In a shaft of sunlight, even when the air is still, you can sometimes see a tribe of dust motes dancing. They move in zigzag paths as if animated, motivated, propelled by some small but earnest purpose. Some of the followers of Pythagoras, the ancient Greek philosopher, thought that each mote had its own immaterial soul that told it what to do, just as they thought that each human has a soul that gives us direction and tells us what to do. Indeed, the Latin word for soul is anima—it is something similar in many modern languages—from which come such English words as “animate” and “animal.”

In fact, those motes of dust make no decisions, have no volition. They are instead the passive agents of invisible forces. They’re so tiny that they’re battered about by the random motion of molecules of air, which have a slight preponderance of collisions first on one side of the mote and then on the other, propelling them, with what looks to us as some mix of intention and indecision, through the air. Heavier objects—threads, say, or feathers—cannot much be jostled by molecular collisions; if not wafted by a current of air, they simply fall.

The Pythagoreans deceived themselves. They did not understand how matter works on the level of the very small, and so—from a specious and oversimple argument—they deduced a ghostly spirit that pulls the strings. When we look around us at the living world, we see a profusion of plants and animals, all seemingly designed for specific ends and single-mindedly devoted to their own and their offspring’s survival—intricate adaptations, an exquisite match of form to function. It is natural to assume that some immaterial force, something like the soul of a dust mote, but far grander, is responsible for the beauty, elegance, and variety of life on Earth, and that each organism is propelled by its own, appropriately configured, spirit. Many cultures all over the world have drawn just such a conclusion. But might we here, as did the ancient Pythagoreans, be overlooking what actually goes on in the world of the very small?

We can believe in animal or human souls without holding to evolution, and vice versa. But if we examined life more closely, might we be able to understand at least a little of how it works and how it came to be, purely in terms of its constituent atoms? Is something “immaterial” present? If so, is it in every beast and vegetable, or just in humans? Or is life no more than a subtle consequence of physics and chemistry?

——

One educated look at how the molecule is shaped and you can figure out what it’s for. Even at the molecular level, function follows form. Before us is a detailed blueprint of breathtaking precision for building complex molecular machines. The molecule is very long and composed of two intertwined strands. Running the length of each strand is a sequence made of four smaller molecular building blocks, the nucleotides—which humans conventionally represent by the letters A, C, G, and T. (Each nucleotide molecule actually looks like a ring, or two connected rings, made of atoms.) On and on the sequence goes, for billions of letters. A short segment of it might read something like this:

Along the opposite strand there’s an identical sequence, except that wherever nucleotide A was in the first strand, it’s T in the second; and instead of G it’s always C. And vice versa. Like this:

This is a code, a long sequence of words written out in an alphabet of only four letters. As in ancient human writing, there are no spaces between the words. Inside this molecule there are, written in a special language of life, detailed instructions—or rather, two copies of the same detailed instructions, because the information in one strand can surely be reconstructed from the information in the other, once you understand the simple substitution cipher. The message is redundant, bespeaking care, conservatism; it conveys a sense that whatever it is saying must be preserved, treasured, passed intact to future generations.

Almost every issue of leading scientific journals such as Science or Nature contains the newly uncovered ACGT sequence of some part of the genetic instructions of some lifeform or other. We’re slowly beginning to read the genetic libraries. The library of our own hereditary information, the human genome, is also becoming increasingly revealed, but there’s a lot to read: Every cell of your body has a full set of instructions about how to manufacture you, encoded in a very compressed format—it takes only a picogram (a trillionth of a gram) of this molecule to specify everything you’ve inherited from your ancestors, back to the first beings of the primeval sea. Yet, there are almost as many nucleotide building blocks, or “letters,” in the microminiaturized genetic information in any of your cells as there are people on Earth.

All words in the genetic code are three letters long. So, if we insert the implicit spaces between the words, the beginning of the first message above looks like this:

Since there are only four kinds of nucleotides (A, C, G, and T), there are at most only 4 × 4 × 4 = 64 possible words in this language. But if the order in which the words are put together is central to the meaning of the message, you can say a great deal with only a few dozen different words. With messages that are a billion carefully selected words long, what might be possible? You must take care in reading the message, though: With no spaces between the words, if you start reading at the wrong place, the meaning will surely change and a lucid message might be reduced to gibberish. This is one reason the giant molecule has special code words meaning “START READING HERE” and “STOP READING HERE.”

As you watch the molecule closely you observe that the two strands occasionally unwind and unzip. Each copies the other, using available A, C, G, and T raw materials—like the metal type stored in an old-fashioned printer’s box Now, instead of one pair, there are two pairs of identical messages. As well as utilizing a language and embodying a complex, redundantly encoded text, this molecule is a printing press.

But what’s the use of a message if nobody reads it? Through copying links and relays, the sequences of As, Cs, Gs, and Ts are revealed to be the job orders and blueprints for the construction of particular molecular machine tools. Some sequences are orders to itself—arranging for the giant molecule to twist and kink so it can then issue a particular set of instructions. Other sequences ensure that the instructions will be followed to the letter. Many three-letter words specify a particular amino acid (or a punctuation mark, like the one that signifies “START”) out there in the surrounding cell, and the sequence of words encoded determines the sequence of amino acids that will make up the protein machine tools that control the life of the cell. Once such a protein is manufactured, it usually twists and folds itself into a three-dimensional shape spring-loaded for action. Sometimes another protein bends it into shape. These machine tools, at a pace determined both by the long double-stranded molecule and by the outside world, then proceed on their own to strip other molecules down, to build new ones up, to help communicate molecular or electrical messages to other cells.

This is a description of some of the humdrum, everyday action in each of the ten trillion or so cells of your body, and those of nearly every other plant, animal, and microbe on Earth. The tiny machine tools perform stupefying feats of molecular transformation. They are submicroscopic and made of organic molecules, rather than macroscopic and made of silicates or steel, but at the molecular level life was tool-using and tool-making from the start.

The long self-replicating double-stranded molecule with the complex message is a sequence of genes, a little like beads on a string. Chemically, it is a nucleic acid (here, the kind abbreviated DNA, which stands for deoxyribonucleic acid). The two strands, wrapped around each other, comprise the famous DNA double helix. The nucleotide bases in DNA are called adenine, cytosine, guanine, and thymine, which is where the abbreviations A, C, G, and T come from. Their names date back to long before their key role in heredity was understood. Guanine, for example, is named unpretentiously after guano, the bird droppings from which it was first isolated. It is a double ring molecule made of five carbon atoms, five hydrogens, five nitrogens, and one oxygen. There’s something like a billion guanines (and roughly equal numbers of As, Cs, and Ts) in the genes of any one of your cells.

Except for some oddball microbes, the genetic information of every organism on Earth is contained in DNA—a molecular engineer of formidable, even awesome talents. One (very long) sequence of As, Cs, Gs, and Ts contains all the information for making a person; another such sequence, nearly identical, for a chimpanzee; others, not so different, for a wolf or a mouse. In turn, the sequences for nightingales, sidewinders, toads, carp, scallops, forsythia, club mosses, seaweed, and bacteria are still more different—although even they collectively hold many sequences of As, Cs, Gs, and Ts in common. A typical gene, controlling or contributing to one specific hereditary trait, might be a few thousand nucleotides long. Some genes may comprise more than a million As, Cs, Gs, and Ts. Their sequences specify the chemical instructions for, say, manufacturing the organic pigments that make eyes brown or green; or extracting energy out of food; or finding the opposite sex.

How this complex information got into our cells, and how arrangements were made for its precise replication and the obedient implementation of its instructions, is tantamount to asking how life evolved. Nucleic acids were unknown when The Origin of Species was first published, and the messages they contain were not to be deciphered for another century. They constitute the demonstration and definitive record of evolution that Darwin sought. Scattered in the ACGT sequences of the diverse lifeforms of our planet is an incomplete history of the evolution of life—not the blood, bones, brains, and the other manufactured products of the genetic factories, but the actual production records, the master instructions themselves, slowly varying at different rates in different beings in different epochs.

Because evolution is conservative and reluctant to tamper with instructions that work, the DNA code incorporates documents—job orders and blueprints—dating back to remote biological antiquity. Many passages have faded. In some places there are palimpsests, where remains of ancient messages can be seen peeking out from under newer ones. Here and there a sequence can be found that is transposed from a different part of the message, taking on a different shade of meaning in its new surrounds; words, paragraphs, pages, whole volumes have been moved and reshuffled. Contexts have changed. The common sequences have been inherited from remote times. The more distinct the corresponding sequences are in two different organisms, the more distantly related they must be.

These are not only the surviving annals of the history of life, but also handbooks of the mechanisms of evolutionary change. The field of molecular evolution—only a few decades old—permits us to decode the record at the heart of life on Earth. Pedigrees are written in these sequences, carrying us back not a few generations, but most of the way to the origin of life. Molecular biologists have learned to read them and to calibrate the profound kinship of all life on Earth.⁵ The recesses of the nucleic acids are thick with ancestral shadows.

We can now almost follow the itinerary of the naturalist Loren Eiseley:

——

A particular sequence of As, Cs, Gs, and Ts is in charge of making fibrinogen, central to the clotting of human blood. Lampreys look something like eels (although they are far more distant relations of ours than eels are); blood circulates in their veins too; and their genes also contain instructions for the manufacture of the protein fibrinogen. Lampreys and people had their last common ancestor about 450 million years ago. Nevertheless, most of the instructions for making human fibrinogen and for making lamprey fibrinogen are identical. Life doesn’t much fix what isn’t broken. Some of the differences that do exist are in charge of making parts of the molecular machine tools that hardly matter—something like the handles on two drill presses being made of different materials with different brand names, while the guts of the two are identical.

Or here, to take another example, are three versions of the same message,⁷ taken from the same part of the DNA of a moth, a fruit fly, and a crustacean:

Moth:

Fruit fly:

Crustacean:

Compare these sequences and recall how different a moth is from a lobster. But these are not the job orders for mandibles or feet—which could hardly be closely similar in moths and lobsters. These DNA sequences specify the construction of the molecular jigs on which newly forming molecules are laid out under the ministrations of the molecular machine tools. Down at this level, it’s not absurd that moths and lobsters might have closer affinities than moths and fruit flies. The comparison of moth and lobster suggests how slow to change, how conservative the genetic instructions can be. It’s a long time ago that the last common ancestor of moths and lobsters scudded across the floor of the primeval abyss.

We know what every one of those three-letter ACGT words means—not just which amino acids they code for, but also the grammatical and lexigraphical conventions employed by life on Earth. We have learned to read the instructions for making ourselves—and everybody else on Earth. Take another look at “START” and “STOP.” In organisms other than bacteria, there’s a particular set of nucleotides that determine when DNA should start making molecular machine tools, which machine tool instructions should be transcribed, and how fast the transcription should go. Such regulatory sequences are called “promoters” and “enhancers.” The particular sequence TATA, for example, occurs just before the place where transcription is to occur. Other promoters are CAAT and GGGCGG. Still other sequences tell the cell where to stop transcribing.⁸

You can see that the substitution of one nucleotide for another might have only minor consequences—you could, for example, substitute one structural amino acid for another (in the “handle” of the machine tool) and in no way change what the resulting protein does. But it could also have a catastrophic effect: A single nucleotide substitution might convert the instructions for making a particular amino acid into the signal to stop the transcription; then, only a fragment of the molecular machine in question will be manufactured, and the cell might be in trouble. Organisms with such altered instructions will probably leave fewer offspring.

The subtlety and nuance of the genetic language is stunning. Sometimes there seem to be overlapping messages using the same letters in the same sequence, but with different functional import depending on how it’s read: two texts for the price of one. Nothing this clever occurs in any human language. It’s as if a long passage in English had two completely different meanings,⁹ something like

ROMAN CEMENT TOGETHER NOWHERE …

and

ROMANCEMENT TO GET HER NOW HERE …

but much better—on and on for pages, perfectly lucid and grammatical in both modes, and, we think, beyond the skill of any human writer. The reader is invited to try.

In “higher” organisms, many long sequences seem to be nonfunctional genetic nonsense. They lie after a “STOP” and before the next “START” and generally remain ignored, forlorn, untranscribed. Maybe some of these sequences are garbled remnants of instructions that, long ago, in our distant ancestors, were important or even keys to survival, but that today are obsolete and useless.* Being useless, these sequences evolve quickly: Mutations in them do no harm and are not selected against. Maybe a few of them are still useful, but elicited only under extraordinary circumstances. In humans some 97% of the ACGT sequence is apparently good for nothing. It’s the remaining 3% that, as far as genetics goes, makes us who we are.

Startling similarities among the functional sequences of As, Cs, Gs, and Ts are seen throughout the biological world, similarities that could not have come about unless—beneath the apparent diversity of life on Earth—there was an underlying and fundamental unity. That unity exists, it seems clear, because every living thing on Earth is descended from the same ancestor 4 billion years ago; because we are all kin.

But how could machines of such elegance, subtlety, and complexity ever arise? The key to the answer is that these molecules are able to evolve. When one strand is making a copy of the other, sometimes a mistake occurs and the wrong nucleotide—an A, say, instead of a G—will be inserted into the newly assembled sequence. Some of them are honest replication errors—good as it is, the machinery isn’t perfect. Some are induced by a cosmic ray or another kind of radiation, or by chemicals in the environment. A rise in temperature might slightly increase the rate at which molecules fall to pieces, and this could lead to mistakes. It even happens that the nucleic acid generates a substance that alters itself—perhaps thousands or millions of nucleotides away.

Uncorrected mistakes in the message are propagated down to future generations. They “breed true.” These changes in the sequence of As, Cs, Gs, and Ts, including alterations of a single nucleotide, are called mutations. They introduce a fundamental and irreducible randomness into the history and nature of life. Some mutations may neither help nor hinder, occurring, for example, in long, repetitive sequences—containing redundant information—or in what we’ve called the handles of the molecular machine tools, or in untranscribed sequences between STOP and START. Many other mutations are deleterious. If you’re crafting superb machine tools and, while you’re not looking, someone introduces a few random changes into the computer instructions for manufacture, there isn’t much chance that the resulting machines, built according to the new, garbled instructions, will work better than the earlier model. Enough random changes in a complex set of instructions will cause serious harm.

But a few of the random changes, by luck, prove advantageous. For example, the sickle-cell trait we mentioned in the last chapter is caused by the mutation of a single nucleotide in the DNA, generating a difference of a single amino acid in the hemoglobin molecules that nucleotide helps code for; this in turn changes the shape of the red blood cell and interferes with its ability to carry oxygen, but at the same time it eventually kills the plasmodium parasites those cells contain. A lone mutation, one particular T turning into an A, is all it takes.

And, of course, not just the hemoglobin in red blood cells, but every part of the body, every aspect of life, is instructed by a particular DNA sequence. Every sequence is vulnerable to mutation. Some of these mutations cause changes more far-reaching than the sickle-cell trait, some less. Most are harmful, a few are helpful, and even the helpful ones may—like the sickle-cell mutation—represent a tradeoff, a compromise.

This is a principal means by which life evolves—exploiting imperfections in copying despite the cost. It is not how we would do it. It does not seem to be how a Deity intent on special creation would do it. The mutations have no plan, no direction behind them; their randomness seems chilling; progress, if any, is agonizingly slow. The process sacrifices all those beings who are now less fit to perform their life tasks because of the new mutation—crickets who no longer hop high, birds with malformed wings, dolphins gasping for breath, great elms succumbing to blight. Why not more efficient, more compassionate mutations? Why must resistance to malaria carry a penalty in anemia? We want to urge evolution to get to where it’s going and stop the endless cruelties. But life doesn’t know where it’s going. It has no long-term plan. There’s no end in mind. There’s no mind to keep an end in mind. The process is the opposite of teleology. Life is profligate, blind, at this level unconcerned with notions of justice. It can afford to waste multitudes.

—

The evolutionary process could not have gone very far, though, if the mutation rate had been too high. In any given environment, there must be a delicate balance—simultaneously avoiding mutation rates so high that instructions for essential molecular machine tools are quickly garbled, and mutation rates so low that the organism is unable to retool when changes in the external environment require it to adapt or die.

There is a vast molecular industry that repairs or replaces damaged or mutated DNA. In a typical DNA molecule, hundreds of nucleotides are inspected every second and many nucleotide substitutions or errors corrected. The corrections are then themselves proofread, so that there is only about one error in every billion nucleotides copied. This is a standard of quality control and product reliability rarely reached in, say, publishing or automobile manufacture or microelectronics. (It is unheard of that a book this size, containing around a million letters would have no typographical errors; a 1% failure rate is common in automobile transmissions manufactured in America; advanced military weapons systems are typically down for repair some 10% of the time.) The proofreading and correction machinery devotes itself to DNA segments that are actively involved in controlling the chemistry of the cell, and mainly ignores nonfunctioning, largely untranscribed, or “nonsense” sequences.

The unrepaired mutations steadily accumulating in these normally silent regions of the DNA may lead (among other causes) to cancer and other illnesses, should the “STOP” be ignored, the sequence turned on, and the instructions carried out. Long-lived organisms such as humans devote considerable attention to repairing the silent regions; short-lived organisms such as mice do not, and often die filled with tumors.¹⁰ Longevity and DNA repair are connected.

Consider an early one-celled organism floating near the surface of the primeval sea—and thereby flooded with solar ultraviolet radiation. A small segment of its nucleotide sequence reads, let’s say,

When ultraviolet light strikes DNA, it often binds two adjacent T nucleotides together by a second route, preventing DNA from exercising its coding function and getting in the way of its ability to reproduce itself:

The molecule literally gets tied up in knots. In many organisms enzymatic repair crews are called in to correct the damage. There are three or four different kinds of crews, each specialized for repairing a different kind of damage. They snip out the offending segment and its adjacent nucleotides (CC, say) and replace it with an unimpaired sequence (CTTC). Protecting the genetic information and making sure it can reproduce itself with high fidelity is a matter of the highest priority. Otherwise, useful sequences, tried-and-true instructions, essential for the adaptation of organism to environment, may be quickly lost by random mutation. Proofreading and repair enzymes correct damage to the DNA from many causes, not just UV light. They probably evolved very early, at a time before ozone, when solar ultraviolet radiation was a major hazard to life on Earth. Early on, the rescue squads themselves must have undergone fierce competitive evolution. Today, up to a certain level of irradiation and exposure to chemical poisons, they work extremely well.

Advantageous mutations occur so rarely that sometimes—especially in a time of swift change—it may be helpful to arrange for an increased mutation rate. Mutator genes in such circumstances can themselves be selected for—that is, those varieties with active mutator genes serve up a wider menu of organisms for selection to draw upon, and serve them up faster. Mutator genes are nothing mysterious; some of them, for example, are just the genes ordinarily in charge of proofreading or repair. If they fail in their error-correcting role, the mutation rate, of course, goes up. Some mutator genes encode for the enzyme DNA polymerase, which we will meet again later; it’s in charge of duplicating DNA with high fidelity. If that gene goes bad, the mutation rate may rise quickly. Some mutator genes turn As into Gs; others, Cs into Ts, or vice versa. Some delete parts of the ACGT sequence. Others accomplish a frame shift, so the genetic code is read, three nucleotides at a time, as usual, but from a starting point offset by one nucleotide—-which can change the meaning of everything.¹¹

This is a marvel of self-reflexive talent. Even very simple microorganisms have it. When conditions are stable, the precision of reproduction is stressed; when there’s an external crisis that needs attending to, an array of new genetic varieties is generated. It might look as if the microbes are conscious of their predicament, but they haven’t the foggiest notion of what’s going on. Those with appropriate genes preferentially survive. Active mutators in placid and stable times tend to die off. They are selected against. Reluctant mutators in quickly changing times are also selected against. Natural selection elicits, evokes, draws forth a complex set of molecular responses that may superficially look like foresight, intelligence, a master Molecular Biologist tinkering with the genes; but in fact all that is happening is mutation and reproduction, interacting with a changing external environment.

——

Since favorable mutations are served up so slowly, major evolutionary change will ordinarily require vast expanses of time. There are, as it turns out, ages available. Processes that are impossible in a hundred generations may be inevitable in a hundred million. “The mind cannot grasp the full meaning of the term of a million or a hundred million years,” Darwin wrote in 1844, “and cannot consequently add up and perceive the full effects of small successive variations accumulated during almost infinitely many generations.”¹²

The time scale problem was formidable when Darwin wrote. Lord Kelvin, the greatest physicist of the late Victorian age, authoritatively announced that the Sun—and therefore life on Earth—could be no more than about a hundred million (later downgraded to thirty million) years old. The fact that he provided a quantitative argument, plus his enormous prestige, intimidated many geologists and biologists, Darwin included. Is it more probable, Kelvin asked,¹³ that straightforward physics was in error, or that Darwin was wrong? There was in fact no error in Kelvin’s physics, but his starting assumptions were mistaken. He had assumed that the Sun shines because of meteorites and other debris falling into it. There was not the faintest hint in the physics of Kelvin’s time of thermonuclear reactions; even the existence of the atomic nucleus was unknown. As late as the first decade of the twentieth century it was believed that the Earth was only 100 million years old, instead of 4.5 billion, and that the mammals had supplanted the dinosaurs only 3 million years ago, instead of 65 million.

On the basis of these misconceptions, Darwin’s critics argued—properly—that even if evolution worked in principle, there might not be enough time for it to do its stuff in practice.* On an Earth created less than ten thousand years ago, it was absurd to imagine that species flowed one into another, that the slow accumulation of mutations could explain the varied forms of life on Earth. It made sense, not merely as an expression of faith, but as legitimate science, to conclude that each species must have been separately created by the same Maker who had only a moment before created the Universe.

The breakup of rocks by the waves, the transport of rock powder by the winds, lava flowing down the sides of a volcano—if the Earth is only a few thousand years old, such processes cannot have much reworked the face of our planet. But the most casual look at the landforms of Earth reveals a profound reworking. So if you imagined from biblical chronology that the world was formed around the year 4000 B.C., it made sense to be a catastrophist—and believe that immense cataclysms, unknown in our time, have occurred in earlier history. The Noachic flood, as we’ve mentioned, was a popular example. If, though, the Earth is 4.5 billion years old, the cumulative impact of small, nearly imperceptible changes over the course of ages could wholly alter our planet’s surface.

Once the time scale for the terrestrial drama had been extended to billions of years, much that had once seemed impossible could now be readily explained as the concatenation of apparently inconsequential events—the footfalls of mites, the settling of dust, the splatter of raindrops. If, in a year, wind and water rub a tenth of a millimeter off the top of a mountain, then the highest mountain on Earth can be flattened in ten million years. Catastrophism gave way to uniformitarianism, championed by Lyell in geology and by Darwin in biology. The accumulation of vast numbers of random mutations was now inevitable, unavoidable. Great cataclysms were discredited and special creation became, both in geology and biology, a redundant and unnecessary hypothesis.

Many advocates of uniformitarianism denied that quick and violent biological change had ever occurred. T. H. Huxley, for example, wrote, “There has been no grand catastrophe—no destroyer has swept away the forms of life of one period, and replaced them by a totally new creation: but one species has vanished and another has taken its place; creatures of one type of structure have diminished, those of another have increased, as time has passed on.”¹⁴ In the light of modern evidence, he was right in general, right for most of the history of the Earth. But he went too far; clearly it is possible to acknowledge the importance of slow, cumulative, background change without denying the possibility of occasional global cataclysms.

In recent years it has become increasingly evident that catastrophes have swept over the Earth, generating vast alterations both in land-forms and in life. Major worldwide discontinuities in the record in the rocks are readily explained by such catastrophes; and abrupt transitions in the forms of life on Earth, occurring in the same epoch, are naturally understood as mass extinctions, times of great dyings. (Of these, the late Permian is the most extreme example, and the late Cretaceous—when the dinosaurs were all snuffed out—the best-known). Previous ecologies are then supplanted wholesale by new teams of organisms. The fossil record shows that long periods of very slow evolutionary change are often interrupted by rarer, episodic intervals of quick change, the “punctuated equilibrium” of Niles Eldredge and Stephen J. Gould.¹⁵ We live on a planet in which both catastrophes and uniform change have played their roles. In the purported distinction between all-at-once and slow-and-steady, as in much else, the truth embraces seemingly antithetical extremes.

The case for special creation has not been strengthened by this new balance. Catastrophism is an awkward business for biblical literalists: It suggests imperfections in either the design or the execution of the Divine Plan. Mass extinctions permit the survivors to evolve quickly, occupying ecological niches formerly closed to them by the competition. The painstaking selection of mutations continues, catastrophes or no catastrophes. But the wiping out of whole species, genera, families and orders of life, the randomness of mutation, the infelicities in the molecular machinery of life, and the slow evolutionary fiddling displayed in the fossil record—of trilobites, say, or crocodiles—all reveal a tentativeness, a hesitancy, an indecision that hardly seems consistent with the modus operandi of an omnipotent, omniscient, “hands-on” Creator.

——

Why are many cave fish, moles, and other animals that live in perpetual darkness blind, or nearly so? At first the question seems ill-conceived, since no adaptive reward would attend the evolution of eyes in the dark. But some of these animals do have eyes, only they’re beneath the skin and don’t work. Others have no eyes at all, although anatomically it’s clear that their ancestors did. The answer seems to be that they all evolved from sighted creatures that entered a new and promising habitat—a cave, say, lacking competitors and predators. There, over many generations, no penalty is paid for the loss of eyesight. So what if you’re blind, as long as you live in pitch darkness? Mutations for blindness, which must be occurring all the time (there being many possible malfunctions in the genetic instructions for vision—in eye, retina, optic nerve, and brain), are not selected against. A one-eyed man has no advantage in the kingdom of darkness.

Similarly, whales have small, internal, and wholly useless pelvises and leg bones, and snakes have four vestigial internal feet. (In the mambas of Southern Africa a single claw from each rudimentary limb breaks through the scaly skin to plain view.) If you swim or slither and never walk anymore, mutations for the withering away of feet do you no harm. They are not selected against. They might even be selected for (feet can be in the way when you’re pouring down a narrow hole). Or if you’re a bird that finds itself on an island devoid of predators, no penalty is levied for the steady atrophy, generation after generation, of wings (until European sailors arrive and club you all to death).

Mutations are occurring all the time for the loss of all sorts of functions. If there’s no disadvantage attached to these mutations, they can establish themselves in the population. Some will even be helpful—shedding formerly useful machinery, say, that is no longer worth the effort of maintaining. There must also be enormous numbers of mutations for biochemical incompetence and other major dysfunctions which result in beings that never survive their embryonic stages. They die before they’re born. They’re rejected by natural selection before the biologist can examine them. Relentless, draconian winnowing is occurring all around us. Selection is a school of hard knocks.

Evolution is just trial and error—but with the successes encouraged and proliferated, the failures ruthlessly extirpated, and prodigious vistas of time available for the process to work itself out. If you reproduce, mutate, and reproduce your mutations, you must evolve. You have no choice in the matter. You get to keep playing the game of life only if you keep winning; that is, if you keep leaving descendants (or close relatives). One break in the train of generations, and you and your particular, idiosyncratic DNA sequences are condemned without hope of reprieve.

——

The English-language edition of this book is printed in letters that trace back to western Asia, and in a language primarily derived from Central Europe. But this is solely a matter of historical accident. The alphabet might not have been invented in the ancient Near East if there had not been a thriving mercantile culture there, if there had been no need for systematic records of commercial transactions. Spanish is spoken in Argentina, Portuguese in Angola, French in Quebec, English in Australia, Chinese in Singapore, a form of Urdu in Fiji, a form of Dutch in South Africa, and Russian in the Kuriles only because of a contingent sequence of historical events, some quite unlikely. Had they run a different course, other languages might be spoken in these places today. The Spanish, French, and Portuguese languages in turn depend on the fact that the Romans had imperial ambitions; English would be very different if Saxons and Normans had not been bent on overseas conquest; and so on. Language depends on history.

That a planet the size of the Earth is a sphere and not a cube, that a star the size of the Sun mainly emits visible light, that water is a solid and a liquid and a gas on any world at the surface temperature and pressure of the Earth—these facts are all readily understood from a few simple principles of physics. They are not contingent truths. They do not depend on a particular sequence of events that could just as well have gone some other way. Physical reality has a permanence and stability, an obsessive regularity to it, while historical reality tends to be fickle and fluid, less predictable, less rigidly determined by those laws of Nature we know. Something like accident or chance seems to play a major role in issuing marching orders to the flow of historical events.

Biology is much more like language and history than it is like physics and chemistry. Why we have five fingers on each hand, why the cross-section of the tail of a human sperm cell looks so much like that of a one-celled Euglena, why our brains are layered like an onion, involve strong components of historical accident. Now you might say that where the subject is simple, as in physics, we can figure out the underlying laws and apply them everywhere in the Universe; but where the subject is difficult, as in language, history, and biology, governing laws of Nature may well exist, but our intelligence may be too feeble to recognize their presence—especially if what is being studied is complex and chaotic, exquisitely sensitive to remote and inaccessible initial conditions. And so we invent formulations about “contingent reality” to disguise our ignorance. There may well be some truth to this point of view, but it is nothing like the whole truth, because history and biology remember in a way that physics does not. Humans share a culture, recall and act on what they’ve been taught. Life reproduces the adaptations of previous generations, and retains functioning DNA sequences that reach billions of years back into the past. We understand enough about biology and history to recognize a powerful stochastic component, the accidents preserved by high-fidelity reproduction.

——

DNA polymerase is an enzyme. Its job is to assist a DNA strand in copying itself. It itself is a protein, configured out of amino acids and manufactured on the instructions of the DNA. So here’s DNA controlling its own replication. DNA polymerase is now on sale at your local biochemical supply house. There’s a laboratory technique, polymerase chain reaction, which unzips a DNA molecule by changing its temperature; the polymerase then helps each strand to reproduce. Each of the copies is in turn unzipped and replicates itself.¹⁶ At every step in this repetitive process, the number of DNA molecules doubles. In forty steps there are a trillion copies of the original molecule. Of course, any mutation happening along the way is also reproduced. So polymerase chain reactions can be used to simulate evolution in the test tube.* Something similar can be done for other nucleic acids:

In the test tube before you is another kind of nucleic acid—this one single-stranded. It’s called RNA (ribonucleic acid). It’s not a double helix and does not have to be unzipped to make a copy of itself. The strand of nucleotides may loop around to join itself, tail in mouth, a molecular circle. Or it may have hairpin or other shapes. In this experiment it’s sitting mixed with its fellow RNA molecules in water. There are other molecules added to help it along, including nucleotide building blocks for making more RNA. The RNA is coddled, jollied, handled with kid gloves. It’s extremely finicky and will do its magic only under very specific conditions. But magic it does. In the test tube not only does it make identical copies of itself, but it also moonlights as a marriage broker for other molecules. Indeed, it performs even more intimate services, providing a kind of platform or marital bed for oddly shaped molecules to join together, to fit into one another. It’s a jig for molecular engineering. The process is called catalysis.

This RNA molecule is a self-replicating catalyst. To control the chemistry of the cell, DNA has to oversee the construction of factotums—a different class of molecules, proteins, which are the catalytic machine tools we’ve been discussing above. DNA makes proteins because it can’t catalyze on its own. Certain kinds of RNA, though, can themselves serve as catalytic machine tools.¹⁷ Making a catalyst or being a catalyst gives you the biggest return for the smallest investment: Catalysts can control the production of millions of other molecules. If you make a catalyst, or if you are a catalyst—the right kind of catalyst—you have a long lever arm on your destiny.

Now in these laboratory experiments, which are being carried out in our time, imagine many generations of RNA molecules more or less identically replicating in the test tube. Mutations inevitably occur, and much more often than in DNA. Most of the mutated RNA sequences will leave no, or fewer, copies, again because random changes in the instructions are rarely helpful. But occasionally a molecule comes into existence that aids its own replication. Such a newly mutated RNA might replicate faster than its fellows or with greater fidelity. If we were uncaring about the fates of individual RNA molecules—and while they may arouse feelings of wonder, they seldom elicit sympathy—and wished only for the advancement of the RNA clan, this is just the kind of experiment we would perform. Most lines would perish. A few would be better adapted and leave many copies. These molecules will slowly evolve. A self-replicating, catalytic RNA molecule may have been the first living thing in the ancient oceans about 4 billion years ago, its close relative DNA being a later evolutionary refinement.

In an experiment with synthetic organic molecules that are not nucleic acids, two closely related species of molecules are found to make copies of themselves out of molecular building blocks provided by the experimenter. These two kinds of molecules both cooperate and compete: They may aid each other’s replication, but they are also after the same limited pool of building blocks. When ordinary visible light is made to shine on this submicroscopic drama, one of the molecules is observed to mutate: It changes into a somewhat different molecule that breeds true—it makes identical copies of itself, and not its pre-mutation ancestor. This new variety, it turns out, is much more adept at replicating itself than the other two hereditary lines. The mutant line rapidly out-competes the others, whose numbers precipitously fall.¹⁸ We have here, in the test tube, replication, mutation, replication of mutations, adaptation, and—we do not think it is too much to say—evolution. These are not the molecules that make us up. They are probably not the molecules involved in the origin of life. There may well be many other molecules which reproduce and mutate better. But what prevents us from calling this molecular system alive?

Nature has been performing similar experiments, and building on its successes, for 4 billion years.

Once even crude replication becomes possible, an engine of enormous powers has been let loose into the world. For example, consider that primitive organic-rich ocean of the Earth. Suppose we were to drop a single organism (or a single self-replicating molecule) into it, considerably smaller than a contemporary bacterium. This tiny being divides in two, as do its offspring. In the absence of any predators and with inexhaustible food supplies, their numbers would increase exponentially. The being and its descendants would take only about one hundred generations to eat up all the organic molecules on Earth. A contemporary bacterium under ideal conditions can reproduce once every fifteen minutes. Suppose that on the early Earth the first organism could reproduce only once a year. Then in only a century or so, all the free organic matter in the whole ocean would have been used up.

Of course, long before that, natural selection would be brought to bear. The genre of selection might be competition with others of your kind—for example, for foodstuffs in an ocean with dwindling stocks of preformed molecular building blocks. Or it might be predation—if you don’t look out, some other being will mug you, strip you down, pull you to pieces, and use your molecular parts for its own ghastly purpose

Major evolutionary advance might take considerably more than one hundred generations. But the devastating power of exponential replication becomes clear: When the numbers are small, organisms may only infrequently come into competition; but after exponential replication, enormous populations are produced, stringent competition occurs, and a ruthless selection comes into play. A high population density generates circumstances and elicits responses different from the more friendly and cheerful lifestyles that pertain when the world is sparsely populated.

The external environment is continuously changing—in part because of the enormous population growth when conditions are favorable, in part because of the evolution of other organisms, and in part because of the ticking geological and astronomical clockwork. So there’s never such a thing as a permanent or final or optimum adaptation of a lifeform to “the” environment. Except in the most protected and static surrounds, there must be an endless chain of adaptations. However it feels on the inside, it might very well be described from the outside as a struggle for existence and a competition between adults to ensure the success of their offspring.

You can see that the process tends to be adventitious, opportunistic—not foresighted, not with any future end in view. The evolving molecules do not plan ahead. They simply produce a steady stream of varieties, and sometimes one of the varieties turns out to be a slightly improved model. No one—not the organism, not the environment, not the planet, not “Nature”—is mulling the matter over.

This evolutionary shortsightedness can lead to difficulties. It might, for example, cast aside an adaptation that is perfectly suited for the next environmental crisis a thousand years from now (about which, of course, no one has a glimmering). But you have to get from here to there. One crisis at a time is life’s motto.