Chapter 1
Two Paradoxes and an Equation
Life got a grip on Earth with almost indecent haste. When our planet was young, it was bombarded by debris left over from the formation of the Solar System, creating a hostile environment in which life could not get a grip. This bombardment also scarred the Moon. Studies of lunar cratering and other evidence tells us that this bombardment tailed off about 3.9 billion years ago, some 600 million years after the formation of the Solar System. But there is evidence that as soon as the bombardment ended, life began.
The oldest evidence is not of life itself, but of a characteristic signature of life. Carbon atoms come in several varieties, called isotopes, which have the same chemical properties but different weights. The most common stable form is called carbon-12, but there is another, slightly heavier, stable form called carbon-13. Living things prefer to take up carbon-12 from their environment, so they produce an excess of the lighter isotope compared with their surroundings. Ancient rocks from Greenland, just over 3.8 billion years old, contain exactly this isotopic signature of life. This suggests that biological processes were going on on Earth as soon as the bombardment ended; the most plausible explanation for this is that the bombardment carried with it the seeds of life, brewed up from the chemical cocktail known to exist in the clouds of gas and dust from which stars form.
Apart from the precursors to life detected in these clouds by spectroscopy, both amino acids and sugars have been found in meteorites and in the dust trails from meteors burning up in the Earth’s atmosphere today. Confirmation of the origin of these complex organic molecules has come from NASA experiments in which amino acids were synthesized under conditions mimicking those that exist in dense interstellar clouds. These organic molecules include glycine, alanine and serine, which are basic parts of protein. And in 2009, a team of NASA scientists announced that they had found glycine in material returned to Earth by the spaceprobe Stardust from comet Wild 2. This was the first confirmed detection of an amino acid in space. Such material must have rained down on the young Earth in profusion at the end of the early bombardment; as a spokesman for the Stardust team put it, ‘our discovery supports the theory that some of life’s ingredients formed in space and were delivered to Earth long ago by meteorite and comet impacts.’
Even so, when we talk about life most people want direct evidence of living creatures – fossils. The earliest known fossils are the remains of colonies of bacteria known as stromatolites. These are found in rocks as old as 3.6 billion years, laid down less than a billion years after the formation of the Solar System, and less than 300 million years after the end of the early bombardment. Stromatolites are not only direct evidence of early life; they are also evidence that by that time there was already a complex ecosystem, with many different kinds of microbes living alongside each other and interacting with one another. Clearly, life itself must have got started even earlier than 3.6 billion years ago, as the isotope evidence implies.
But the stromatolites also highlight one of the most important features of life on Earth. The chemistry of life always takes place in a special environment, protected from the outside world. That special place is the cell, a tiny bag of watery liquid containing all of the requirements for life. Within the cell, DNA and RNA can go about their business of making copies of themselves (reproduction) and providing the instructions for the manufacture of protein molecules. This is the essential chemistry of life. But it has to be contained in a secure environment.
The best way to appreciate the importance of the cell is to look at the role of enzymes, the proteins that encourage the essential chemical reactions of life to take place. Enzymes are not particularly rugged molecules. If they get too hot or too cold, they fall apart. If their surroundings are too acid or too alkaline, they fall apart. If they fall apart, they can no longer do their job, and life stops. So they have to operate inside a protective wall, a special kind of wall which allows some molecules in but keeps others out, and which allows some molecules out but keeps others in. This wall is called a semi-permeable membrane, and it is the wall that surrounds the bubble of a cell. One of the defining characteristics of life – perhaps the defining characteristic – is that the region where life processes go on inside the cell is not in chemical equilibrium with its surroundings. Equilibrium equals death. Life maintains itself in a non-equilibrium state. The American biologist Lynn Margulis sums this up by saying ‘life is a self-bounded system.’
This has deep implications for our understanding of the origin of life on Earth. It is now widely accepted that a rain of meteorites and comets delivered the basic components of life to Earth at the end of the early bombardment. Most suggestions about how the transition from non-life to life occurred involve the production of the molecules of life first (DNA, RNA and proteins, though not necessarily in that order), followed by the ‘invention’ of the cell. One popular idea is that complex compounds were concentrated in a thin layer of material, either trapped in a layer of clay or spread across a surface, and chemistry did the rest. Another is that the crucial chemical processes took place in a hot, chemically rich environment like the ones found today in deep-sea vents, where superhot water is produced by volcanic activity. There are other variations on the theme, all of which hark back to Charles Darwin’s speculation, in a letter he wrote to Joseph Hooker in 1871, that:
We could conceive in some warm little pond, with all sorts of ammonia and phosphoric salts, light, heat, electricity, etc., present that a protein compound was chemically formed, ready to undergo still more complex changes.
But what would happen to such a compound? It would be more likely to be washed away, or destroyed, than to combine with other complex molecules and do something interesting. The interesting things that could lead to life would only take place in a sheltered environment. Where better than inside a cell? To my mind, it is far more likely that cells came first, in the form of bubbles made of semi-permeable membrane, and were taken over by life. And there is strong evidence to support this contention.
Researchers at NASA’s Ames Center carried out experiments at the end of the twentieth century in which vacuum-sealed chambers about the size of shoe boxes were cooled to 10 degrees above the absolute zero of temperature, equivalent to −263 °C. A mixture of water, methane, ammonia and carbon dioxide in the chamber was allowed to freeze onto pieces of aluminium or caesium dioxide in the chambers, simulating the way ices form on dust grains in interstellar clouds. Then, the mixture of molecules was bathed in ultraviolet radiation, simulating the radiation from young stars. You will not be surprised to learn that the result was the production of a whole slew of organic molecules, including alcohols, ketones, aldehydes, and very large molecules with as many as forty carbon bonds linking their atoms. The news soon attracted the attention of another researcher, David Deamer, of the University of California, Santa Cruz. Years before, in the 1980s, Deamer had caused a sensation among astrobiologists with his studies of a piece of rock from space called the Murchison meteorite, which landed in Australia in 1969. Searching for signs of organic material, Deamer had ground some of the rock from the meteorite to dust, and washed it over with water to rinse out any organic molecules. To his astonishment, he found hundreds of microscopic globules floating in the water, each one made up of a double ‘skin’, like little balloons. And when he took some of the frozen material from the Ames experiment and placed it in warm water, Deamer found exactly the same kind of balloons, or bubbles, called vesicles. They were between 10 and 40 micrometres across, about the same size as red blood cells, and essentially indistinguishable from the vesicles obtained from the Murchison meteorite. They were like cells, but without the chemistry of life.
Further research revealed what was going on to make the vesicles. Some of the more complex molecules produced by the action of ultraviolet light on the ice grains, both in the lab (for sure) and in space (presumably), are members of a family known as lipids, with a distinct ‘head’ and ‘tail’ structure, like a tiny tadpole. The head end of the molecule is attracted to water, but the tail end is repelled by water. When such molecules are put in water, they naturally form a double layer, with the head outwards and the tails inwards. And these double-layered ‘walls’ promptly curl up into tiny balls. This must have happened in the warm waters of the young Earth, trapping things like amino acids and sugars inside the vesicles, in a contained environment where it was possible for the processes that led to life to take place. Without those barriers, the important molecules of life would have been so diluted in the ocean that no interesting chemistry would have taken place. The icing on the cake is that given the raw materials, little balls like this grow, by inserting more lipids into the skin of the bubble, and if they grow big enough they spontaneously divide into two spheres.
It’s even possible – though this is not essential to the explanation of how life got started on Earth – that vesicles exist inside comets, so that life itself, not just the precursor chemistry of life, was carried down to Earth at the end of the early bombardment. A team of researchers from NASA’s Ames laboratory (Max Bernstein, Scott Sandford and Louis Allamandola) discussed something similar in an article in Scientific American in July 1999:
An intriguing possibility is the production, within the comet itself, of [organic material] poised to take part in the life process . . . there are repeated episodes of warming for periodic comets such as Halley when they approach the Sun [and] ample time for a very rich mixture of complex organics to develop . . . It is even conceivable that liquid water might be present for short periods within the bigger comets . . . it is quite plausible that comets played a more important active role in the origin of life.
The most extreme version of this idea was developed in the 1970s and 1980s by Fred Hoyle and Chandra Wickramasinghe, who were convinced that Earth was seeded with life from space. Whether or not these speculations are correct, the important point to emphasize is that the conservative view today is that the Earth was seeded with complex organic molecules, one step from being alive, very early in its history. Life on Earth did not have to get started from scratch from a mixture of simple molecules such as carbon dioxide, water and methane.
It still may not seem likely that the step from non-life to life could occur, even within the protecting enclosure of a vesicle. But there must have been many hundreds of billions of vesicles on the young Earth, and it only had to happen once! As Darwin says, near the end of Origin of Species:
probably all the organic beings which have ever lived on this earth have descended from some one primordial form, into which life was first breathed.
This insight has since been amply confirmed by studies of the genetic material – the DNA – from many different kinds of living thing, and of the workings of the cell itself, which uses exactly the same basic chemistry (for example, to process energy) in all living things. Interestingly, this evidence also suggests that the ‘primordial cell’ was a heat-loving bacterium that may have come to life near an underwater volcanic vent. Since the young Earth was covered in water and highly active geologically, this is not very surprising and is entirely consistent with the idea that the chemistry of life developed inside a vesicle from space. It just tells us where that vesicle was when life began.
THE COSMIC LOTTERY AND THE DRAKE EQUATION
Can all of this tell us anything about the likelihood of life elsewhere in the Universe? Since we only have the example of the Earth to go by, you might think that it is impossible to draw cosmic conclusions about the existence of life elsewhere. Perhaps it is very easy for life to get started, and all Earth-like planets harbour life; or perhaps it is very difficult for life to get started, and ours is the only inhabited planet. Either possibility, and every possibility in between these extremes, is consistent with the evidence that life exists on one planet – as the statisticians say, you can’t generalize from a sample of one. Or perhaps you can. Some astronomers, notably Charles Lineweaver, of the Australian National University, argue that the speed with which life got started on Earth is an important additional piece of evidence to add to the fact that there is life here at all.
Most astronomers agree that the Moon was formed in a huge impact between a Mars-sized object and the Earth, about 100 million years after the formation of the Solar System. The energy from this impact would have melted the Earth’s crust and sterilized the planet, providing a ‘blank slate’ for the beginning of life on Earth. After that event, for about 600 million years the Earth was subjected to a heavy but decreasing bombardment from space, and it is possible that life got started several times and was wiped out several times before the bombardment ended. There is some controversial evidence that there was a final catastrophe called the ‘Late Heavy Bombardment’ before this process ended, but Lineweaver does not find any evidence to support this idea, and whether or not the Late Heavy Bombardment occurred the subsequent argument is not affected. Either way, as the long bombardment ended life began. Lineweaver makes the scientific case for the common-sense argument that if life were rare in the Universe it would be unlikely ‘that biogenesis would occur as rapidly as it seems to have occurred on Earth’.
He uses the example of a lottery to indicate how the statistics work. If a gambler buys a lottery ticket every day for three days, and loses on the first two days but wins on the third, a statistician would conclude that the odds of winning are unlikely to be close to 1, and are more likely to be about 1 in 3 than about 1 in 100. If a large number of gamblers each buy lottery tickets for 12 consecutive days, and then we pick one of them and discover that he won at least once during the first three days, we conclude that it is likely that the odds of winning are quite good. Statisticians can specify just how good the odds are, in terms of the ‘confidence level’. A 95 per cent confidence level, for example, means that their conclusions are likely to be right 19 times out of 20, since 5 per cent is one twentieth of 100 per cent. The 95 per cent level is usually regarded as a good benchmark. In this particular example, the proper statistical conclusion is that the chance of winning the lottery with any one ticket is at least 0.12, at the 95 per cent confidence level. This means you could expect to win slightly more than once in every ten goes. So if the prize was worth at least ten times the cost of a ticket, it would be well worth entering this particular lottery.
As far as biogenesis is concerned, all the Earth-like planets in the Milky Way are our sample of gamblers, and Earth itself is our chosen individual, who won the lottery very early on. Life emerged (the lottery was won) certainly within 600 million years of the end of the bombardment, and perhaps much sooner than that. The statistical calculation then leads to the conclusion, at the 95 per cent confidence level, that life exists on at least 13 per cent of all the Earth-like planets that are at least a billion years old, and that the proportion of such planets with life is ‘most probably close to unity’. Or, in everyday language, life is common in the Universe.
The same argument applies whether life originated on Earth itself, or originated in space and was brought down to Earth in comets, or originated on some other planet and has spread through the Galaxy by hitching a ride on pieces of cosmic rubble (panspermia) or even by being deliberately spread by intelligent aliens (directed panspermia). No matter how life on Earth got started, it got started so quickly that it is highly likely that it exists on other planets as well. This is seen by some people as an important piece of information to feed in to an equation devised by the astronomer Frank Drake to quantify the chance of finding intelligent life elsewhere in the Universe. I have my doubts about the usefulness of this ‘Drake equation’; but it does at least provide a neat way of summing up our ignorance about the subject.
The Drake equation was a product of the excitement about space generated by the launch of the first artificial Earth satellites in the late 1950s. At that time, Frank Drake was working at the Green Bank radio observatory, in West Virginia, and was interested not only in the possibility of other intelligent civilizations existing, but in the prospect of communicating with them using radio telescopes. Early on, the handful of astronomers who took the idea seriously coined the term Communication with ExtraTerrestrial Intelligence, or CETI, to describe what they hoped would be the outcome of their work, and in 1974 signals were beamed into space from the giant radio telescope at Arecibo, in Puerto Rico, in the hope that one day some being ‘out there’ might detect them and reply. ‘Some day’ will be a long time coming – for reasons I’ve never understood the signal was beamed towards a cluster of stars known as M13, in the direction of the constellation Hercules. The cluster is 25,000 light years away, and radio waves travel at the speed of light (c), so even if any aliens pick up the message and reply immediately their answer won’t get back here until 50,000 years from now! Even that would be too soon for many people. The prospect of actually communicating with ET caused so much alarm in some quarters that the perception of politicians and the public, in the USA in particular, that such beings might not be friendly led to a slight change of name, to the Search for ExtraTerrestrial Intelligence, or SETI. Nowadays, some radio astronomers still listen out for alien signals, but they do not broadcast messages to the stars. It’s just too bad if all other civilizations are equally paranoid, so that everybody is listening and nobody is speaking!
That change of name still lay in the future, though, when Drake organized the first scientific conference devoted to the possibility of CETI (or SETI), at Green Bank in 1961. The aim of the meeting was to raise awareness of the possibilities and provide propaganda for fund-raising for the search, and Drake succeeded brilliantly in this aim by coming up with, as the basis for discussion at the meeting, the equation that now bears his name.
This is based on the fact that the probability of two things each happening is equal to the probability of one thing – one event – happening on its own, multiplied by the probability of the other event happening. The chance of rolling a three on a die, for example, is 1/6, since there are just six possible ways the die can end up. The chance of rolling a three on a second die is also 1/6. So the chance of rolling two dice and getting two threes is (1/6) × (1/6), or 1/36. Simple. And the same basic rule applies if you have a string of events and want to know the probability of all of them happening together.
Drake tried to think of all the factors that could affect the emergence in our Galaxy of a technological civilization capable of communicating across interstellar space. He started with the number of stars in the Milky Way, which he called N*. Next came the fraction of stars that are like the Sun, fs. Then, we have to estimate the fraction of those stars that have planets, fp. The fraction of those planets that lie in the life zone around their parent star is denoted by ne. Lineweaver’s calculation comes in to the next number, the fraction of those planets on which life actually does arise, fi. Pure guesswork comes in to any estimate of the fraction of those planets where intelligence like ours arises, fc. And as a kind of joker in the pack, Drake added in a number to provide scope for a guess of the percentage of the lifetime of a planet during which it is occupied by a civilization able and willing to communicate with other intelligences; he called this fl.
Putting everything together, if N is the number of civilizations beyond Earth that we might be able to communicate with in the Milky Way today,
This is the Drake equation.
The good news is that we start out with a large number of stars – N* is at least several hundred billion, and may be a trillion. Even better, although no extrasolar planets were known in 1961, we now know that planetary systems are common, although the jury is still out on whether or not Earth-like planets are common. And if we take Lineweaver’s statistical argument at face value, fi is probably close to 1 and almost certainly bigger than 0.13. The bad news is that if even one of the other numbers in the equation is zero, then N = 0, no matter how big all the other numbers are.
Almost as bad, it isn’t immediately obvious how you can quantify those other numbers, although that is more or less what I shall be trying to do in the rest of this book. Worst of all, though, Drake made a gross oversimplification (entirely justified in the context of what he was trying to do and what was known in 1961, but no longer valid) by using just one number to represent an estimate of the fraction of those planets where intelligence like ours arises, fc. This number is itself best understood as the product of a long string of other numbers representing the probability of the various events in the history of life on Earth that led to the emergence of our own civilization, and, as I shall explain, any one of those numbers could be close to zero, making fc itself, and therefore N, absolutely tiny.
Human experience suggests that the last number in the equation, fl, is also small. American Michael Shermer has a take on what civilization is that is different from that of many other people. From one point of view, all of human civilization is a continuous whole; but Shermer points out that many separate civilizations have risen and fallen since records began, and only one of them produced the kind of technology that enables, for example, the construction of radio transmitters. He has looked at the lifetimes of sixty terrestrial civilizations, ranging from Sumeria through Babylonia, Egypt, Greece and Rome and into the modern era, including eleven in China, four in Africa, three in India, two in Japan, six in Central and South America, and six modern states in Europe and America. The total lifetime covered by the sixty civilizations is 25,234 years, giving an average lifetime, fl, of 420.6 years. Worse, for the twenty-eight civilizations since the fall of Rome the average lifetime is only 304.5 years – and the ability to send and receive radio signals has only arisen once. On that basis, Shermer calculates that even with optimistic estimates for the other numbers in the Drake equation, there are at most three other radio-transmitting civilizations in the Galaxy today.
None of this has stopped people trying to use the Drake equation, or some variation of it, to calculate N, using their own preferred guesses for the numbers on the right hand side of the equation. It’s a sign of how futile this approach is that the ‘answers’ they get range from zero to a few hundred billion. The Drake equation is best regarded not as an equation that can actually be solved to give a realistic measurement of the number of technological civilizations in our Galaxy, but as a kind of mnemonic to remind us of the sort of things we have to take into account when considering the possibility of finding intelligent life elsewhere. Perhaps it isn’t surprising that, faced with the complexities involved, some people prefer to go back to arguments based on statistics and probability theory, with some rather startling conclusions.
THE INSPECTION PARADOX AND THE COPERNICAN PRINCIPLE
One of the most crucial numbers in the Drake equation is fl, corresponding to the lifetime of a technological civilization. Your best guess for the value of this number probably depends on how optimistic you are about the fate of our own civilization. No less an authority than the then President of the Royal Society, Sir Martin Rees, was gloomy enough to suggest that we may have less than a hundred years left; but it is possible to find optimists (many of them science fiction authors) who believe that humankind can overcome all the problems facing us and develop the resources of our Solar System for millions, perhaps even tens of millions, of years. Probability theory can improve on such hunches and give an insight into the likely lifetime of our own civilization. It can also provide insight into our past, using the same statistics that apply to catching a bus.
We’ve all had the experience of waiting ages for a bus, and then seeing two (or even three) arriving one behind the other. Somehow, most of the times you go to catch a bus you seem to have to wait longer than half of the average interval between buses. How can this be so? Surely it ought to even out, with the next bus coming soon on some days and a longer wait on others? But a little thought shows that it really is true that on average you wait ‘longer than average’ for your bus.
It happens like this. Suppose the buses on your route start out from the depot at regular 10-minute intervals. They may get bunched up by traffic, but the average interval is still 10 minutes – if one bus is delayed by 2 minutes, the interval between this bus and the one in front is now 12 minutes, but the interval between it and the one behind is now 8 minutes, so the average interval is still 10 minutes. If you stood at the bus stop all day and measured the intervals between each bus, that’s what you would find – varying gaps, but with an average of 10 minutes. But that is not what happens when you catch a bus. You arrive at the bus stop at some random time, and get on the first bus that turns up. If the buses were evenly spaced, your average wait would be 5 minutes – half the interval between buses. But because the buses are not evenly spaced, you are more likely to arrive at the stop during a long gap than during a short gap. If there is a gap of 20 minutes after one bus, and then two buses arrive within a minute of each other, it is 20 times more likely that you will arrive at the stop during the long gap than during the short one. So you will probably have to wait longer than half the average interval between buses. Put that way, it’s common sense; the probability theorists can quantify all this, and have given the phenomenon the name ‘the inspection paradox’.
The inspection paradox can also explain why there were a lot of old people around in Shakespeare’s day even though the life expectancy at the time was low. The average life expectancy at birth was low because many children and babies died in infancy. If someone survived to become an adult, they had a good chance of living to a decent old age. Even in a population where the life expectancy at birth is, say, 20, there will be people who live to be 70 or older. No matter how old you are, you will still have some life expectancy, and at any age your total life expectancy is greater than the life expectancy at birth for the general population. Everybody has better than average life expectancy – another example of the inspection paradox. Everybody in good health reading this can take comfort in the knowledge that you have a better than average life expectancy. By the same token, the fact that our civilization is still ‘alive’ means that it has a better than average life expectancy compared with all the civilizations that have ever existed in the Milky Way Galaxy – but this is based on the idea, which Michael Shermer disagrees with, that all of human history represents a single ‘civilization’. As the mathematician Amir Aczel has put it, referring to the longevity of life on Earth, not just of human civilization, but taking this perspective, ‘our ability to inspect ourselves is an outcome of the fact that we’ve been here on this planet for a long time [and] the conditional probability that we have been around for longer than other civilizations . . . is high.’ And ‘assuming other civilizations exist, chances are that we are among the first in our galaxy to arrive at this level of advancement.’ Curiously, he has arrived at the same conclusion as Shermer – that we are probably alone in the Galaxy – starting out from a completely different assumption about what constitutes a civilization!
But that only tells us about the past. What about the future of human civilization?
In 1993, Richard Gott, of Princeton University, provoked debate when he published a paper in the leading science journal Nature in which he used this kind of statistical reasoning to estimate the total lifetime of our species. He found that at the 95 per cent confidence level Homo sapiens is likely to be around for a total (including our history to date) of between 200,000 years and 8 million years; the lower end of this range corresponds to us going extinct more or less tomorrow, and even the upper end is hardly impressive on an astronomical timescale.
The only assumption that Gott put in to his calculation is that you are a random example of all the intelligent observers that ever exist. He calls this the ‘Copernican principle’, by which he means the assumption that we do not occupy a special place in the Universe – in fact, Copernicus only said that we do not occupy a central place in the Universe, but the extension of this idea to say that we occupy a nonspecial place is quite common. Slightly less grandiosely, it is sometimes called the ‘Principle of Terrestrial Mediocrity’ – the idea that we occupy an ordinary planet orbiting an ordinary star in an ordinary galaxy.
Before setting out the consequences of this assumption, Gott illustrates the power of this approach with two examples from his own life. The essence of the argument is that if you observe something at random for the first time, there is a 95 per cent chance that you are seeing it in the middle 95 per cent of its lifetime. Just 2.5 per cent of the time you might be seeing it near the beginning of its life, and just 2.5 per cent of the time you might be seeing it near the end of its life. If you know how old the thing is when you first see it, you can use these figures to set limits on its future lifetime – with standard probability calculations similar to those involved in the bus ‘paradox’, the future lifetime of the thing you are observing should lie between 1/39 times and 39 times its past lifetime, at the 95 per cent confidence level.
In 1969, on a trip to Europe, Gott saw both Stonehenge and the Berlin Wall for the first time. Stonehenge is about 3,700 years old, and in 1969 the Berlin Wall was 8 years old. This kind of calculation would imply a future lifetime for Stonehenge of at least a hundred years, and a future lifetime for the Berlin Wall (as of 1969) of no more than a couple of hundred years. The Wall fell in 1989, but Stonehenge is still there, in line with these estimates.
Applying the same logic to humankind, Gott starts from the estimate, based on fossil evidence, that modern people, Homo sapiens sapiens, have been around for 200,000 years. This suggests that we are likely to be around for at least another 5,000 years, but for no more than 8 million years. Gott points out that the average lifetime for a mammalian species is about 2 million years, and that our immediate ancestor, Homo erectus, was around for just under 1.5 million years, so his calculation is very much in line with what we know from the fossil record. Perhaps our descendants will evolve into something as different from us as we are from erectus, and still have a technological civilization; but further calculations by Gott paint a much more gloomy picture.
The same kind of statistical arguments apply to the number of people alive on Earth today, compared with the number already born and the number yet to be born. In this case, you are a random ‘observer’ picked out from the population of all the people that ever were or ever will be simply by being born. The probability equations tell us that 50 per cent of all such ‘observers’ are born at a time when the population is at least half of its maximum value. The present human population of the Earth is nearly 7 billion, and the planet’s estimated carrying capacity is about 12 billion, so on that basis it is not surprising that you are alive today. It is an example of the fact that you are a random intelligent observer picked out by chance from all of the intelligent observers in the past, present and future. In 1993, using a version of the calculation previously applied to timescales, Gott calculated that the number of people still to be born was at least 1.8 billion, and estimated that this total would be reached in the first decade of the twentieth century. Applying the same calculation today, at least another 2 billion people are still to be born, and this will take roughly another ten years. If Gott is right, sooner or later (and most probably sooner) there will be a population crash and a collapse of civilization. That may not apply with such force to other civilizations; but Gott has more bad news for proponents of SETI.
SETI advocates still pin their hopes mainly on radio communication. In 2004, one of the most prominent of these advocates, the Microsoft billionaire Paul Allen, donated a further $13.5 million, on top of earlier donations totalling $11.5 million, towards the construction of the ‘Allen Array’, a dedicated SETI radio telescope. But he may have been wasting his money (although, fortunately, the Allen Array can be used for conventional radio astronomy as well). Our civilization has only used radio for about 120 years, and the probability calculation says that our future lifetime as a radio-transmitting civilization is likely to be between three years and five thousand years. This doesn’t necessarily imply the collapse of civilization – our descendants may move on to something superior to radio communication, just as we no longer use smoke signals. But it does suggest a severe constraint on our chances of making contact with ET.
Assuming that we are a typical example of a radio-transmitting civilization, and plugging this number into his version of the Drake equation, Gott estimates that the number of radio-transmitting civilizations in the Galaxy today is no more than 121. The chance of one of them being within range is so small as to make any radio SETI project futile, if Gott’s calculations are correct. It may be futile in any case, since the most likely flaw in his argument is the Copernican principle itself. Perhaps we are not typical observers, after all. That’s the most likely resolution of the most famous ET ‘paradox’ of them all, formulated clearly by the physicist Enrico Fermi, and now given his name, although he was not, in fact, the first person to puzzle over it.
PANSPERMIA AND THE FERMI PARADOX
Fermi was one of the most important physicists of the twentieth century. Among his many achievements he predicted the existence of the particle known as the neutrino, and he received the Nobel Prize in 1938 for his work on radioactivity and nuclear reactions. With war looming in Europe, instead of going back to Fascist Italy, the Fermi family went on from the Nobel awards ceremony in Stockholm to the United States, where Fermi became the leader of the group at the University of Chicago that built the first nuclear reactor, known at the time as an ‘atomic pile’. The pile ‘went critical’ at 2.20 p.m. on 2 December 1942.
Fermi had a great ability to see to the heart of a problem, and to express complex ideas in simple language. He was a master of the art of making rough estimates – called order of magnitude calculations – of the solution to complicated problems, and it was this that led him to the Fermi paradox. He became so well known for this that these kind of puzzles are often referred to as ‘Fermi questions’. A simple example of a Fermi question is: How many barley sugar sticks will fit into a 1-litre jar? The point is not to get a precise answer, but to make an educated guess. A stick is roughly cylindrical, about 2 cm long and 1.5 cm in diameter (0.75 cm in radius). The volume of such a cylinder is π multiplied by the square of the radius multiplied by the length, which comes out as about 25/7 cubic cm, if we use the approximation π = 22/7. But the sticks don’t fit tightly together in the jar, so as a guess we might say that 20 per cent of the volume is air. A litre is 1,000 cubic cm, so the sticks actually take up 800 cubic cm, and the number of sticks needed to do the job is 800 divided by 25/7, which is about 220 (it’s exactly 224, but it would be silly to quote the ‘answer’ that accurately). It’s this kind of calculation that astronomers use to estimate things like the number of stars in a galaxy without actually counting them all, and which Fermi used to come up with his famous puzzle.
Although Fermi died in 1954 at the age of only 53, without leaving a memoir of the occasion when he came up with his paradox, the exact details of the story were reported by the physicist Eric Jones, on the basis of interviews with Fermi’s contemporaries, in the August 1985 edition of the journal Physics Today. It happened in the summer of 1950, when Fermi was at the Los Alamos laboratory where the first nuclear bomb had been developed a few years earlier. This was during the height of public interest in flying saucers (UFOs), triggered by sightings of secret aircraft during the post-war period. There had also been a spate of disappearances of rubbish bins (trash cans) from the streets of New York, and the New Yorker had just published a cartoon suggesting that the trash cans were being stolen by aliens. Fermi and his colleagues had been laughing over the cartoon on the way to lunch, and this led them into a discussion about the (im)possibility of travelling faster than light. Over lunch, the conversation turned to other matters. Then, suddenly, Fermi asked out loud, ‘Where is everybody?’ His colleagues realized that he was referring to extraterrestrial intelligences, and since it was Fermi asking the question, they took it seriously. He quickly made an order of magnitude calculation which implied that even if they were restricted to travelling slower than the speed of light, aliens should have long since colonized the entire Galaxy and the Earth should have been visited many times. The neatest summing up of the Fermi paradox is contained in the question, ‘If they are there, why aren’t they here?’ In other words, if there are extraterrestrial intelligences, why haven’t they visited us?
Nobody took much notice of the puzzle, except as a topic for coffee-time discussion, until 1975. Then, like two buses coming along together, two scientific papers appeared that stimulated a much broader discussion. Writing in the Journal of the British Interplanetary Society, David Viewing re-stated the puzzle, giving full credit to Fermi. The same year, Michael Hart published a paper in the Quarterly Journal of the Royal Astronomical Society in which he phrased the question slightly differently – why are there no intelligent visitors from other worlds on Earth today? Unlike Viewing, though, Hart offered four possible categories of explanation for the puzzle:
1. It may be physically impossible to get from there to here
2. They are there, but they have no wish to contact us
3. They are there, but they have not yet had time to reach us
4. They have been here but have left no trace and are not here now
There’s one other possibility, which is in essence a variation on Hart’s category 4 – perhaps we are the aliens. This is another way of looking at the idea of panspermia, mentioned earlier. And although it does not affect the main argument of this book, it is not only fascinating in its own right, but offers a powerful insight into how easy it is for life to spread across the Milky Way, given the enormous span of cosmic time available.
Panspermia literally means ‘life everywhere’, and speculation about the possibility of life everywhere goes back to ancient times. But the science of panspermia can be seen to have started with some remarks made by William Thomson, later Lord Kelvin, in his Presidential Address to the meeting of the British Association for the Advancement of Science, in 1871. Thomson was following up work by Louis Pasteur in the 1860s, which had finally proved that living things do not spontaneously come forth from non-living things – that maggots, for example, are not a product of rotting meat, but come from eggs laid by flies. He declared that all life comes from life – all living things have ancestors. As Thomson put it, ‘dead matter cannot become living without coming under the influence of matter previously alive. This seems to me as sure a teaching of science as the law of gravitation.’ We would now say that there was, at least once and very long ago, an occasion when a living molecule arose from non-life, but that does not affect the subsequent thrust of Thomson’s argument.
Thomson made an analogy with the way life appears on a newly formed volcanic island. ‘We do not hesitate to assume that seed has been wafted to it through the air, or floated to it on a raft,’ rather than being generated spontaneously from the dead rocks. The Earth, he said, is in the same situation:
Because we all confidently believe that there are at present, and have been from time immemorial, many worlds of life besides our own, we must regard it as probable in the highest degree that there are countless seed-bearing meteoritic stones moving about through space. If at the present instant no life existed upon earth, one such stone falling upon it might, by what we blindly call natural causes, lead to its becoming covered with vegetation.
Few of Thomson’s contemporaries took the idea seriously. But one who did was the Swede Svante Arrhenius, a chemist who won the Nobel Prize in 1903 for his work on electrolysis, and who was a pioneering investigator of the atmospheric ‘greenhouse effect.’ His grandson, Gustaf Arrhenius, was, incidentally, one of the people who studied the isotopic evidence for early life on Earth.
Instead of considering seeds being carried through space inside meteorites, Svante Arrhenius speculated that microorganisms like bacteria might be carried high into the Earth’s atmosphere and escape, to be blown across space by the pressure of the Sun’s radiation. Such microorganisms can remain inert for long periods of time before reviving, perhaps long enough for them to cross interstellar space and land on some other Earth-like planet. Arrhenius estimated the travel times for such seeds of life, starting from Earth, as 20 days to Mars, 80 days to Jupiter, and 9,000 years to Alpha Centauri, the nearest star to the Sun. And if they could travel one way, why not the other, with the Earth having been seeded by microbial life from space?
Panspermia has never been a mainstream idea in astrobiology, but since the speculations of Thomson and Arrhenius people have returned to it from time to time in different contexts, and strengthened its scientific credentials without ever making a completely convincing case. One variation on the theme picks up Arrhenius’ original idea, and adapts it to take account of the conditions that we now know life would be likely to encounter on its journey through space. A star like the Sun produces a large amount of ultraviolet radiation, which would be lethal for microorganisms that escape into space. But Jeff Secker, of Washington State University, and his colleagues Paul Wesson and James Lepock, at the University of Waterloo, Canada, calculated what would happen if the microorganisms were encased in tiny grains of ice or dust. This would still not be enough to protect them from the radiation from a star like the Sun. But in old age a star swells up to become a so-called red giant, which would be luminous enough to push the particles out across space, but does not produce the damaging ultraviolet rays. Secker and his colleagues calculate that the journey time for such grains carrying the seeds of life across space is about 20 light years per million years. Since there are dozens of stars within 20 light years of the Sun, this suggests that life could move easily from one planetary system to the next, and could cross the entire Galaxy, spreading life everywhere, in a few billion years.
The better the protection, of course, the longer the organisms could survive. Jay Melosh, of the University of Arizona, has shown that microorganisms could survive for many millions of years deep inside large chunks of rock. This is intriguing within the context of our own Solar System, because large impacts from space can eject just such chunks of rock from the surface of a planet. Indeed, there are meteorites found on Earth which have been identified, partly from isotope evidence, as coming from Mars. One of these pieces of Martian rock, dubbed ALH 84001, has been the subject of intense investigation since claims were made that microscopic tube-like structures in the rock resemble fossilized bacteria. There is very little evidence to support this claim. But what the presence of meteorites like ALH 84001 on Earth does tell us is that if there were life on Mars it could be carried to Earth in this way, after spending millions of years wandering about the inner Solar System after the impact in which it was blasted off the surface of Mars. It’s slightly harder for life from Earth to get transported to Mars, or anywhere else, in this way, because of the Earth’s stronger gravitational pull. And, alas, it is virtually impossible for large chunks of rock to be completely ejected from the Solar System, to carry the seeds of life to other planetary systems.
It would be a lot easier if the seeds of life were deliberately directed at suitable planets. This idea, called directed panspermia, was developed by Francis Crick, co-discoverer of the structure of DNA, and another molecular biologist, Leslie Orgel. It would certainly be easy, using technology only slightly more advanced than our own, to send out probes packed with the kind of blue-green algae that have proved so successful as life forms on Earth towards any interesting looking planetary systems. And it would be a quick way to ‘colonize’ the Galaxy, if you feel that blue-green algae are suitable representatives of life on Earth. So are we the colonists? Did our primordial ancestral cells arrive on Earth as a gift from another civilization? Are ‘they’ here because we are them? It’s certainly technically possible. But the big question is why would any intelligent beings do this? Even Crick and Orgel have never been able to give a satisfactory answer to that one, and in his book Life Itself Crick admits that ‘as a theory it [directed panspermia] is premature.’ Orgel is equally undogmatic. He has been quoted as saying ‘my opinion is that we have no way of knowing anything about the possibility of life in the Cosmos. It could be everywhere, or we could be alone.’
I agree with Crick’s conclusion. But I have not ignored the idea of directed panspermia, because it introduces the notion that, for whatever motives, an alien civilization, or civilizations, might decide to send spaceprobes to visit other planetary systems. This possibility provides what I consider to be the definitive resolution of the Fermi paradox. Before unveiling that resolution, however, it is only fair to mention the other suggested ‘answers’ to the puzzle, if only to say why they are implausible.
The most detailed and accessible compilation of proposed ‘answers’ to the Fermi paradox has been made by Stephen Webb, a physicist at the Open University, in his book Where is Everybody? If the few examples I have room for here whet your appetite for such speculation, that is the place to look for more, although the weirdness of some of the suggestions only adds strength to my own argument.
Many people, of course, believe that aliens are here, and that some of them are even in the habit of abducting people in their flying saucers. There is no credible evidence for this, and although in some quarters the term Unidentified Flying Object, or UFO, has become synonymous with ‘flying saucer’, UFO buffs are missing the point that just because something is unidentified doesn’t automatically mean that it is an alien spacecraft. If a car goes past me too fast for me to identify the make, in a sense it is an unidentified moving object; but that doesn’t automatically mean that it is a top-secret rocket-propelled prototype. The logical suggestion is that on closer inspection I would be able to identify it as a known make of car. So many reported UFOs turn out on inspection to be identifiable, as known atmospheric phenomena, weather balloons, aircraft, and the like – or even (I kid you not) the planet Venus – that the simplest explanation for the small percentage that remain formally unidentified is that they are also caused by natural phenomena, but we don’t have enough information to decide which category they belong to, just as I don’t have enough information to decide the make of the fast car.
A closely related idea to the flying saucer story is the suggestion that we live in a cosmic zoo, or a kind of wilderness area where primitive creatures such as ourselves are being left to develop at their own pace. This often carries with it the implication, explicit in many science fiction stories including the Star Trek movie franchise, that once we reach a certain level of technological ability (or perhaps, once we grow up enough to stop fighting among ourselves) we will be welcomed into some sort of galactic club of advanced civilizations, albeit as a junior member. Among the many objections to this idea is: why didn’t the galactic club take over the Earth in the billions of years when it was only occupied by single-celled life forms? Did they only discover us a few million years ago, when primates were clearly already on the road to intelligence? And why is there absolutely no evidence of their activity out among the stars of the Milky Way?
At the other extreme from the galactic club there is the suggestion that all the alien civilizations have stayed at home, because they are not interested in space travel (a variation on the theme, which I alluded to tongue-in-cheek earlier, that everyone is listening for a signal but nobody is transmitting). This argument might hold water if we are thinking about just one alien civilization. But if you want to believe that what we call civilization is common in the Galaxy, you would then have to accept that none of them want to investigate their surroundings, because it turns out, as I shall explain, to be very easy to make a mark on the Milky Way. We will soon do so ourselves, if we give the lie to Martin Rees’s gloomy prognostications and survive the present century.
The most serious objection to the idea that ‘if aliens are there, they ought to be here’ is that interstellar travel is tedious and difficult – even though the whole point of Fermi’s insight is that he realized, from simple order of magnitude calculations, that on the timescale of human civilization (let alone the timescale of life on Earth) it is relatively quick and at one level easy. There are even examples from human history – strictly speaking, prehistory – that highlight the point.
What’s difficult about travelling to the stars is doing it within a human lifetime, and it’s more than doubly difficult to do so and come home again. We should not ignore the possibility that other life forms may be much more long-lived than us, and would find a journey lasting a few hundred years no more tedious than we find a flight across the Atlantic, nor the possibility of a civilization so much more advanced than ours that it can take advantage of the shortcuts through spacetime allowed by the general theory of relativity, or tap in to the quantum field energy of empty space to power their ships. Either possibility only makes it even more likely that aliens will colonize the Galaxy. But it could be done by beings like us with only slightly more advanced technology than our own. Possible propulsion systems include nuclear-electric rockets, fusion rockets, the interstellar ramjet, and (my favourite) ‘starsailing’ with the aid of powerful planet-based lasers. A neat and only slightly dated overview of the possibilities is provided by Robert Forward and Joel Davies in their book Mirror Matter (Forward, himself a physicist, is one of the leading proponents of starsailing), and I shall not elaborate on them here. What matters is that whatever your means of propulsion, if you go slowly enough you can get to the nearest stars without too much effort. And if colonies are established on planets orbiting those stars, they can reach the stars near to them but farther from us without too much effort.
The key to colonizing the Galaxy is that you don’t come back. The archetypal example from human prehistory is the way Polynesian people spread from island to island across the Pacific, eventually reaching all the way to New Zealand, and even to isolated Easter Island, some 1,800 km from the nearest land. They didn’t visit new islands and then go back; they settled on new islands, and used those islands in turn as bases from which to send out people who settled on more remote islands. None of this was planned. It wasn’t part of a grand scheme to colonize the Pacific. It just happened, because of population pressure, or from a basic urge to find out what lay beyond the horizon.
In a similar way, although our ancestors evolved in East Africa, their descendants spread out on foot all over the world. They even crossed the land bridge between what are now Siberia and Alaska, and got all the way down into South America. Like the Polynesian voyages, none of this was planned. It just happened that in each generation (or just in some generations) a few people moved on a little way from their neighbours, in search of food and water, or just to see what lay over the next hill, or to get away from the crowd. The whole process took about 30,000 years. Australian-born radio astronomer Ronald Bracewell, from Stanford University, summed this up by pointing out that it was quicker for humans to walk from Africa to South America than it would have been for human-level intelligence to have evolved independently in South America. The last stage of the journey, from present-day Canada to Patagonia, covers about 13,000 km, a distance that could have been traversed in only a thousand years at a rate of merely 13 km per year. His conclusion is that it is much more likely that intelligent life would spread across the Galaxy from the first ‘intelligent planet’ than that it would evolve independently several (or many) times on several (or many) different planets.
Bracewell’s conclusion draws on a calculation made by Michael Hart and published in his paper in the Quarterly Journal of the Royal Astronomical Society in 1975. Hart used the example of the possible future colonization of the Galaxy from Earth. Making reasonable assumptions about possible propulsion systems allowed by the laws of physics, he assumes that ‘we eventually send expeditions to all of the 100 nearest stars’, which all lie within about 20 light years of the Sun. ‘Each of these colonies has the potential of eventually sending out their own expeditions, and their colonies in turn can colonize, and so forth.’ With no pause between trips, ‘the frontier of space exploration would then lie roughly on the surface of a sphere whose radius was increasing at a speed of 0.10 c. At that rate, most of our Galaxy would be traversed within 650,000 years.’ Of course, the assumption of no pause between trips is over-optimistic, but even if the length of time between voyages is the same order of magnitude as the length of a voyage, the time needed to colonize the Galaxy would only be doubled, to 1.3 million years. This is not much more than one eight-thousandth of the age of the Galaxy. More pessimistic estimates of the time required, based on how long it takes for colonies to become established, range from a few million years to 500 million years – but even that is still very short compared with the age of the Galaxy. And, as Fermi appreciated, the precise numbers don’t matter, just the order of magnitude. If we could do it, so could they. So – why aren’t they here?
The argument strikes with even more force once you realize that it isn’t even necessary to subject living creatures to the boredom and hazards of interstellar travel.
The most powerful argument for the non-existence of other technological civilizations in our Galaxy stems from the work of two mathematical geniuses who each made major contributions both to the science of computing and to the Allied victory in World War Two. Alan Turing is remembered as a cryptographer who was the leading member of the team at Bletchley Park, in Buckinghamshire, which cracked the German codes during that war. But as early as 1936 he had written a scientific paper, ‘On Computable Numbers’, which laid out the fundamental principles of machine computing. Turing proved that it is possible in principle to build a machine, now referred to as a universal Turing machine, which could solve any problem that could be expressed in the appropriate machine language. To a generation brought up on personal computers, this seems blindingly obvious – but it was the proof that such machines could be built that started science and technology down the road to the modern computer. In Turing’s own words, the universal machine ‘can be made to do the work of any special-purpose machine, that is to say carry out any piece of computing’ if fed with the right program. The computer built at Bletchley Park as part of the code-breaking effort was an example of a special-purpose machine, designed to do only one job. But the first general purpose computer was soon developed, with the aid of Hungarian-born John von Neumann, who among many other things worked on the Manhattan Project, which developed the first nuclear bomb.
Von Neumann’s interest in the idea of a computer that could solve any problem led him to think about the nature of intelligence and of life. Intelligent beings are in a sense universal Turing machines, since they can solve many different kinds of problems, but they have the additional ability to reproduce. Was it possible in principle for there to exist self-reproducing Turing machines in a non-biological sense? Von Neumann proved that it is indeed possible. The process involves just a few simple steps. First, the computer program stored in the memory banks of the machine instructs the machine to make a copy of the program and store it in some sort of memory bank (nowadays, we could imagine this to be an external hard drive). Then, the program instructs the machine to make a copy of itself, with a blank memory. Finally it tells the machine to move the copy of the program from the storage device into the new machine. Von Neumann showed, as long ago as 1948, that living cells must follow exactly the same steps when they reproduce, and we now understand this in terms of nucleic acids as the ‘program’ and proteins as the ‘machinery’ of the cell. First, the DNA is copied. Then, as the cell divides in two the copy of the DNA is moved into the new cell.
A self-reproducing non-biological automaton is now often referred to as a ‘von Neumann machine’. Neither Turing nor von Neumann lived to see these ideas developed. Turing was only 41 when he killed himself, in 1954, after years of harassment by the authorities for his homosexuality; von Neumann died of cancer in 1957 at the age of 53. It was Ronald Bracewell who suggested that probes could be used to explore the Milky Way, and the American Frank Tipler who presented the full force of the argument which shows how quickly such von Neumann machines could visit every interesting planet in the Galaxy.
The key point is that a technological civilization only has to build one or two probes in order to colonize (by proxy) the entire Milky Way. Such a probe would be programmed to use the raw materials that it found among the asteroids and other cosmic rubble in a planetary system, plus the energy of the parent star, to build copies of itself, and send those copies off to explore other planetary systems. One or two probes sent from Earth to the Asteroid Belt between Mars and Jupiter could mine the raw materials there to make a fleet of identical probes which could set off to explore nearby stars, maintaining contact with home by radio. Each time one of them arrived in a new planetary system, as well as reporting its findings it would set about building copies of itself and repeating the process. Making the very modest assumption that probes could travel at one fortieth of the speed of light and that they were programmed to seek out stars with planets, it would take less than 10 million years (in a galaxy 10 billion years old) from the construction of the first probe to visit every interesting planet in the Galaxy. And all it costs is the construction of one initial probe (or at most a few copies, as backups).
Even with our present technology, we could get a decent-sized probe to the nearest star. It would involve sending the probe on a close flyby of Jupiter, using the gravity of Jupiter like a slingshot to speed it up and send it diving past the Sun, where the Sun’s gravity would give it a further boost, sending the probe out of the Solar System at a speed of about 0.02 per cent of the speed of light. When the probe arrived at its target star, it could use the gravity of the star and its planets to slow itself down. It would take several thousand years for the probe to reach its target, but if civilization still existed on its home planet and anyone was interested, it could be programmed by radio to construct not just a replica (or replicas) of itself but an improved, faster version (assuming technology back home had advanced in the millennia since the probe was launched). It might take thousands of years for any interesting news to come back from the first probe, or probes; but as they reproduced and spread faster and faster through the Galaxy news about different planetary systems would come flooding in several times each year.
This is so nearly within our present technological ability that it is quite clear that within a couple of decades at most (unless civilization collapses) we will be able to start this process. And by ‘we’, I don’t necessarily mean the full might of government-sponsored agencies such as NASA. Individuals such as Paul Allen already pay for radio telescopes to search for ET; the Paul Allens of the next generation may well be able to pay to explore (or at least start the exploration of) every planet in the Galaxy. Such an individual might well be hoping to get news from the nearest planetary systems within their own lifetime, rather than caring much about what happens millions of years from now. But since it literally costs no more to explore the entire Galaxy than to explore the nearest planetary system, who could resist going the whole hog?
That is why the possibility of constructing von Neumann probes is such a powerful argument that we are alone in the Milky Way. All the arguments that ‘they’ are out there but are, for whatever reason, avoiding contact with us require that every technological civilization is working together to keep their presence secret. But it isn’t just the case that you only need one civilization to break ranks and send a few probes out into the Galaxy. All it needs is for one individual being to send one probe and every interesting planet will be visited in a few million years.
Fermi asked, ‘If they are there, why aren’t they here?’ The solution to the puzzle is that they are not there. But that raises an even more important question. Not ‘Are we alone?’ but ‘Why are we alone?’ If they are not there, why are we here? What is it that is special about our location in the Universe, in both space and time, that has allowed the development of the only technological civilization in the Galaxy? Why is Earth the only intelligent planet? That is the theme of the rest of this book – the reason why we are here to ask such questions.