Our expanding universe has existed for roughly ten billion years. Even to get from Earth’s formation to our century took as long, Martin Rees remarks, as journeying half across America at the rate of one step per two thousand years. Our sun will shine steadily for five billion more years before swelling and vaporising our planet. The universe, however, will probably last for at least another hundred billion. It is widely expected to collapse eventually, but might first dilate by a factor of one followed by a million zeroes. Its expansion is possible even if it has always been infinitely large. Infinitely many galaxies, scattered across infinite space, could keep getting further apart. Infinity being a large number, some of the galaxies could well contain exact duplicates of you and me, Rees points out. With sufficiently many typing monkeys, even Hamlet would get typed many times.
The universe might perhaps expand eternally, becoming ever colder. Given enough time, strange things could happen. An electron could find its motions controlled by an equally insignificant positron tugging at it ever so weakly from ten billion light years away. (A light year is roughly ten trillion kilometres.) After a number of centuries so great that its zeroes were as many as all the atoms within reach of our telescopes, neutron stars would collapse to black holes – entities so dense that even light rays cannot overcome their gravitational pull – through their parts suddenly chancing to rush together, as quantum theory allows. Conceivably, though, only another fifty billion years separate us from a time when everything will start collapsing towards a Big Crunch. Its violence would rival the Big Bang in which the universe began – or which at least opened the cosmic period known to us.
Rees’s pages survey these and many more marvels. The material of the expanding universe is already very thinly spread. It is as if a snowflake’s atoms had dispersed through a region as large as the Earth. Some ten billion years ago, in contrast, everything detectable by our telescopes occupied a volume no bigger than a golf ball, undergoing random fluctuations which gave rise to entire galaxies later. Smaller fluctuations perhaps generated black holes tiny enough to fit inside atomic nuclei, yet each as massive as a mountain. These might nowadays be ending their ‘black hole evaporation’ (a process discovered by Stephen Hawking) in bangs detectable from two million light years away. Bigger black holes, their evaporation too slow to be detected, probably litter our galaxy in large numbers as the remnants of stellar explosions.
Other such remnants are neutron stars. Roughly the size of a town, they are so compact – at some hundred million tons to the teaspoonful – that they can rotate hundreds of times a second without flying apart. The resulting rapid pulses at radio wavelengths were jocularly referred to as ‘LGM’, meaning ‘little green men’, before their true nature was confirmed. The stellar explosions themselves, ‘supernovas’, scatter the products of stellar cookery (nucleosynthesis) so that planets and persons can be made from them. Our bodies contain atoms cooked inside many stars, some not even in our galaxy.
All these wonders have been discovered with technology yet more wondrous. A century and a half ago, Auguste Comte cited the composition of the stars as something which would never be known. Astronomers now determine it by analysing starlight. Helium was discovered by this means before being recognised on Earth. The sky’s radio waves, too, are remarkably informative. Linked across a continent, radio telescopes act like a single gigantic dish. Each operates with superb efficiency. All the energy it will ever have collected for astronomers to analyse is less than is needed for picking up a cigarette. Gravitational waves have been detected through the star movements they cause, slower than the hour hand of a watch. Even neutron star ‘starquakes’ of a few micrometres (thousandths of a millimetre) have been detected.
Rees, the Astronomer Royal, is a combination of theoretician and observer. Only heavy contributions from theory allow micrometre-sized starquakes to be ‘seen’: deduced, that is to say, from tiny variations in a neutron star’s rotation. In the same way, we can ‘see’ a Big Bang, but not a universe which has existed eternally in a Steady State with new matter constantly being created to fill its expanding space. One of Rees’s first projects was counting distant radio sources, now thought to be regions where matter is swallowed by gigantic black holes. He found too many to fit comfortably into a Steady State universe.
He gives odds of only about ten to one in favour of a hot Big Bang, however – a sign of considerable open-mindedness, almost everyone else being much more firmly convinced. His attitude towards ‘dark matter’ is similar. Our galaxy is held together gravitationally by far more matter than is actually visible. The protons, neutrons and electrons which make stones and trees and rabbits are probably comparatively rare, over 90 per cent of the galaxy being ‘dark’ material with unknown components. Rees is a leading advocate of the idea that slow-moving ‘cold dark matter’ encouraged galaxies to form. Even though this theory is now winning, he is quick to say that the ‘hot dark matter’ of his competitors could help.
As he stresses, investigating such questions can be fruitful for physics. The highest energies physicists can yet produce are trillions of times below the Big Bang energies at which the elegance of Nature’s laws stands most fully revealed. They therefore ask astronomers for evidence left behind by the Ultimate High Energy Experiment, conducted some ten billion years ago in an exploding laboratory.
Before the Beginning is at its finest when discussing life’s place in the universe. Recent findings challenge the notion that Nature’s fundamental laws, the ones reigning at ultrahigh energies, completely dictate the physics of our low-energy surroundings. There are signs that such things as the strengths of physical forces (gravity and electromagnetism, for instance) and the masses of elementary particles (such as the proton and electron) differ randomly from one to another of vastly many regions, each stretching much further than telescopic searches can range. We can deduce that almost all such regions are probably hostile to life.
How could anybody deduce it? Humans can have direct knowledge only of the cosmic district visible to human telescopes. These can probe no farther than light or radio waves have travelled since the universe cooled enough to give them free passage. In this neck of the cosmic woods, life may not be all that rare. Rees estimates that even our own galaxy carries millions of Earth-like planets. Admittedly, failure to detect extraterrestrials suggests that most of the planets are ‘Earth-like’ only in offering suitable environments, not in actually bearing intelligent life. Life’s beginnings, or the development of powerful brains, could be so much a matter of luck that Earth may carry the galaxy’s only brainy species. But the search for extraterrestrials strikes Rees as worth every cent of its paltry budgets. Finding them would be of tremendous interest. And the laws of physics say nothing to exclude them, at least locally, for analysis of starlight shows that force strengths and elementary particle masses are identical throughout our cosmic locality.
How are matters elsewhere, though? Rees gradually became converted to the theory that our neck of the woods is unusual in being ‘fine-tuned for life’. Imagine a machine whose knobs you can twiddle to decide various characteristics of the cosmos, or at least of some huge region. Some decide the strengths of gravity, of electromagnetism and of the forces whose actions control the properties of the atomic nucleus. Others control the masses of the proton, the neutron, the electron and other particles. Still others dictate the early cosmic expansion speed, perhaps, or the amount of turbulence. Now, the signs are that the knobs would need to be positioned with extreme care if the resulting situation weren’t to be utterly hostile to life.
Turn the knobs a little inaccurately and you get a world of black holes only, or of light rays and not much else, or a world filled with destructive radiation, or collapsing within a few seconds, or remaining life-excludingly hot until particles are too spread out to form galaxies, or without long-living stable stars like the Sun, or without any atoms, or with just the simplest atoms and no chemistry. Arguments in this field go well beyond the claim that life couldn’t have evolved without carbon, a claim sometimes thought unimaginative, although Rees finds it interesting. Marginally different knob positions would have led to a mainly carbon-free universe (or huge cosmic region) in which life might well never have evolved.
Why keep adding ‘or huge cosmic region’? Because, as Rees points out, what’s conventionally called ‘the universe’ could be ‘just one member of an ensemble’; there may be ‘countless others’, universes ‘in which the laws are different’; ‘other universes are a natural expectation from current cosmology’. His talk of ‘laws which are different’ is compatible with thinking that fundamental laws are everywhere the same. The idea is that these most basic laws wouldn’t control the knobs. A gigantic, cooling cosmos could have divided into domains whose knobs were turned differently, rather as a freezing pond divides into variegated ice crystals. Whether we called domains beyond the one we inhabited ‘other universes’, or just ‘other huge regions’, would then be a matter of taste.
Physicists talk not of ‘knob-twiddling’ but of ‘varied symmetry-breaking’. They note that symmetries might have been broken (knobs turned) identically for as far as human telescopes can probe. This is because initially minuscule regions, each filled with symmetries broken in a way peculiar to itself, could have ‘inflated’ enormously at early instants.
Inflation at early instants was first proposed as a means of eliminating any need to tune the early cosmic expansion speed. The speed had seemed to require tuning with immense accuracy if galaxies were ever to form. Rees compares the feat to standing at the bottom of a well and throwing a stone so delicately that it stops rising exactly at the top. If the proto-galactic clouds had expanded marginally faster, no stars would have formed inside them; if marginally more slowly, then they would have collapsed to black holes. Inflation, people hoped, would automatically have resulted in precisely the right speed. Many now argue, however, that getting inflation to work would itself demand extremely accurate tuning. Yet they still often favour the inflationary scenario because it provides so many cosmic domains, universes’, whose knobs could have been twiddled randomly. At least a few such domains might then be expected to be life-permitting.
Alternatively, new universes might be born inside black holes. Itself perhaps born in this way, ours might continually be giving birth to others. Because of the peculiarities of Einsteinian space, these could grow from the black holes of their parent without exploding into it. (Think of bubblegum swelling from a teenager’s lips without pushing them apart.) Or an eternally inflating Large-U Universe could be constantly sprouting new domains – small-u universes, necks of the cosmic woods – in which inflation had slowed to life-permitting speeds.
Many further mechanisms have been proposed for creating multiple universes. Would this all be irrefutable guesswork? Not so. Already, some multi-universe theories have been refuted on grounds of physics. Already, too, there are indications that at least some force strengths and particle masses may not be independently variable. Many dream that every force and mass will eventually be found to have been dictated by some Fundamental Theory. Yet Rees is far from alone in thinking our universe ‘fine-tuned for life’. An attractive suggestion is then that there exist vastly many universes with very varied characteristics. Unsurprisingly, we find ourselves in one of the perhaps extremely rare universes which are life-permitting.
As Rees explains, we can here use what the cosmologist Brandon Carter calls ‘the anthropic principle’. This reminds us that living beings exist only in life-permitting circumstances, so setting up observational selection. If your net has a wide mesh, don’t expect to observe tiny fish in it. If your body is made largely of water, don’t be surprised to find yourself on a star-warmed planet.
Applying the anthropic principle to multiple universes would be elegantly explanatory, Rees insists. Every bit as explanatory as the ‘superstrings’ on which so many theoretical physicists are working – entities a hundred million trillion times smaller than atomic nuclei. We cannot genuinely see other universes, but only in a very stretched sense has anyone ‘seen’ even an atomic nucleus. In that stretched sense, mayn’t we now have ‘seen’ multiple universes? Mayn’t we have detected them indirectly, as what would tidily explain the fine tuning? Or are we perhaps seeing a Fine Tuner? During his brief Foreword to Rees’s book, Stephen Hawking remarks that its author ‘has always worked closely with the observations’ so that the book unfolds without introducing the word ‘God’; after all, this word stands for ‘a theoretical concept’. Still, I think Rees ought at least to have mentioned that many scientists regard divine fine tuning as no more wildly speculative than superstrings, or than his multiple universes.
Once intelligent life has evolved, what are its long-term prospects? Rees refers to Freeman Dyson’s idea that it might survive eternally in an ever-expanding, ever-cooling universe, since intelligent information processing demands less and less energy as temperatures fall. Again, he says, John Barrow and Frank Tipler have argued for infinite information processing (and therefore infinite ‘subjective time’ for our appropriately odified descendants) in the final moments of a Big Crunch. Mental acrobatics could take ever tinier fractions of a microsecond as the collapsing cosmos grew hotter and hotter.
Such gloriously prolonged futures would never materialise if we succumbed to a pollution crisis, say, or to biological warfare. Can we be confident, Rees asks, that there is less than a 10 per cent chance of human extinction in the coming century? And if intelligent life were wiped from our planet, mightn’t this be the extinction of all the intelligence which would ever have evolved within reach of our telescopes? In his eyes, this consideration supplies strong grounds for learning how to send humans far away from an endangered Earth. While it offers slim pickings scientifically, manned space flight is a crucial ‘insurance policy’.
Is intelligence indeed a rarity, cosmically speaking? Rees is impressed by another point made by Brandon Carter. The time intelligent life took to evolve on our planet was very roughly equal to the total time available: why? (A few billion years to get to intelligence, and only another few billion before the Sun swells and Earth fries. Why are those periods as nearly equal as they are?) Carter reasons that intelligence would typically need much longer to evolve. It is as if hugely many convicts had struggled to open the combination locks on their cells by random dial-twisting. Average time required: many days. Time available: a couple of minutes. We might then expect only one or two to have succeeded, in about a minute.
This reasoning is highly controversial. It is curious, then, that Rees is much more inclined to doubt another of Carter’s arguments, the ‘doomsday argument’ which Carter himself regards as straightforward. In a nutshell: one shouldn’t expect to have been born extremely early, for instance in the earliest billionth, among all humans who would ever have lived. This could be like expecting one’s name to have been drawn very quickly from a hat containing several billion others. Now, mayn’t this suggest that the human race won’t last long enough to colonise its galaxy? For if it did, wouldn’t you and I have been astonishingly early members of it?
This sketch of the doomsday argument leaves out something which Carter insists on, but which is missing from Rees’s book. Your grounds for thinking that a hat contained many names might be very powerful. If so, couldn’t they survive the discovery that your name had been drawn quickly? Consider all humans who will ever have lived. What’s your chance of being in the earliest 5 per cent? Carter’s reply is not that it must be 5 per cent, which is what Rees portrays him as saying.
Again, the number of names in a hat is something definite. In contrast, the number of humans who will ever have lived may depend, as Rees notes, on facts which have yet to be created – the fact, for instance, of somebody’s freely pushing a nuclear warfare button. This means, or so I argued in a recent book, that Carter’s doomsday argument acts strongly only against people confident in a long future for humans. But even so, can’t the argument be rather important?
Apart from pollution and war, what could lead to human extinction? About fifteen years ago, Rees investigated a possibility to which his book returns. Our cooling universe could have become ‘supercooled’. Like a statue balancing upright, it would then be unstable against a violent push. Experiments at ultra-high energies might supply the push. Things would at first go awry only inside a tiny bubble, but this would expand at nearly the speed of light. Earth would be destroyed. Then the rest of the galaxy. Then the next galaxy, and so on. The ultra-high energies in question are, Rees calculates, hundreds of times above those attainable even by the Large Hadron Collider now being built. Still, he comments, ‘caution should surely be urged (if not enforced) on experiments that create energy concentrations that may never have occurred naturally.’ This may be the most important point in Rees’s important book.
Lee Smolin’s The Life of the Cosmos insists that our cosmos is unified in a way once thought unique to living beings. Einstein’s general relativity is Machian, which means (to use a traditional way of making the point) that it attributes a car’s difficulties in accelerating to the presence of all the stars. Again, gauge theory (essential to modern physics) treats the properties of different kinds of particle as interdependent. And quantum theory reveals that once two particles have interacted they typically display a dramatic ‘wholeness’; their natures remain ‘entangled’ no matter how far apart they move. By itself, this theme might have justified the book’s title.
Smolin’s main theme, however, is something different, and it is this which has captured people’s attention. Joining the many who now see our universe as ‘fine-tuned’ in ways making life possible, he offers a highly original explanation. Like Rees, Smolin has many and varied universes, yet his vary in inherited ways. Universes give birth to more universes, the differences between child and parent being only small. Large ones can arise only through many generations of slow evolution. The word ‘evolution’ feels right because universes undergo ‘cosmological natural selection’. Like the organisms Darwin described, universes with appropriate properties produce more offspring. Having appropriate properties means having many black holes, which means having many stars which can become black holes, which in turn means having much carbon and oxygen. Stars, carbon and oxygen are good at producing living beings. People therefore speak of ‘fine-tuning for life’. In reality, the fine-tuning is for black holes because (this being a scenario which Rees takes fairly seriously) it is from black holes that new universes are born.
Smolin calls his theory ‘frankly speculative’. Whereas his random variations between parent and child universes have to be tiny, most cosmologists assume they would typically be huge. Again, universes born from black holes are pretty controversial. So is the usefulness of carbon and oxygen for producing the stars which become black holes. Stars are born inside giant clouds. These were at risk, Smolin thinks, of being heated so much by the earliest stars that few later ones could form. What allows them to radiate away enough heat? The answer, he suggests, is carbon monoxide and carbon dust.
Using arguments which have impressed the likes of Rees, Smolin calculates that a universe chosen randomly from the possibilities would have, at most, one chance in a billion followed by two hundred and twenty zeroes of having stars like those we see. Now, the chances of getting life and the chances of getting the stars can appear much the same, especially if we buy Smolin’s theory that tuning for stars involves tuning for carbon and oxygen. His ‘cosmological natural selection’ can thus seem a viable alternative to ‘anthropic’ observational selection of whichever universes produce living beings.
Smolin deserves great credit for this; he has pioneered a new way of thinking. Yet his book suffers from its aggressive efforts to destroy its ‘anthropic’ competitors. He attacks anthropic reasoning on two grounds. First, it ‘leads to no testable predictions’. Second, it offers ‘to explain almost anything’, so ends up explaining nothing. But these seem very weak grounds.
In point of fact, anthropic reasoning encourages many predictions. It encourages us to believe in many and varied universes – and hence to predict that at least one mechanism for producing them will be found plausible despite advances in physics. Suppose physicists found some Fundamental Theory dictating that universes all had to be much the same. This would be a blow to the anthropic approach, wouldn’t it? So would studies suggesting that life would evolve in more or less any universe, no matter what its force strengths and particle masses. So, too, would proof that early cosmic inflation (which, as Rees says, most cosmologists now take very seriously) couldn’t have occurred. For it would then follow that physical force strengths and elementary particle masses, if settled by chance, would have been settled in many different ways inside the region probed by our telescopes. Yet we know for sure that they weren’t.
Again, users of anthropic reasoning don’t ‘offer to explain almost anything’. Smolin’s complaint is that one could find, among all the varied universes which such folk believe in, many which were quite as unlikely as the ‘fine-tuned universes’ that contain living beings. To accept the entire lot ‘is simply to give up looking for a rational explanation’, he protests. His thought appears to be that, no matter what oddities were observed, lovers of the anthropic principle would just comment, ‘When universes come in such variety, don’t be surprised that some are like this.’
In reality, however, these people think that what’s observed can be explained by variety supplemented by observational selection. Our universe appears fine-tuned for permitting living beings to evolve. Can we avoid believing in divine selection of its properties, or in Smolin’s ‘cosmological natural selection’? Believing in a million billion other universes couldn’t help explain what we see, if we could equally well have seen just any of the others. But anthropic reasoning stresses that nobody could have seen the life-excluding universes. To explain is to render unmysterious. The anthropic approach need not involve supposing that our universe was from its first moments certain, or almost certain, to become life-permitting. Chance could have played a crucial role here (in particular in the form of random symmetry-breaking or ‘knob-twiddling’). But rendering unmysterious isn’t the same as showing to be fated, or almost fated. Why is the observed fish longer than a centimetre? You answer that your net can’t hold anything shorter. That’s not saying that catching a fish had been inevitable or nearly inevitable.
Smolin’s aggressiveness is particularly unfortunate since his theory can appear to need help from anthropic reasoning. The trouble is that there seems a better way of generating black holes. Just have more turbulence. Cosmologists face a tough problem in explaining why the early universe was so smooth. Why not a chaos full of regions of extreme, black-hole-generating compression? An attractive answer is that while hardly any universes begin smoothly, only those which do can give rise to observers.
Smolin does grant respectability to some answers of this sort, yet only very grudgingly. We mustn’t, he insists, give them very much work to do, lest we start offering to explain just anything. We must employ them only when considering universes which contain a few more black holes than ours, ‘a modest increase’. These possible universes don’t destroy his theory, he suggests, if no observers could evolve in them. But how about the criticism that tremendous early turbulence, very easily achieved, would have produced not ‘a modest increase’ but tremendously many black holes? He lays himself wide open to this criticism.
He tentatively considers the following reply. Although early cosmic inflation smoothed away the turbulence, it produced a universe gigantic enough to beat turbulent ones at the game of generating black holes. His preferred reply, however, is that he needn’t prove that our universe beats all comers. He offers a Darwinian analogy. An eagle can be maximally fit for giving birth to more eagles; that doesn’t mean it has to produce more offspring than a dandelion. There can be peaks of many different heights in a Darwinian ‘fitness landscape’.
Regrettably, the analogy appears to fail. Eagles aren’t competing with dandelions, yet different universes would compete in Smolin’s quasi-Darwinian scheme of things. The sort producing most offspring would win. There would soon be hardly any others, proportionately. It would be as if the many dandelions were crowding out the eagles. So for Smolin to save his theory, shouldn’t he perhaps be saying that only the eagles could ever be observed?