious to us as it was to St Augustine. Cosmologists who study it are forced to jettison commonsense ideas, invoking extra spatial dimensions and postulating that space and time may have an inherently "grainy" structure on very tiny scales.

The sun-and the earth with it-came into being when the universe was approximately two thirds its current age. Minuscule though they may be in terms of their size, on a cosmic scale, they are thus of a respectable age. The sun will continue to shine for five billion years-longer than it has taken for our earth to evolve from a lump of molten rock to its present state. It will then swell up into a red giant, engulfing the inner planets and vaporising any life that remains on earth.

It is hard to say what life might remain on the earth at that point. The sun has been getting hotter since its birth and will continue to do so. This heating has been balanced on earth, to some extent, by trends that have tended to cool the planet; the most clear cut of which, over geological timescales, has been a long-term drop in carbon dioxide and a resultant weakening of the earth's natural greenhouse effect. This trend, however, cannot continue for long; carbon dioxide, once more plentiful in the atmosphere than oxygen, is now present at a mere 300 parts per million. Even a drop to zero would not cool the planet a great deal in the face of increased solar luminosity-but it would kill off the plants which rely on carbon dioxide as the raw material for their photosynthetic growth, and thus remove oxygen from the atmosphere.

Without the cushion provided by ever-lower carbon dioxide levels, over the next billion years the earth will start to feel the warming of the sun much more than it has in the past, and when the surface reaches about 50ºC, the increased levels of water vapour will permit a new form of "runaway" greenhouse effect that will quickly raise the temperature high enough to boil the remaining oceans. Current estimates suggest that the biosphere cannot survive much beyond a billion years.

So by the time the sun finally licks the earth's face clean, life on earth will either be extinct, spread beyond its original planet or in a form impervious to extremes of temperature. Just as the universe is still young, so it seems that the emergence of intelligence and complexity is near its cosmic beginnings; we are far from the culmination of evolution. Intelligently-controlled modifications could lead to faster and more dramatic changes than Darwinian natural selection allows. The future may lie in artefacts created by us and in some way descended from us, that develop via their own directed intelligence. Such entities might see the death of the sun as a minor or sentimental matter.

In the 1960s, Arthur C Clarke imagined the "Long Twilight" after the death of the sun and today's other hot stars as a realm at once majestic and slightly wistful. "It will be a history illuminated only by the reds and infrareds of dully glowing stars that would be almost invisible to our eyes; yet the sombre hues of that all-but-eternal universe may be full of colour and beauty to whatever strange beings have adapted to it. They will know that before them lie, not the... billions of years that span the past lives of the stars, but years to be counted literally in trillions. They will have time enough, in those endless aeons, to attempt all things and to gather all knowledge. But for all that, they may envy us, basking in the bright afterglow of creation; for we knew the universe when it was young."

from big crunch to eternal expansion

But even after the longest twilight, night will fall. There will come a time when the dimmest, slowest-burning stars are done. While academic cosmologists publish, month after month, hundreds of scientific papers discussing the ultra-early universe, they have written little about this long-range future. But it is an area ripe for speculation. I can claim to have made one of the first scientific contributions to "cosmic futurology" in a short 1968 paper entitled "The Collapse of the Universe: an Eschatological Study." Many cosmologists suspected then that the expansion that currently characterises our universe would cease and reverse itself. Galaxies would then fall towards each other, eventually crashing together into a "Big Crunch." I described how, as galaxies merged together in the countdown to the crunch, individual stars would accelerate to almost the speed of light (rather as the atoms speed up in a gas that is compressed). Eventually these stars would be destroyed as the blue-shifted radiation from other stars rushing towards them made the sky above them hotter than the fires within.

Currently, though, the big crunch is out of favour; more recent long-range cosmic forecasts have predicted that the expansion of the universe will continue for ever, with its contents becoming ever colder and more diffuse. Ten years after my paper, the Princeton theorist Freeman Dyson-who would not countenance the Big Crunch because it "gave him a feeling of claustrophobia"-made scientific eschatology more respectable in an influential article called "Time without End: Physics and Biology in an Open Universe," published in the austere scholarly journal Reviews of Modern Physics. "The study of the remote future," Dyson wrote, having noted that the handful of papers on the subject were written in an apologetic or jocular style, "still seems to be as disreputable today as the study of the remote past was 30 years ago." He set out to change the state of affairs with a rigorous study of the physics of the far future and the prospects for some sort of life persisting there.

Dyson charted the processes that would take place after the suns had burned out, and their timescales; the loss of planets from dead stars, the slow evaporation of galaxies, the decay of black holes (a process that takes 1060 years-that is one with 66 zeros behind it for a stellar-mass hole, and longer for a giant one) and the eventual transmutation of all remaining matter into iron. That last alchemical endgame is spun out for such a large number of years that to write it down would need as many zeros as there are atoms in all the galaxies we can see. That is how long you would have to wait before a giant quantum fluctuation caused an entire star to "tunnel" into a black hole. Throughout it all, Dyson imagined, life might persist in some form.

The universe's usable energy reserves are finite, and at first sight this might seem to be a basic restriction on everlasting life. But Dyson showed that this constraint was not fatal. Just as an infinite series can have a finite sum (for instance 1 + H + G + ... = 2) so there is no limit to the amount of information processing that could be achieved with a finite expenditure of energy. And the nature of mathematics ensures that there could be infinite novelty generated in that infinite information processing. It might be an odd, stripped down sort of "life," a life abstracted to such an extent that only the most mathematical of minds could conceive it as life-but it would not have to be repetitive, and thus might not have to be dull. The rate at which the calculations were made would get slower and slower and enforced periods of hibernation longer and longer-as Woody Allen said, "eternity is very long, especially toward the end"-but life in some form could persist.

what freeman dyson didn't know

Recent developments in physics have modified Dyson's picture in two ways. First, we now suspect that atoms themselves live for "only" 1036 years-not for ever. In consequence, the cold remnants of stars and planets (and any complex entities made of atoms) will erode away as the atoms within them decay. Thoughts and memories would only survive beyond this stage if downloaded into circuits and magnetic fields in clouds of electrons and positrons-maybe something that would resemble the alien intelligence in The Black Cloud, the most imaginative of Fred Hoyle's science fiction novels of the 1950s.

Second, Dyson saw no limit to the scale of artefacts that could some day be constructed. He envisioned the observable universe getting ever vaster, as the expansion of space slowed down and galaxies whose light has not yet had time to reach us (galaxies further away in space than the big bang is in time) slowly came within range of possible communication and "networking," offering scope for ever larger cosmic construction projects. But within the past few years, cosmologists have discovered to their surprise that the expansion of the universe is not slowing down at all: some repulsive force or antigravity seems to be pushing galaxies apart at an accelerating rate. This means that, rather than more and more galaxies coming into contact with us, we will lose contact with most of the galaxies we can now see: these galaxies will accelerate towards the speed of light, eventually receding "over the horizon" into the unobservable. In the very distant future, our descendants (if any) would have lost contact with everything except the small group of galaxies that contains our own milky way (by then bereft of the starlight that makes it milky). There would therefore be a permanent limit on how large and complex anything can get. Some theorists suggest that this limit on the energy and complexity available would preclude infinite novelty and that eternity would become very dull.

The implications of accelerating expansion, though, need to be put in the context of the other uncertainties. First, we cannot be sure that the regions beyond our present horizon are like the parts of the universe we see, and this might influence future expansion. For another thing, there is a theoretical possibility that the nature of the universe, and the laws that hold sway in it, might be subject to radical and unexpected change. Very pure water can stay liquid below 0ºC-a state known as supercooling-only to freeze in a flash when a snowflake falls on it. It has been suggested that our present "vacuum"-the fundamental stuff of space time-may also be "supercooled," and thus that some comparatively minor event could trigger a change to a new state of being, governed by quite different laws. There are occasional scares that this sort of catastrophe could be induced artificially by experiments that crash particles together with high energy. Those in charge of the Brookhaven national laboratory in the US were cautious enough to commission an expert assessment on whether any of their planned experiments could end the universe this way or trigger some other global catastrophe. (Another scare was that a new kind of atom called "strange matter" could be created in accelerators, which would by contagion gradually transform the entire earth into this new form-a scenario reminiscent of Kurt Vonnegut's Ice Nine, where the oceans are transformed into a new kind of ice with a high freezing point.) It is reassuring to note that "cosmic ray" particles, hurtling through space with far more energy than any terrestrial experiment can generate, have been hitting each other, on and off, for billions of years, without tearing the fabric of space or destroying the possibility of matter.

the future of time travel

To see any of these wonders far ahead in time would require a time machine-something like that contraption of "black and brass" imagined by HG Wells, just over a century ago. As Wells's chrononaut gently eased the throttle forward, "night came like the turning out of a light, and in another moment came tomorrow." As he sped up "the palpitations of night and day merged into one continuous greyness." Eventually, after some unpleasantness with the Eloi and the Morlocks, he ends up 30m years hence, in a world where all familiar forms of life have become extinct. He then returns to the present, bringing strange plants as evidence of his trip.

Time travel clashes less with current scientific concepts than it did with those of Wells' readers at the end of the 19th century. Einstein has taught us that time is not "absolute." If you move almost as fast as light, or if you are exposed to strong gravity near a black hole, time is stretched. This may seem counter-intuitive. But our intuitions are based on a constricted range of experience. Few of us have travelled faster than a millionth of the speed of light (the speed of a jet airliner) and we live on a planet where gravity is far weaker than it is near black holes. But time travel still takes place, if only on a small scale. Richard Gott, another Princeton theorist, calculates that the American astronaut Story Musgrave is a millisecond younger than he would be if he had not travelled repeatedly into space. The designers of the extremely accurate clocks at the heart of the GPS satellite navigation system have had to take into account the fact that the clocks on their satellites measure time at a slightly different rate to clocks on the earth.

Time travel into the far future violates no fundamental physical laws. A spaceship that could travel at 99.99 per cent of the speed of light would allow its crew to "fast forward" into the future. An astronaut who managed to navigate into the closest possible orbit around a rapidly-spinning black hole without falling in could, in a subjectively short period, view an immensely long future time span in the external universe. Such adventures may be technically challenging, but they are not impossible. Nor, rather more surprisingly, is travel into the past-as far as we know.

Time, as Wells and his chrononaut knew, is a fourth dimension, but it is different from the three dimensions that define space. In three-dimensional space we can move to the left as easily as to the right, backward as well as forward, up as well as down. But we seem to be dragged forward in time willy-nilly. In some ways this seems like a good thing, since returning to the past involves the risk of changing it in such a way that makes history, or the universe itself, internally inconsistent. Science fiction writers like Isaac Asimov and Poul Anderson have imagined busy time- police trying to keep reality straight, stopping people from murdering their grandmother, or engineering the defeat of Charles Martel at the battle of Tours, or whatever. Physicists, less in need of plot lines, have preferred to have such things banned by the laws of nature rather than the laws of man. But insisting that time travel cannot change the past is not the same as saying that time travel cannot happen.

More than 50 years ago, the great logician Kurt Gdel (Princeton again) invented a bizarre hypothetical universe, consistent with Einstein's theory, that allowed "time loops," in which events in the future "cause" events in the past that then "cause" their own causes, introducing a lot of weirdness to the world but no contradictions. (The film, The Terminator, in which a son sends his father back in time to save-and inseminate-his mother, wonderfully combines the insights of the greatest Austrian-American mind, Gdel, with the talents of the greatest Austrian-American body, Arnold Schwarzenegger.) Several later theorists have used Einstein's general theory of relativity (which tells you how to change the shape of spacetime) to design "time machines" that might create temporal loops. But these machines are not things you can build in a Victorian basement. Some of them need to be of effectively infinite length; others need vast amounts of energy.

Despite the remoteness of these concepts from realistic technology, Gdel's discovery raised a fundamental question: is time-travel into the past precluded by a "chronology protection law?" The science-fiction writer Larry Niven believes that time travel would create a chronology protection law; if time travel is allowed, time travellers will endlessly change the universe until one of them changes something so basic that time travel becomes impossible, at which point the universe stabilises with a new set of laws that rule out time travel. One piece of evidence for the protection of chronology sometimes put forward is that time travel is not, in fact, observed: "tourists from the future" don't swell the crowds at historical events, as far as we can tell. But this may tell us only that no time machine has yet been made, not that it is impossible. Even theories that allow time machines suggest that they have limitations: a time machine could not enable its users to travel back prior to the date of its construction.

If there is no absolute chronological protection, perhaps there is some subtler means whereby, while time travel is possible, it never changes what has already happened. This would restrict the time traveller's free will, but that is nothing new. Physics already constrains us: we cannot exercise our free will by walking on the ceiling. Another option is that time-travellers would shift into a parallel universe, where events played out differently rather than repeating themselves exactly, as in the film Groundhog Day. Perhaps in the sunless future, as galaxies fade over the cosmic horizon like astronauts sinking into black holes, it will be through time travel to parallel universes that our descendants stave off the ennui of eternity. As the long day wanes and the slow moons climb, they will make off for newer worlds, endlessly seeking a perfect past.

A dangerous moment for life

A hackneyed anecdote among astronomy lecturers describes a worried questioner asking: "how long did you say it would be before the sun burnt the earth to a crisp?" On receiving the answer, "five billion years," the questioner responds with relief: "thank God for that, I thought you said five million." What happens in far-future aeons may seem blazingly irrelevant to the practicalities of our lives. But I don't think it is. It is widely acknowledged that the Apollo programme's pictures of the island earth, its fragile beauty contrasted with the stark moonscape, changed the way we see ourselves in space-strengthening the collective ties that bind us to our environment. No new facts were added to the debate; just a new perspective. A new perspective on how we see ourselves in time might do something similar. (That is the hope of the Long Now Foundation, an organisation devoted to the construction of clocks and other instruments that can last thousands of years.)

Taking the truly cosmic long view, the view in which billions of years pass as hours, would require pictures of the earth, not from the neighbouring moon, but from light years away, from some far-off star. Within a few decades, new generations of space telescopes should provide the technology to take just such pictures-to produce images of the planets around other stars. Will they have biospheres? Will they harbour intelligent life? We know too little about how life began, and how it evolves, to be able to say whether alien intelligence is likely or not. The cosmos could already be teeming with life: if so, nothing that happens on earth would make much difference to life's long-range cosmic future. On the other hand, advanced life may be rare-so rare, perhaps, that it is unique to our earth. The emergence of intelligence may require such an improbable chain of events that it has not occurred anywhere else-not around even one of the trillion billion other stars within range of our telescopes. Claims that advanced life is widespread must confront the famous question first posed by the physicist, Enrico Fermi: "why aren't the aliens here?" Why haven't they visited earth already, or at least manifested their existence in some way? This argument gains extra weight when we realise that some stars similar in many ways to our sun are billions of years older: if life were common, its emergence should have had a head start on planets around these ancient stars.

But if earth is the unique abode of intelligence, the fate of our planet could have an importance that is truly cosmic: it might conceivably be the difference between a near eternity filled with ever more complex and subtle forms of life and one filled with nothing but base matter. Yet this fate is finely balanced. One does not need to believe that the Brookhaven national laboratory is a threat to believe that technological civilisation carries risks that get greater with every century. The 20th century brought us the bomb; the 21st offers the more far-reaching threats of catastrophically replicating rogue nano-machines and engineered diseases. The latter are the biggest worry. Within a few years, thousands-even millions-of individuals may acquire the capability to make and disseminate weapons that could cause widespread epidemics. An organised network of al Qaida type terrorists would not be required: just one fanatic or weirdo with the mindset of those who now design computer viruses. In a few decades, such individuals may be able to trigger truly global catastrophes. Even if all nations applied a strict precautionary principle to dangerous procedures, the chances of effective worldwide enforcement are small. There is also, of course, the possibility of accidents in the most respectable of institutions.

Pessimism on these matters seems the rational stance. But technology also offers hope. Perhaps, by the end of the century, self-sustaining communities will have been established away from the earth-on the moon, on Mars, or freely floating in space. Although it will be little consolation to those on earth, life's long-term cosmic potential could thereafter not be quenched by any terrestrial catastrophe: life would have "tunnelled through" its era of maximum jeopardy. This is the best reason for prioritising programmes of manned, rather than unmanned, spaceflight. (New technologies may offer another option too: downloading our blueprint into inorganic memories that could be launched into space.)

The most crucial location in space and time (apart from the big bang itself) could be here and now. This new century, on this planet, may be a defining moment for the cosmos. Our actions could initiate the irreversible spread of life beyond the solar system. Or, in contrast, through malign intent, or through misadventure, 21st-century technology could jeopardise life's cosmic potential when its evolution has still barely begun.
The author is the astronomer royal and fellow of King's College, Cambridge