Estimates for the number of communicating civilizations in the galaxy, based on the so-called Drake equation, are meaningless without a plausible estimate for the probability that life will emerge on an Earth-like planet. In the absence of a theory of the origin of life, that number can be anywhere from 0 to 1. Distinguished scientists have been known to argue that life on Earth is a freak accident, unique in the observable universe and, conversely, that life is almost bound to arise in the course of time, given Earth-like conditions. De Duve, adopting the latter position, coined the phrase that ‘life is a cosmic imperative’. De Duve’s position would be immediately verified if we were to discover a second sample of life that we could be sure arose from scratch independently of known life. Given the current absence of evidence for life beyond Earth, the best way to test the hypothesis of the cosmic imperative is to see whether terrestrial life began more than once. If it did, it is possible that descendants of a second genesis might be extant, forming a sort of ‘shadow biosphere’ existing alongside, or perhaps interpenetrating, the known biosphere. I outline a strategy to detect the existence of such a shadow biosphere.
1. Historical background
The subject of astrobiology rests on the popular hypothesis that life emerges readily in Earth-like conditions, and is therefore likely to be widespread in the universe, a point of view eloquently expressed by Christian de Duve in his evocative phrase that ‘life is a cosmic imperative’ . Fifty years ago, however, opinions were very different. Most biologists thought that the origin of life on Earth was a freak chemical accident involving a sequence of events so low in probability that it would be unlikely to be repeated anywhere else in the observable universe. Jacques Monod summarized the mood by declaring that ‘the universe is not pregnant with life’, and therefore that ‘Man at last knows that he is alone in the universe’ . George Simpson, one of the great neo-Darwinists of the postwar years, dismissed SETI, the search for extra-terrestrial intelligence, as ‘a gamble at the most adverse odds with history’ . Biologists such as Monod and Simpson based their pessimistic conclusions on the fact that the machinery of life is so stupendously complex in so many specific ways that it is inconceivable it would emerge more than once as a result of chance chemical reactions. Francis Crick expressed a similar position when he wrote ‘The origin of life appears… to be almost a miracle, so many are the conditions which would have had to be satisfied to get it going’ . In short, to profess belief in extra-terrestrial life of any sort, let alone intelligent life, in the 1960s and 1970s, was tantamount to scientific suicide. One might as well have expressed a belief in fairies.
What, then, has changed? Why is it now scientifically respectable to search for life beyond Earth? Oft cited as an explanation is the fact that many extra-solar planets have now been discovered, with a strong likelihood that billions of Earth-like planets may exist in our galaxy alone. As a result, there are probably a very large number of available habitats for life as we know it. However, this explanation will not do. Although no planets were identified beyond the Solar System until recently, most astronomers nevertheless supposed all along that they existed. Sometimes astrobiologists point to the recent discovery of many species of organic molecules in space, providing evidence that abundant ‘raw materials’ for life are scattered throughout the universe. That may be so, but the path from simple building blocks such as amino acids to a metabolizing, replicating cell is so long and tortuous, the fact that the first small step might have already been taken in space is irrelevant. Then there is the ‘follow the water’ fallacy. Wherever there is liquid water on Earth, there seems to be life, therefore, it is sometimes claimed, when we find water on other planets and moons, life should exist there too! Water does in fact seem to be abundant in the Solar System and beyond, so (it is reasoned) life should also be abundant too. Unfortunately this simplistic reasoning confuses a necessary with a sufficient condition. To be sure, liquid water is necessary for life (at least as we know it), but it is far from sufficient. The reason life on Earth inhabits almost all aqueous niches is because Earth has a contiguous biosphere, and life has invaded those niches; it has not arisen there de novo. Another reason given for the current optimism about life beyond Earth is the dawning recognition that life can survive in a much wider range of physical conditions than was recognized hitherto, opening up the prospect for life on Mars, for example, and generally extending the definition of what constitutes an ‘Earth-like’ planet. But this at most amounts to a factor of two or three in favour of the odds for life. Set against that is the exponentially small probability that any given complex molecule will form by random assembly from a soup of building blocks. In short, habitability does not mean inhabited. It is natural that we should concentrate on the habitable planets in our search for life—by the ‘keys under the lamppost’ principle—but at this stage we cannot put any level of confidence—none at all—on whether such a search will prove successful.
The correct explanation for the shift in mood lies, I suspect, with fashion. The pendulum has swung from scepticism to credulity without very much changing on the actual scientific front. True, we do know far more today about life’s basic processes, and about the physical conditions on other planets. But we still know almost nothing about the overwhelmingly important factor, namely, the probability that life will emerge on a planet given that it resembles Earth, for the simple reason that we have little or no idea beyond a few general scenarios and just-so stories how life actually emerged from non-life.
My point is thrown in sharp relief by studying the famous Drake equation, first written down by Frank Drake in 1961 at the dawn of SETI to guesstimate the number of communicating civilizations in our galaxy. Here it is where R* is the rate of formation of Sun-like stars in the galaxy; fp the fraction of those stars with planets; ne the average number of Earth-like planets in each planetary system; fl the fraction of those planets on which life emerges; fi the fraction of planets with life on which intelligence evolves; fc the fraction of those planets on which technological civilization and the ability to communicate emerges; and L the average lifetime of a communicating civilization.
Today we have good estimates of R* and fp, and should soon have a handle on ne too, when the results of the Kepler mission are released. The last three factors are notoriously hard to predict, but at least we have an accepted and well worked out theory—Darwin’s theory of evolution—that could in principle tell us something about the probability of the evolution of intelligence and even of civilization. And once civilizations exist, it is not hard to imagine that at least some fraction will survive for a long time. That leaves the factor, fl, the number of Earth-like planets on which life arises. The uncertainty in fl completely dominates the right-hand side of Drake’s equation, and makes discussion of the error bars on the other terms utterly moot. To take a simple (but absurd) illustration, if the simplest known microbe had formed by chance assembly from an ocean of molecular building blocks, the odds against it happening twice are so large they dwarf the number of atoms in the universe. But of course, that may not be how life on Earth formed at all. Chance may have played only a subordinate role. The origin of life may have been more ‘law-like’ than ‘chance-like’, so that it will emerge more or less automatically wherever conditions permit. Perhaps, life is indeed a cosmic imperative, somehow ‘built into’ the laws of physics in a fundamental way, and therefore an expected product of an intrinsically bio-friendly universe. Perhaps. But the trouble is these sentiments are philosophical, not scientific. Unlike Darwin’s theory of evolution, we have no accepted theory of life’s origin, only a collection of scenarios and conjectures. And without a proper theory, it is meaningless to assign probabilities to outcomes. Furthermore, there is nothing in the laws of physics that singles out ‘life’ as a favoured state or destination. The laws of physics (and chemistry) are ‘life blind’—they are universal laws that care nothing for biological states of matter specifically, as opposed to non-biological states. If there is a ‘life principle’ in nature, then it has yet to be elucidated. Perhaps such a principle lurks in the realm of complexity theory or information theory or in the properties of self-organizing systems, but so far there is no hard evidence for it. How, then, can we test the audacious and appealing idea of the cosmic imperative?
2. Finding a second sample of life
If we discovered a second sample of life that we could be sure had arisen from scratch, independently of life as we know it, then the case for the cosmic imperative would be immediately made. The most obvious and direct way in which such a discovery might occur is if SETI succeeded. If astronomers detected unmistakable signs of alien technology, then although such technology itself might be the product of machine as opposed to biological intelligence, it would almost certainly imply a biological precursor. A message from an extra-terrestrial source would enable us to assign a value very close to 1 for fl. Pending that dramatic event, what else might we look for? Future space-based systems such as the proposed Terrestrial Planet Finder might be able to detect the presence of gases in the atmospheres of other planets that would indirectly imply life—such as oxygen and methane in combination. Unfortunately it may be a long time before systems with the requisite resolution are available. Finding life on Mars from a sample return mission or an expedition might resolve the issue, but then again it might not. Mars and Earth are known to trade rocks, and it is clear that at least some fraction of microbial inhabitants of these rocks could survive the journey . By hitching a ride on impact ejecta, Mars life would readily infect Earth, and vice versa, so the two planets are not quarantined and in effect constitute a single biosphere .
Sometimes it is suggested that if we make life in the laboratory, it would prove that life starts up easily. But this is another fallacy. Synthetic biology demands sophisticated equipment and technicians, purified and refined substances, high-fidelity control over physical conditions and, above all, an organic chemist who has a preconceived notion of the entity to be manufactured—in other words, an intelligent designer. Astrobiologists want to know, however, how life began without any fancy equipment, purification procedures, environment stabilizing systems and, in particular, without an intelligent designer. Life may be easy to make in the laboratory, but still be exceedingly unlikely to happen spontaneously. After all, organic chemists can make plastics quite easily, but we do not find them in nature.
Another argument often used in favour of the cosmic imperative is that life established itself on Earth rather rapidly once conditions became suitable. For about 700 million years, our planet was pounded by large comets and asteroids. This heavy bombardment phase abated about 3.8 billion years ago, and already by 3.5 billion years ago microbial organisms were flourishing . As Carl Sagan once expressed it, ‘the origin of life must be a highly probable affair; as soon as conditions permit, up it pops!’ . The reasoning, of course, is that because the formation of life on Earth was relatively quick and easy, then life could be expected to arise similarly on other Earth-like planets. Unfortunately this argument is also flawed. The reason that Earth was singled out for Sagan’s comment is precisely because Carl Sagan specifically, and human beings in general, are the product of terrestrial biology. Now life takes billions of years to evolve as far as beings like us who can study astrobiology. However, Earth’s ‘habitability window’ is not unlimited. In about another 800 million years, the Sun will be so hot it will boil the oceans, and our planet will become uninhabitable. So there is a finite period of time, roughly 4.5 billion years, in which intelligent life had better emerge on Earth, if it is to emerge at all. But unless life had started promptly after the bombardment, it may never have evolved to the level of intelligence before the habitability window closed. It is therefore no surprise that life ‘popped up’ so quickly—it had to, on account of the fact that we are here! This argument has been placed on a more rigorous and quantitative footing by Carter  and Hanson . Obviously, one cannot draw strong conclusions from a sample of one; the best one can say is that a quick start to life on Earth is consistent both with the cosmic imperative and with the hypothesis that the average time for life to form under Earth-like conditions is in fact very much longer than the age of the Earth.
All this indecisiveness is very discouraging, but there is a glimmer of hope. We might be able to test the cosmic imperative in a more direct way. No planet is more Earth-like than Earth itself, so if life does arise readily in Earth-like conditions, then surely it should have formed many times over right here. Well, how do we know it did not?
3. The shadow biosphere
According to the orthodox picture, all life on Earth descended from a common origin, often expressed, following Darwin, by analogy with a tree. There is strong evidence that all life so far studied in detail is closely inter-related: organisms use a universal genetic code, they all employ nucleic acids to store information and proteins for structural and enzymatic functions. Proteins are made by ribosomes. It is unlikely that so many specific features would have evolved independently from separate origins; rather, they were surely present in a common ancestral organism (often known as LUCA—the last universal common ancestor) and have been retained as frozen accidents. Even the so-called extremophiles—microbes that thrive in conditions that would be lethal to most life that we know—possess the foregoing biochemical features, and share many genes with less exotic organisms. All known extremophiles have been positioned on the same tree of life as you and me.
Nevertheless, it is now apparent that the vast majority of terrestrial species are microbes, and biologists have only just scratched the surface of the microbial realm. Most micro-organisms have not been cultured or characterized, let alone genetically sequenced. At the present time, we simply do not know what they are. One cannot tell by looking whether a microbe is a bacterium or a novel organism with a radically different internal structure and biochemistry. To fully identify a microbe, it is necessary to elucidate its biochemistry and molecular architecture, and to obtain some form of sequence information to position it on the tree of life. It is, therefore, entirely possible that among the billions of microbes contained in, say, a sample of soil or seawater, some are representatives of life as we do not know it—‘weird life’, to use the preferred term. And even if all microbes so far sampled are standard life, there may be unsampled niches, perhaps beyond the reach of even the hardiest extremophiles, that are inhabited by weird micro-organisms. If so, such organisms might be what McKay calls Life 2.0 —the living descendants of an origin of life quite separate from the one that gave rise to standard life .
One scenario for multiple terrestrial origins goes like this. Earth was heavily bombarded by large comets and asteroids for about 700 million years after the formation of the Solar System. The biggest impacts delivered such a large amount of energy that they could have boiled the oceans and sterilized the Earth’s surface . However, those same impacts would have ejected prodigious amounts of material into solar orbit. Suppose life got going on Earth during a quiescent period between sterilizing impacts, for example, in a 10 million year window. Following the next big impact, Earth’s surface was devoid of life, but micro-organisms may have survived within the ejected material. (There is good evidence that microbes are not killed by the shock of planetary ejection .) Cocooned inside rocks, these microbes would have been spared direct exposure to harsh space conditions, and in particular, they would have been shielded from much of the radiation. In a dormant condition, they could have survived for many millions of years . Some fraction of this ejected material will impact Earth eventually, thus returning viable microbes to their planet of origin. Meanwhile, however, Life 2.0 had started, so Earth would then be hosting two forms of life from two independent origins. This process may have been iterated many times. In such a manner, our planet may once have accommodated, and may still accommodate, multiple forms of life, and multiple trees of life.
Remarkably little attention has been paid to the possibility of weird (i.e. non-standard) life on Earth, although astrobiologists have thought a lot about weird life on other planets. Searches for weird terrestrial life fall into two categories. First is the case of ecological separation. Life 1.0 and Life 2.0 might inhabit non-overlapping regions of physical and/or parameter space. Consider, for example, hyperthermophiles. The current upper temperature limit for known hyperthermophiles is 122°C. If a different form of microbial life was detected in a deep ocean volcanic vent system occupying a temperature range of, say, 160–180°C, then it would stand out as a candidate for alternative life because of the discontinuity in the temperature range. A list of extreme environments to search for weird life includes, in addition to ocean volcanic vents, strong UV regions such as the upper atmosphere and high plateaux, regions of extreme cold (Antarctica and mountain tops), aridity (Atacama desert), highly saline or high/low pH aqueous environments, heavily contaminated mining sites and high-radiation environments such as uranium mines and nuclear waste deposits.
Much harder to identify would be weird microbes intermingled with standard life, especially if present at low relative density. Here, two approaches suggest themselves. One could devise a crude filter that would eliminate or at least inhibit the metabolism of standard life in the hope that it would leave any weird life unaffected. Then the weird life would eventually come to predominate. For example, a culture medium laced with a polymer that targets an enzyme like aminoacyl-tRNA synthetase (which attaches specific amino acids to tRNA in conformity with the standard genetic code), and disables it, would stop all standard life in its tracks. Or a polymer loaded with a quantum dot or metallic nanoparticle could be devised to target some specific universal feature of standard life, and then the system irradiated with a laser or microwaves to kill the host cells, but leave any weird cells unscathed.
A second approach would be to make educated guesses about the nature of weird life. Synthetic biologists seek to create novel forms of life in the laboratory, so they are adept at imagining alternative ways that organisms could make a living . The problem about looking for life as we do not know it, is that we do not know quite what to look for. Any general signatures of life, such as carbon cycling or chiral specificity, will be masked by standard life. But if we guess that weird life might exploit a specific molecule, such as an amino acid absent in standard life, then methods could be devised to detect that molecule. An extreme case would be if weird life uses not merely a different suite of amino acids or nucleotides, but a different set of elements. Most biologists think that carbon is essential, but the secondary vital elements are negotiable. One case of interest is phosphorus. According to Wolfe-Simon, arsenic can substitute for phosphorus for many biological functions, and has the added advantage that it offers a redox potential by the reduction of arsenate to arsenite . To be sure, poly-arsenates are far less stable against hydrolysis than their phosphorus counterparts, but in a low-temperature environment that may not be too great a disadvantage. A search for arsenic life has begun, by culturing organisms from arsenic-rich environments such as Mono Lake, in phosphorus-depleted conditions. Successive arsenic enhancements and phosphorus depletions will eventually eliminate all standard life, so that any organisms that survive will have radically new biochemistry.
A final example of a biological filter concerns chirality. Standard life uses left-handed amino acids and right-handed sugars. The laws of physics are, however, indifferent to the chiral signature of organic molecules, and a second genesis might well produce life with the opposite chirality, i.e. right-handed amino acids and/or left-handed sugars. A culture medium made of ‘mirror molecules’ would prove indigestible for standard life, but may be palatable to ‘mirror’ life (i.e. life with reversed chirality). A pilot experiment of this sort was performed by Hoover & Pikuta, and yielded intriguing results that led to the identification of a new class of organisms able to (somehow) metabolize L-sugars . These organisms do not appear, however, to be the sought-after ‘mirror’ life, so the situation clearly involves some subtleties. Nevertheless, the use of chirality as a discriminator between the standard and weird life remains a promising approach.
It is sometimes argued that had more than one form of life existed on Earth, then a ‘winner’ would emerge to displace the rest. But there is no compelling evidence for this assumption. Bacteria and archaea are two genetically very different forms of microbial life that have peacefully coexisted for 3 billion years, even though they are competing for resources in similar niches. Survival strategies widespread in one domain (e.g. methanogenesis among archaea) have not spread to the other. Moreover, microbial species represented in small relative numbers are not observed to be ‘squeezed out’ by the majority, but can remain stable, long-term components of the biosphere as minority players. If an alternative form of life is ecologically separated from standard life, or has reversed chirality, it would not be in direct competition anyway.
4. When is a tree merely a branch?
In the event that a new form of life is discovered, a major challenge will be to determine whether it really started from scratch independently of standard life, i.e. that it represents a genuine example of Life 2.0, or whether it is simply a hitherto undiscovered side branch on the known tree of life. As I have stressed, this issue is of crucial significance to astrobiology. If it can be established that life on Earth has originated more than once, it implies that life will emerge readily in Earth-like conditions, and it will therefore be very likely to arise on other Earth-like planets too. It is exceedingly unlikely that life would have happened twice on Earth but never on all other Earth-like planets. On the other hand, if weird life is merely a highly divergent outlying branch on the same tree as familiar life, we could not draw the momentous conclusion that life is widespread in the universe. At best, the discovery of such a shadow biosphere might extend the parameter range over which extra-terrestrial life is expected to exist. It is therefore of crucial importance to develop criteria for distinguishing a single tree of life from a forest of independent trees.
The most general criterion is biochemical difference. The more the new life differs biochemically from known life, the more plausible it is that we would be dealing with multiple genesis events. If, for example, we found a form of life that used the same amino acids and nucleotides as standard life, but had adopted a different genetic code, it would be plausible to conclude that a common precursor code split into two and evolved apart. For example, it is very likely that the familiar triplet genetic code has evolved from a simpler and more primitive precursor, perhaps a doublet code . It is conceivable that organisms evolved from a doublet code to a variety of triplet codes, many of which may have been eliminated by natural selection because they are in some sense less efficient. It may also be that ancient micro-organisms have survived to the present day, still using the earlier doublet code. Such organisms would not be a genuinely new form of life; rather, they would be ‘living fossils’ occupying a new, deep branch on the known tree of life. By contrast, if we discover ‘mirror life’, i.e. organisms with the opposite chirality from standard life, then a strong case could be made for an independent origin because it is hard to imagine an earlier achiral form of life that splits into left- and right-handed versions because achiral molecules lack the necessary complexity to support biological systems.
I have avoided providing either a general definition of life, or a specific scenario for biogenesis, as they are not necessary for the purposes of the general argument. I have tacitly assumed that life emerged from non-life abruptly, after the fashion of a phase transition in physics, in which case one may meaningfully discuss the concept of a well-defined origin of life at a specific place and time. However, if the transition from non-life to life was a complicated sequence of events extended over time, the notion of multiple genesis events becomes blurred, or even vacuous . Certainly it would be much harder to establish whether a shadow biosphere is descended from an independent second genesis if there is no clear demarcation between the living and non-living realms. Absolute physical isolation would provide a clear cut case, but that is extremely unlikely if both genesis events occurred on Earth. There is, of course, the possibility of a shadow biosphere of extra-terrestrial origin, coexisting with a regular biosphere of terrestrial origin.
It is hard to imagine a discovery of greater significance to astrobiology than a shadow biosphere on Earth consisting of a different form of life descended from an independent origin. Obviously, the existence of such a biosphere is a long shot, but it certainly cannot be ruled out on the basis of our current scientific understanding . If Earth has, or once had, a shadow biosphere, it would then be very likely that life will have arisen on many Earth-like planets across the universe, as many astrobiologists currently assume, but with little or no justification. Because Earth-like planets are likely to be common, life could then be regarded as a truly cosmic phenomenon.
The philosophical ramifications of this shift in viewpoint are immense . So long as we know of only a single sample of life, it is possible to argue that biology is a freak local aberration, the product of a chemical fluke so improbable that it will not have happened anywhere else in the observable universe. Although individual human beings may imbue their lives with significance, life as a whole would be a collection of insignificant freak physical systems, restricted to an infinitesimal patch of the cosmos. By contrast, if life is a ‘cosmic imperative’, emerging more or less automatically in myriad locations, we could say that the universe possesses intrinsically bio-friendly physical laws, so that life, and perhaps mind, could be regarded as having universal significance. They would be fundamental, as opposed to incidental, cosmic phenomena. One could then join with the physicist John Wheeler  and declare that we truly are ‘at home in the universe’.
One contribution of 17 to a Discussion Meeting Issue ‘The detection of extra-terrestrial life and the consequences for science and society’.
- This journal is © 2011 The Royal Society