carries the instructions for assembling a new rabbit. But the ability to create order through metabolism also depends on information -
the instructions for building and maintaining the equipment that creates the order. An adult rabbit, with its ability to both reproduce and metabolise, is prefigured and presupposed in its living filaments in the same way that a cake is prefigured and presupposed in its recipe. This is an idea that goes right back to Aristotle, who said that the 'concept' of a chicken is implicit in an egg, or that an acorn was literally 'informed' by the plan of an oak tree. When Aristotle's dim perception of information theory, buried under generations of chemistry and physics, re- emerged amid the discoveries of modern
The filament of D N A is information, a message written in a code of chemicals, one chemical for each letter. It is almost too good to be true, but the code turns out to be written in a way that we can understand. Just like written English, the genetic code is a linear language, written in a straight line. Just like written English, it is digital, in that every letter bears the same importance. Moreover, the language of D N A is considerably simpler than English, since it has an alphabet of only four letters, conventionally known as A, C, G and T.
Now that we know that genes are coded recipes, it is hard to recall how few people even guessed such a possibility. For the first half of the twentieth century, one question reverberated unanswered through biology: what is a gene? It seemed almost impossibly mysterious. Go back not to 1953, the year of the discovery of D N A ' s symmetrical structure, but ten years further, to 1943. Those who will do most to crack the mystery, a whole decade later, are working on other things in 1943. Francis Crick is working on the design of naval mines near Portsmouth. At the same time James Watson is just enrolling as an undergraduate at the precocious age of fifteen at the University of Chicago; he is determined to devote his life to ornithology. Maurice Wilkins is helping to design the atom bomb in the United States. Rosalind Franklin is studying the structure of coal for the British government.
1 4 G E N O M E
In Auschwitz in 1943, Josef Mengele is torturing twins to death in a grotesque parody of scientific inquiry. Mengele is trying to understand heredity, but his eugenics proves not to be the path to enlightenment. Mengele's results will be useless to future scientists.
In Dublin in 1943, a refugee from Mengele and his ilk, the great physicist Erwin Schrodinger is embarking on a series of lectures at Trinity College entitled What is life?' He is trying to define a problem. He knows that chromosomes contain the secret of life, but he cannot understand how: 'It is these chromosomes . . . that contain in some kind of code-script the entire pattern of the individual's future development and of its functioning in the mature state.' The gene, he says, is too small to be anything other than a large molecule, an insight that will inspire a generation of scientists, including Crick, Watson, Wilkins and Franklin, to tackle what suddenly seems like a tractable problem. Having thus come tantalisingly close to the answer, though, Schrodinger veers off track. He thinks that the secret of this molecule's ability to carry heredity lies in his beloved quantum theory, and is pursuing that obsession down what will prove to be a blind alley. The secret of life has nothing to do with quantum states. The answer will not come from physics.3
In New York in 1943, a sixty-six-year-old Canadian scientist, Oswald Avery, is putting the finishing touches to an experiment that will decisively identify D N A as the chemical manifestation of heredity. He has proved in a series of ingenious experiments that a pneumonia bacterium can be transformed from a harmless to a virulent strain merely by absorbing a simple chemical solution. By 1943, Avery has concluded that the transforming substance, once purified, is D N A . But he will couch his conclusions in such cautious language for publication that few will take notice until much later.
In a letter to his brother Roy written in May 1943, Avery is only slightly less cautious:4
If we are right, and of course that's not yet proven, then
characteristics of cells — and that by means of a known chemical substance it is possible to induce predictable and hereditary changes in cells. That is something that has long been the dream of geneticists.
Avery is almost there, but he is still thinking along chemical lines.
'All life is chemistry', said Jan Baptista van Helmont in 1648, guessing.
At least some life is chemistry, said Friedrich Wohler in 1828 after synthesising urea from ammonium chloride and silver cyanide, thus breaking the hitherto sacrosanct divide between the chemical and biological worlds: urea was something that only living things had produced before. That life is chemistry is true but boring, like saying that football is physics. Life, to a rough approximation, consists of the chemistry of three atoms, hydrogen, carbon and oxygen, which between them make up ninety-eight per cent of all atoms in living beings. But it is the emergent properties of life — such as heritability
- not the constituent parts that are interesting. Avery cannot conceive what it is about D N A that enables it to hold the secret of heritable properties. The answer will not come from chemistry.
In Bletchley, in Britain, in 1943, in total secrecy, a brilliant mathematician, Alan Turing, is seeing his most incisive insight turned into physical reality. Turing has argued that numbers can compute numbers. To crack the Lorentz encoding machines of the German forces, a computer called Colossus has been built based on Turing's principles: it is a universal machine with a modifiable stored program.
Nobody realises it at the time, least of all Turing, but he is probably closer to the mystery of life than anybody else. Heredity is a modifiable stored program; metabolism is a universal machine. The recipe that links them is a code, an abstract message that can be embodied in a chemical, physical or even immaterial form. Its secret is that it can cause itself to be replicated. Anything that can use the resources of the world to get copies of itself made is alive; the most likely form for such a thing to take is a digital message - a number, a script or a word.5
In New Jersey in 1943, a quiet, reclusive scholar named Claude Shannon is ruminating about an idea he had first had at Princeton l 6 G E N O M E
a few years earlier. Shannon's idea is that information and entropy are opposite faces of the same coin and that both have an intimate link with energy. The less entropy a system has, the more information it contains. The reason a steam engine can harness the energy from burning coal and turn it into rotary motion is because the engine has high information content — information injected into it by its designer. So does a human body. Aristotie's information theory meets Newton's physics in Shannon's brain. Like Turing, Shannon has no thoughts about biology. But his insight is of more relevance to the question of what is life than a mountain of chemistry and physics. Life, too, is digital information written in D N A . 6
In the beginning was the word. The word was not D N A . That came afterwards, when life was already established, and when it had divided the labour between two separate activities: chemical work and information storage, metabolism and replication. But D N A contains a record of the word, faithfully transmitted through all subsequent aeons to the astonishing present.
