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.