Science in the Soul

by Richard Dawkins

  AMONG THE MANY dishonesties of the well-financed intelligent-design cabal is the pretence that the designer is not the God of Abraham but an intelligence unspecified, who could equally well be an extraterrestrial alien.*2 Presumably the motive is to circumvent the First Amendment’s prohibition on the establishment of religion, especially following Judge William Overton’s 1982 decision in McLean v. Arkansas Board of Education, in which he struck down the state legislature’s attempt to ensure ‘balanced treatment’ in the schools for ‘creation science’.

  The religious affiliation of these people is not in doubt, and their ingroup communications do not bother to hide their agenda. Jonathan Wells, one of the Discovery Institute’s leading propagandists and the author of Icons of Evolution, is a lifelong member of the Unification Church (the Moonies). He wrote the following testimony in a Moonie in-house journal, under the heading ‘Darwinism: why I went for a second PhD’ (note that ‘Father’ is the Moonies’ name for Reverend Moon himself):

  Father’s words, my studies, and my prayers convinced me that I should devote my life to destroying Darwinism, just as many of my fellow Unificationists had already devoted their lives to destroying Marxism. When Father chose me (along with about a dozen other seminary graduates) to enter a PhD program in 1978, I welcomed the opportunity to prepare myself for battle.

  This quotation alone casts doubt on any claim Wells might have had to be taken seriously as a disinterested seeker after truth – which would seem a fairly minimal qualification for a PhD in science. He publicly admits to undertaking a scientific research degree not in order to discover something about the world but for the specific purpose of ‘destroying’ a scientific idea that his religious leader opposed. Phillip Johnson, the born-again Christian law professor generally regarded as the leader of the gang, openly admits that his motive for opposing evolution is its ‘naturalism’ (as opposed to supernaturalism).

  The claim that the intelligent designer might be an alien from outer space may be disingenuous, but this doesn’t stop it from serving as the basis for an interesting and revealing discussion. Such a constructive discussion, within science, is what I shall undertake in this essay.

  The problem of recognizing an alien intelligence arises, in its starkest form, in that branch of science known as SETI, the Search for Extra-Terrestrial Intelligence. SETI deserves to be taken seriously. Its practitioners are not to be confused with those who complain of having been abducted in flying saucers for sexual purposes. For all sorts of reasons including the reach of our listening devices and the speed of light, it is extremely unlikely that our first apprehension of an alien intelligence will be a corporeal visitation. SETI scientists do not anticipate meeting extraterrestrial visitors in the flesh but in the form of radio transmissions whose intelligent origin should, it is hoped, be evident from their patterning.

  A strong case can be made for the probable existence of intelligent life elsewhere in the universe. It gains support from the principle of mediocrity, that salutary lesson from Copernicus, Hubble and others. Earth was once thought to be the only place in existence, surrounded by crystalline spheres bedecked with tiny stars. Later, when the size of the Milky Way galaxy was understood, it too was thought to be the only place, the locus of all that is. Then Edwin Hubble came along as a latter-day Copernicus to downgrade even our galaxy to mediocrity: it is only one among a hundred billion galaxies in the universe. Today, cosmologists look at our universe and seriously speculate that it may be one of many universes in the ‘multiverse’.

  Similarly, the history of our species was once thought to have been roughly coterminous with the history of everything. Now, to borrow Mark Twain’s crushing analogy, our history’s proportionate duration has shrunk to the thickness of the paint on top of the Eiffel Tower. If we apply the principle of mediocrity to life on this planet, doesn’t it warn us that we would be foolhardy and vain to think that the Earth might be the only site of life in a universe of a hundred billion galaxies?

  It is a powerful argument, and I find myself persuaded by it. On the other hand, the principle of mediocrity is emasculated by another powerful principle, known as the anthropic principle: the fact that we are in a position to observe the world’s conditions determines that those conditions had to be favourable to our existence. The name comes from the British mathematician Brandon Carter, although he later preferred – with good reason – the ‘self-selection principle’. I want to borrow Carter’s principle for a discussion of the origin of life, the chemical event that forged the first self-replicating molecule and hence triggered natural selection of DNA and ultimately all of life. Suppose the origin of life really was a stupendously improbable event. Suppose the accident of primeval-soup chemistry which engendered the first self-replicating molecule was so prodigiously lucky that the odds against it were as low as one in a billion per billion planet years. Such fantastically low odds would mean that no chemist could entertain the smallest hope of repeating the event in a laboratory. The National Science Foundation would laugh in the face of a research proposal whose admitted chance of success was as low as one in a hundred per year, let alone one in a billion per billion years. Yet so great is the number of planets in the universe that even these minuscule odds yield an expectation that the universe contains a billion planets bearing life. And (here comes the anthropic principle) since we manifestly live here, Earth necessarily has to be one of the billion.

  Even if the odds against life arising on a planet are as low as one in a billion billion (which puts it well beyond the range we would classify as possible*3), the plausible calculation that there are at least a billion billion planets in the universe provides an entirely satisfying explanation for our existence. There will still plausibly be one life-bearing planet in the universe. And once we have granted that, the anthropic principle does the rest. Any being contemplating the calculation necessarily has to be on that one life-bearing planet, which therefore has to be Earth.

  This application of the anthropic principle is astonishing but watertight. I have oversimplified it by assuming that once life has originated on a planet, Darwinian natural selection will lead to intelligent and reflective beings. To be more precise, I should have been talking about the combined probability of life’s originating on a planet and leading, eventually, to the evolution of intelligent beings capable of anthropic reflection. It could be that the chemical origin of a self-replicating molecule (the necessary trigger for the origin of natural selection) was a relatively probable event but later steps in the evolution of intelligent life were highly improbable. Mark Ridley, in Mendel’s Demon (confusingly rebranded in America as The Cooperative Gene), suggests that the really improbable step in our kind of life was the origin of the eucaryotic cell.*4 It follows from Ridley’s argument that huge numbers of planets are home to something like bacterial life but only a tiny fraction of planets will have made it past the next hurdle to a level equivalent to the eucaryotic cell – what Ridley calls complex life. Or one might take the view that both those hurdles were relatively easy and that the really difficult step for terrestrial life was the attainment of the human level of intelligence. In this view, we would expect the universe to be rich in planets housing complex life but perhaps with only one planet harbouring beings capable of noticing their own existence and therefore of invoking the anthropic principle. It doesn’t matter how we distribute our odds among these three ‘hurdles’ (or indeed other hurdles, such as the origin of a nervous system). So long as the total odds against a planet’s evolving a life form capable of anthropic reflection do not exceed the number of planets in the universe, we have an adequate and satisfying explanation for our existence.

  Although this anthropic argument is entirely watertight, my strong intuitive feeling is that we do not need to invoke it. I suspect that the odds in favour of life’s arising and subsequently evolving intelligence are sufficiently high that many billions of planets do indeed contain intelligent life forms, many of them so superior to ours tha
t we might be tempted to worship them as gods. Fortunately or unfortunately, we very likely won’t encounter them: even such apparently high estimates still leave intelligent life marooned on scattered islands, which might well be, on average, too far apart for their inhabitants ever to visit one another. Enrico Fermi’s famous rhetorical question ‘Where are they?’ could receive the disappointing answer: ‘They are all over the place but too widely spaced to meet.’ Nevertheless, my belief, for what it’s worth, is that the odds against intelligent life are nothing like as great as the anthropic calculation allows us to countenance. And therefore I think it is well worth putting quite a lot of money into SETI. A positive result would be an exhilarating biological finding, equalled in the history of biology perhaps only by Darwin’s discovery of natural selection itself.

  If SETI ever does pick up a signal, it will likely be from the high, or godlike, end of the spectrum of cosmic intelligences.*5 We shall have a huge amount to learn from the aliens, especially about physics, which will be the same for them as it is for us although they’ll know much more about it. Biology will be very different, though – just how different will be a fascinating question. Communication will be all one-way. If Einstein is right about the limiting speed of light, dialogue will be impossible. We may learn from them, but we won’t be able to tell them about us in return.

  How, then, would we recognize intelligence in a pattern of radio waves picked up by a giant parabolic dish and known to originate from deep space and not be a hoax? A tentative candidate was the pattern first detected by Jocelyn Bell Burnell in 1967 and jokingly called by her the LGM (Little Green Men) signal. This rhythmic pulse, with a periodicity of just over one second, is now known to have been a pulsar; indeed, hers was the first discovery of a pulsar. A pulsar is a neutron star rotating on its own axis with a beam of radio waves sweeping round and round as if from a lighthouse. That a star could rotate with a ‘day’ measured on a scale of seconds is an extremely surprising fact – not the only surprising fact about neutron stars. But for present purposes the important point is that the periodicity of Bell Burnell’s signal is not an indicator of intelligent origin but an unaided product of ordinary physics. Plenty of very simple physical phenomena, from dripping water to pendulums of all kinds, are capable of yielding rhythmic patterns.

  What next might occur to a SETI researcher as diagnostic of intelligent life? Well, if we assume that the aliens actively want to signal their presence, we can ask what we would do if we were trying to transmit evidence of our intelligent presence. Certainly not emit a rhythmic pattern like Bell Burnell’s LGM signal, but what else? Several people have suggested prime numbers as the simplest kind of signal that could originate only from an intelligent source. But how confident should we be that a pattern of pulses based on prime numbers could come only from a mathematically sophisticated civilization? Strictly speaking, you can’t prove that there is no inanimate physical system capable of generating prime numbers. You can say only that no physicist has ever discovered a non-biological process capable of generating them. Strictly speaking, the same caution goes for any signal. However, there are certain kinds of signals – of which those based on prime numbers may be the simplest example – which would be so convincing as to leave alternatives looking absurd.

  Disquietingly, biologists have proposed models that are capable of generating prime numbers but do not involve intelligence. Periodical cicadas emerge for breeding every seventeen years (in some varieties) or every thirteen years (in other varieties). Two theories to account for this odd periodicity depend on the fact that 13 and 17 are prime numbers. I’ll describe just one of these theories. Its premiss is that plague-year breeding is an adaptation to foil predators by swamping them. But then predator species evolved their own periodic breeding pattern to cash in on cicada plagues (or bonanzas, as they would see it). In an evolutionary arms race, the cicadas ‘replied’ by lengthening the period between plague years. The predators lengthened theirs in response. (Remember that this shorthand language of ‘reply’ and ‘response’ implies no conscious decisions, only blind natural selection.) When in the course of the arms race the cicadas reached an interval, such as six years, which was divisible by some other number, the predators found it more profitable to drop their breeding interval, to, say, three years, thereby hitting the cicada bonanza with alternate peaks of their own breeding cycle. Only when the cicadas hit a prime number did this become impossible. The cicadas continued to lengthen their stride until they reached a number that was too large to allow the predators to synchronize directly, yet prime and therefore impossible to meet with some multiple of a smaller period.

  Well, that may not seem a very plausible theory, but it doesn’t need to be for my purpose. I simply need to show that it is possible to conceive of a mechanistic model that does not involve conscious mathematics yet still manages to generate prime numbers. The cicada example shows that while prime numbers may not be generatable by non-biological physics, they can be generated by non-intelligent biology. Even the implausible cicada story is a cautionary tale to warn us that at least it is not necessarily obvious that prime numbers are diagnostic of intelligence.

  The difficulty of diagnosing intelligence in a radio signal is itself a cautionary tale that calls to mind the historical analogy of the argument from design. There was a time when everybody (with a very few, very distinguished exceptions, such as David Hume) thought it completely obvious that the complexity of life was unmistakably diagnostic of intelligent design.*6 What should give us pause is this: Darwin’s nineteenth-century contemporaries could claim the right to be as surprised by his remarkable discovery as we should be surprised today if a physicist discovered an inanimate mechanism capable of generating prime numbers. Perhaps we should entertain the possibility that other principles, comparable to Darwin’s, remain to be discovered – principles capable of mimicking an illusion of design as convincing as the illusion manufactured by natural selection.

  I am not inclined to predict any such event. Natural selection itself, properly understood, is powerful enough to generate complexity and the illusion of design to an almost limitless extent. Bear in mind that elsewhere in the universe there could be variants of natural selection that, although based on essentially the same principle as Darwin discovered on this planet, might be almost unrecognizably different in detail. Bear in mind, too, that natural selection can midwife other forms of design. It doesn’t stop with its direct productions, such as feathers, ears and brains. Once natural selection has produced brains (or some extraterrestrial equivalent of brains), those brains can go on to produce technology (extraterrestrial equivalents of technology), including computers (or extraterrestrial equivalents of them) which are, like brains, capable of designing things. The manifestations of deliberate engineering design – indirect rather than direct productions of natural selection – can burgeon into new reaches of complexity and elegance. The point here is that natural selection manifests itself in the form of design at two levels: there is, first, the illusion of design, which we see in a bird’s wing or a human eye or brain; and, second, there is ‘true’ design, which is a product of evolved brains.*7

  And now to my central point. There really is a profound difference between an intelligent designer who is the product of a long period of evolution, whether on this planet or a distant one, and an intelligent designer who just happened, without any evolutionary history. When a creationist says that an eye or a bacterial flagellum or a blood-clotting mechanism is so complex that it must have been designed, it makes all the difference in the world whether the ‘designer’ is thought to be an alien produced by gradual evolution on a distant planet or a supernatural god who didn’t evolve. Gradual evolution is a genuine explanation, which really can theoretically yield an intelligence of sufficient complexity to design machines and other things too complex to have come about by any process other than design. Hypothetical ‘designers’ jumped up from nothing cannot explain anything, because they can’t e
xplain themselves.

  There are some man-made machines that commonsense, if not strict logic, tells us could not have come about by any process other than intelligent design. A jet fighter, a moon rocket, a motorcar, a bicycle – these are surely intelligently designed. But the important point is that the entity that did the designing – the human brain – is not. There is overwhelming evidence that the human brain evolved through a graded series of almost imperceptibly improving intermediates, whose relics may be seen in the fossil record and whose analogues survive all around the animal kingdom. Moreover, Darwin and his twentieth- and twenty-first-century successors have provided us with a luminously plausible explanation for the mechanism that propels evolution up the graded slopes, the process I have dubbed ‘Climbing Mount Improbable’. Natural selection is not some desperate last resort of a theory. It is an idea whose plausibility and power hit you between the eyes with stunning force, once you understand it in all its elegant simplicity. Well might T. H. Huxley cry out, ‘How extremely stupid of me not to have thought of that!’

  But we can go further. Not only does natural selection explain the bacterial flagellum, the eye, the feather, and brains capable of intelligent design. Not only can it explain every biological phenomenon ever described. It is the only plausible explanation for these things that has ever been proposed. Above all, the argument from improbability – the very argument that the advocates of intelligent design fondly imagine supports their case – turns around to kick that case over with devastating force and lethal effect.

  The argument from improbability states, incontrovertibly, that some phenomenon in nature – something like a bacterial flagellum, say, or an eye – is too improbable to have simply happened. It has to be the product of some very special process that generates improbability. The mistake is to jump to the conclusion that ‘design’ is that very special process. In fact it is natural selection. The late Sir Fred Hoyle’s jocular analogy of the Boeing 747 is useful, although it, too, turns out to make the opposite point to the one he intended. The spontaneous origin of the complexity of life, he said, is as improbable as a hurricane blowing through a scrapyard and spontaneously assembling a Boeing 747. Everybody agrees that airliners and living bodies are too improbable to be assembled by chance. A more precise characterization of the kind of improbability we are talking about is specified improbability (or specified complexity). The ‘specified’ is important, for reasons that I explained in The Blind Watchmaker. I began by pointing out that randomly hitting the number that opens the large combination lock on a bank vault is improbable in the same sense as hurling scrap metal around and assembling an airliner: