The Fear of God

Review of The Fifth Miracle: The Search for the Origin of Life, by Paul Davies (1999)


Phillip E. Johnson

A condensed version of this review essay was published in the Weekly Standard for March 22, 1999. This is the original manuscript version.

Paul Davies is a professor of mathematical physics, a highly successful scientific popularizer, and the winner of the 1996 Templeton Prize of over one million dollars for "progress in religion."  He has said in interviews that the book that set him on his career path was Anglican Bishop John T. Robinson's Honest to God, a 1960s manifesto of theological liberalism that also inspired such religious reformers as Episcopal Bishop John Shelby Spong to attempt to save Christianity by removing its supernatural elements. Davies explained to one reporter that "I was drawn to [Robinson's] idea of God as a sort of 'timeless ground of being' on which the cosmic order is built. Since science proceeds from the assumption that nature unfailingly obeys rational mathematical laws, these laws must be rooted in something. But not a cosmic magician!"

If Bishop Spong means to save Christianity by secularizing it, Paul Davies wants to save science from the bleak reductionism encapsulated in physicist Steven Weinberg's oft-quoted remark that "the more the universe seems comprehensible the more it seems pointless."  He describes his own middle ground between supernaturalism and materialism as "a vision of a self-organizing and self- complexifying universe, governed by ingenious laws that encourage matter to evolve towards life and consciousness."  In The Fifth Miracle, Davies takes this vision from cosmology into evolutionary biology, thus bringing a fresh viewpoint to a field that has long been dominated by  materialists who insist that no purposeful forces guided the evolution of life.

The fifth miracle of Davies' title refers to Genesis 1:11: "Let the Land Produce Vegetation." (The first four Biblical miracles are the creation of the universe, the creation of light, the creation of the firmament and the creation of dry land.)  It is proverbial in the popular science publishing world that God is good for sales, especially since Steven Hawking sold millions of copies of an otherwise unremarkable book by promising that a unified physical theory would enable us "to know the mind of God." Commercial requirements alone seem to have dictated that word "miracle," since Davies begins the book by disavowing it. Like other evolutionary scientists he starts with the presumption that "it is the job of science to solve mysteries without recourse to divine intervention."  Life is not a miracle because scientists wish it to be a product of natural forces which they can explain.

If the origin of life was not a miracle, Davies does think that it was a very mysterious event.  Not long ago he thought that science was close to solving the mystery, but upon investigating the subject to write this book he became convinced that "we are missing something fundamental about the whole business."  A satisfactory theory of the origin of life requires not just more knowledge of the kind we already possess, but "some radically new ideas." So what is the fundamental thing that scientists are currently missing, and what kind of radically new ideas does Davies have in mind?

At the middle of the twentieth century, the reigning belief was that life began with an immensely improbable chance event.  This view was dramatically stated in a Scientific American article in 1954 by the Harvard biochemist George Wald.  Wald conceded that the spontaneous generation of something as complex as a living organism seemed impossible, but he insisted that such statistical miracles are possible and even probable given enough time.  He estimated that two billion years were available, and he argued that "Given so much time, the 'impossible' becomes possible, the possible probable, and the probable virtually certain. One has only to wait: time itself performs the miracles."

Wald's view that chance and time are all that is necessary is now out of favor.  Today's dominant view is most authoritatively stated in Nobel laureate Christian de Duve's 1995 book, Vital Dust: Life as a Cosmic Imperative (1995).  De Duve argues that life is not the product of chance but of law-driven chemical steps, each one of which must have been highly probable in the right circumstances.  This reliance upon laws favoring life is a giant step in the direction Davies wants to go, and  away from the view that the origin of life was a freakish accident unlikely to occur elsewhere.  George Wald's position has dropped from sight in part because it has become clearer that even the simplest conceivable life form (still much simpler than any actual organism) would have to be so complex that accidental self-assembly would be practically miraculous even in two billion years.  (Natural selection can't help until biological reproduction has started,  a point to which I will return in a moment.)   In addition, the time available for a statistical miracle to occur has been dramatically shrinking. Evidence of life goes back about 3.8 billion years, almost to a time when conditions on the early earth were inhospitable for even the hardiest bacteria. If life evolved in a geological instant, it must have been by law-directed chemical steps that would be likely to occur again in similar conditions.

That is why many scientists, including both de Duve and Davies, are confident that some kind of unicellular life (or evidence of past life) is likely to be found on other planets wherever conditions are sufficiently favorable. Whether there is any more extensive "cosmic imperative" that would compel relatively simple organisms to evolve in the direction of human-style intelligence is more doubtful.  But our own galaxy alone has some 200 billion stars, and if a few million of these have planets bearing life, there are lots of opportunities for Darwinian evolution to bring life to a stage of consciousness and intelligence.  Maybe the laws that make primitive life inevitable also make intelligent life sufficiently likely that the necessary evolutionary steps would have happened many times.

So far Davies and de Duve are in agreement. They part company over whether existing chemical laws are sufficient to explain the origin of life, or whether something essentially different remains to be discovered.  Orthodox prebiological chemists, including de Duve, see the problem as one of conventional chemistry.  Once the right chemicals are in place at the right time, the necessary reactions inevitably follow and life emerges.  Hence the experiments that seem important to them are ones that show that some of the necessary chemicals (mainly amino acids) could have been synthesized on the early earth, or could have come to this planet on comets or meteorites. They concede that many of the specific steps leading to life remain to be explained, but they are confident that the picture can be filled in eventually on the basis of the known laws of chemistry, supplemented perhaps by chance events that are not forbiddingly unlikely.  Chance and law had to do the whole job, because nothing else was available.

For Davies, the solution to the riddle of life lies not in just getting the chemicals together, but in explaining the origin of the genetic information, which he calls the "software" of the organism.  A living cell is a masterpiece of miniaturized complexity, and its complex operations are coordinated by a program or "blueprint" inscribed in the 4-letter chemical alphabet of the DNA, and then translated into the twenty-letter alphabet of the proteins.  What most needs to be explained is not the chemicals but the information in the software, just as the important thing about a computer program or a book is the information content and not the physical medium in which that information is recorded.  Davies believes that, once the life processes of metabolism and reproduction are under way, natural selection can supply increases in information.  But metabolism and reproduction can not get started until an enormous amount of complex information is already in existence.  What was the source of the initial information input?

That genetic information exists, and is enormously complex, is not controversial.  The arch-materialist Richard Dawkins states, at the beginning of The Blind Watchmaker, that "Biology is the study of complicated things that give the appearance of having been designed for a purpose," and that each cell "contains a digitally coded database larger, in information content, that all 30 volumes of the Encyclopedia Britannica put together."  Materialists assume that this information is an emergent product of chemistry, and that it somehow forms when the right chemical combinations get going from the right combination of chance and law.

Davies says, however, that the leap from chemistry to biology requires something in addition to chance and law, because of the fundamentally informational character of life.  Law produces the same simple pattern over and over again.  Chance produces disordered, unspecified sequences that show no consistent patterns.  The problem is that the genetic information, like the information in Windows 98 or the Bible, is both highly specified and random (i.e., non-repeating).  These characteristics are essential for any book or program with a high information content, and explain why the nucleotide sequences of an organism's DNA necessarily must be independent of any chemical properties that cause the parts of the molecule to bind together.  A book whose letters reflected only the chemical properties of ink and paper would express no information beyond what is already contained in the laws.

The heart of the problem is that physical laws are simple and general, and by their nature they produce only repetitive order.  Law-governed processes can produce simple repetitious patterns, as in crystals, but they can't produce the complex, specified  sequences by which the nucleotides of DNA code for proteins -- any more than they can produce the sequence of letters on a page of the Bible.   Random sequences, on the other hand, are by definition non- patterned.  To say in this context that sequences are random means that they are non-repeating and hence cannot be produced by a formula such as  "do X over and over again."  A random assortment of letters also contains no significant information unless the sequence is also specified by some independent requirement.  Again, think of your computer's operating system or the Bible as an example.  Only a very small number of  highly specific sequences of instructions will give you a working program, or an intelligible Bible.  Random deviations from this specified sequence will introduce disorder, and the situation will only be worse if you add recurring patters of mindless repetitions governed by invariable laws.

In short, meaningful sequences require some third force that works both against repetitive order on the one hand, and chaotic chance on the other. Mixing the two together just gives us the worst of both worlds.  Here is a pastiche of Davies' sentences from his concluding chapter to give a flavor of the kind of third force he has in mind:

"A law of nature of the sort we know and love will not create biological information, or indeed any information at all. (210) ... The whole point of the genetic code, for example, is to free life from the shackles of non-random chemical bonding.  (211) ... The key step that was taken on the road to biogenesis was the transition from a state in which molecules slavishly follow mundane chemical  pathways, to one in which they organize themselves to follow their own pathways. (211) ... Once this essential point is grasped, the real problem of biogenesis is clear.  Since the heady success of molecular biology, most investigators have sought the secret of life in the physics and chemistry of molecules.  But they will look in vain for conventional physics and chemistry to explain life, for this is a classic case of confusing the medium with the message.  The secret of life lies, not in its chemical basis, but in the logical and informational rules it exploits. (212) ...  Real progress with the mystery of biogenesis will be made, I believe, not through exotic chemistry, but from something conceptually new. (216)."

But exactly what is this "something conceptually new?"  Davies understood that many scientists would think he was describing "a miracle in nature's clothing."  He appealed to de Duve himself for protection, saying that "Deterministic thinking, even in the weaker forms of de Duve and [Stuart] Kauffmann, represents a fundamental challenge to the existing scientific paradigm.... Although biological determinists strongly deny that there is any actual design, or predetermined goal, involved in their proposals, the idea that the laws of nature may be slanted towards life, while not contradicting the letter of Darwinism, certainly offends its spirit.  It slips an element of teleology back into nature, a century and a half after Darwin banished it." (218-219).  

I heard Davies present this thesis at a scientific conference in Italy in September 1998, at which Christian de Duve and I were also participants.  I thought Davies had walked right up to the brink of saying that an intelligent agent had participated in the origin of life, and evidently de Duve thought so too.  He accosted Davies immediately afterwards, and continued with probing questions into the evening.  De Duve, as remorseless in his logic as he was courteous in his manner, asked Davies if he was implying that information came first, and chemistry only thereafter. Davies answered that no, he didn't mean that.*  So it seemed that chemical laws created the information after all.  De Duve continued: "You are reviving vitalism and [Aristotelian] final causes!"  Again, Davies again pleaded not guilty.  He assured de Duve that he hadn't really meant that the new laws would contradict the existing laws.  After a lot more of this de Duve smiled benignly and said "I must have misunderstood you." 

Put to the test, it seemed that Davies was harmless after all.  All his revolutionary talk just meant that we don't have all the answers yet, which is what de Duve and the orthodox people have been saying all along. Everybody agrees that new knowledge is needed.  The argument Davies had seemed to be making was that we need to discover some fundamentally new third factor that is beyond both chance and law.  Then he seemed to flinch when he contemplated the consequences of his own logic.  He was caught between what he wanted to say (chance and law aren't enough) and what he didn't quite dare to say (something beyond chance and law had to be involved).  His awkward solution was to call the necessary third entity another kind of  "law," even though he acknowledged that it would have to be something fundamentally different from anything previously known as a law.

One reason Davies had to retreat was that he risked being labeled as a vitalist or even as a creationist.  It is tolerable for a cosmologist to say that the laws of physics are rooted in some rational principle, because that implies only a deism that leaves the laws inviolate after the ultimate beginning.  It is quite another thing to say that some information-creating intelligence was involved in the origin of life, billions of years after the Big Bang.  Regardless of any disclaimers Davies might make, an intelligent force which operates in the history of life would be seen by scientific materialists as a cosmic magician and by the conventionally religious as the God of the Bible.

Davies also had a scientific reason for hesitating to commit himself to the need for a third factor. He believes that evolutionary biologists have proved that, once the life process has somehow been jump-started, the Darwinian mechanism can do all rest of the information-building. If that is true, then it is reasonable to agree with de Duve that there must be some quasi-Darwinian process operating in the prebiotic environment.  Quoting again from The Fifth Miracle:

"Can specific randomness be the guaranteed product of a deterministic, mechanical,  law-like process, like a primordial soup left to the mercy of familiar laws of physics and chemistry?  No it couldn't.... If you have found the foregoing argument persuasive, you may be forgiven for concluding that a genome really is a miraculous object. However, most of the problems I have outlined above apply with equal force to the evolution of the genome over time.  In this case we have a ready-made solution to the puzzle, called Darwinism.  Random mutations plus natural selection are one sure-fire way to generate biological information, extending a short random genome over time into a long random genome. Chance in the guise of mutations and law in the guise of selection form just the right combination of randomness and order needed to created 'the impossible object.' The necessary information comes, as we have seen, from the environment. (P. 89)

That amounts to saying that the law/chance combination can do the job after all when the package is labeled Darwinism, and that the Darwinian magician can draw the informational rabbit out of the environmental hat when the rabbit never was in the hat in the first place!  We are entitled to ask for experimental confirmation of so marvelous a tale, and of course it won't be forthcoming.  The standard examples of Darwinism-in-action involve only cyclical variations in fundamentally stable populations of peppered moths and finches. Random mutation is typically an information- reducing entity, even if there are rare exceptions.  Some mutations are fitness enhancing, to be sure, as when a bacterium becomes resistant to an antibiotic, but this does not necessarily mean that they increase genetic information. Lee Spetner's book Not By Chance (Judaica Press 1998) explains that a mutations that causes a bacterium to become resistant to streptomycin, for example, is information-reducing.  It disables a site on a ribosome where the drug normally attaches, and so the drug molecule can no longer do its damage to the bacterium.  By analogy, a random change in an elaborate computer program might on rare occasions improve the computer's performance; if it disabled some feature that was causing trouble in a particular environment.  But that is not how computers and their software are built up in the first place. In fact biologists believe in the creative power of the mutation/selection mechanism for exactly the same reason that prebiological chemists like de Duve believe that chemical reactions can create genetic information.  They are philosophical materialists, and identify science with that philosophy, and so they assume that nothing other than law and chance was available to do all the creating that had to be done.

One reason that many scientists think that "materialism" and "science" are just two words for the same thing is that they assume that the only alternative to law and chance is miracle, by which they mean a cosmic magician's arbitrary interference with the stately order of natural law. But why should they assume that?  If we go back to the computer and its software as an example, it is evident that intelligent design is also part of the natural order.  A computer does not operate by magic, nor does it contravene the laws of physics and chemistry.  Its operations are within the laws and as predictable as other systems which scientists study, even though a computer does not come into existence until an intelligent entity designs its hardware and software. If genetic information is comparable to software, it must be designed by an entity with the capability of a software designer.  That's just a fact, and science doesn't progress by denying facts in order to take refuge in comfortable philosophical assumptions.

Even the most unyielding Darwinists seem to have some sense that their biological mechanism is inadequate.  In his 1998 book Unweaving the Rainbow, Richard Dawkins castigates his rival Stephen Jay Gould for misleading the public by using "bad poetry" to describe evolution.  Yet Dawkins himself employs dubious metaphors to make his points, especially when he tries to illustrate the power of  natural selection by comparing it to a computer running a program.  He does this most famously in Chapter Three of The Blind Watchmaker, where he explains how a computer can write a text of Shakespeare by selecting the right letters from a random array. Many lesser science writers have followed his example.  I found a succinct version of this bad poetry recently in an article by the science editor of the London Guardian, announcing a public lecture by Dawkins:

"Churchmen used to argue that self-creating life was as likely as a monkey randomly batting typewriter keys and typing out the Bible without a mistake. But think of monkeys and typewriters Darwin's way, says Professor Cesare Emiliani of Miami. It might take a monkey an eternity to type the 6 million characters of the Bible by chance, but suppose natural selection were a rubber that erased each mistyped letter - or each unsatisfactory mutation - immediately? Assume 13 mistakes for each successful letter, and at the rate of a keystroke per second, you could have the whole of Holy Writ in 13 times 6 million seconds, or two and a half years." [Tim Radford, "And Darwin created us all. . .;" The Guardian (London), February 6, 1999, P. 1.]

Of course this absurdity has nothing to do with the natural selection known to biologists, which amounts to nothing more than the unremarkable observation that organisms may fail to leave viable offspring due to genetic inadequacies, in which case the "bad genes" will tend to disappear from the population.   Darwin made the observation remarkable by calling it "selection," a metaphor which misleadingly suggests intelligent choice. Metaphor aside, death and sterility do not create genetic information or write Bibles. Professor Emiliani's "eraser" would have to be a computer with the Bible already in its memory, so it could compare the letters typed by the monkey with those in its copy of the Bible to see which letters were correct.  It could more easily dispense with the sham of the random letters and just find the file and hit the "print" key.  The hero of the story is not the monkey or the piece of rubber, but the software designer, who put the Bible in the eraser's memory and programmed it with the ability to spot the meaning an apparently meaningless letter would eventually have when the rest of the text was in place.  Attributing all opposition to the Darwinian scenario to "churchmen," by the way, is an all-too-typical Darwinian appeal to prejudice.

When used to demonstrate what natural selection can do, the analogy to a computer is either a howler or a fraud.  I don't say that in order to complain, but to point out why Darwinists have to resort to such bad poetry to defend their system.  The reason is that, once the problem of biological evolution is framed as "information creation" rather than "variation within the type," the inadequacy of the peppered moth and finch beak examples is glaringly obvious. Computers with intelligently designed programs become the basis of thought-experiments, because the Darwinists at some level understand that their mechanism cannot do the necessary information-creating unless it has the capabilities of Professor Emiliani's eraser.

I hope that Paul Davies will think again about how the materialists have misled him, and find the courage to follow his own insights to their logical conclusion.  Law and chance are not enough to create the thing that makes life what it is, which is the software bearing the information. A third factor is needed. Whether we call it "intelligence" or find some euphemism like "new laws," the third factor has to have the capacities we normally associate with intelligence. If the materialists want to believe that a mechanism involving only law and chance can do the job, they have every right to try to pursue a research strategy based on that belief.  But science does not mean believing what you want to believe, or holding fast to a philosophy regardless of the evidence. It means having the boldness to follow the trail of the evidence, even if it threatens to lead you into unfamiliar territory.

* But maybe he did mean just that. In his article "Bit before it?" (New Scientist, January 30, 1999, Pg. 3), Davies wrote that "Normally we think of the world as composed of simple, clod-like, material particles, and information as a derived phenomenon attached to special, organised states of matter. But maybe it is the other way around: perhaps the Universe is really a frolic of primal information, and material objects a complex secondary manifestation."