The Information - James Gleick [143]
If you want to know what a mouse is, ask instead how you could build a mouse. How does the mouse build itself? The mouse’s genes switch one another on and off and perform computation, in steps. “I feel that this new molecular biology has to go in this direction—to explore the high-level logical computers, the programs, the algorithms of development.…
“One would like to be able to fuse the two—to be able to move between the molecular hardware and the logical software of how it’s all organized, without feeling they are different sciences.”
Even now—or especially now—the gene was not what it seemed. Having begun as a botanist’s hunch and an algebraic convenience, it had been tracked down to the chromosome and revealed as molecular coiled strands. It was decoded, enumerated, and catalogued. And then, in the heyday of molecular biology, the idea of the gene broke free of its moorings once again.
The more was known, the harder it was to define. Is a gene nothing more or less than DNA? Is it made of DNA, or is it something carried in DNA? Is it properly pinned down as a material thing at all?
Not everyone agreed there was a problem. Gunther Stent declared in 1977 that one of the field’s great triumphs was the “unambiguous identification” of the Mendelian gene as a particular length of DNA. “It is in this sense that all working geneticists now employ the term ‘gene,’ ”♦ he wrote. To put it technically but succinctly: “The gene is, in fact, a linear array of DNA nucleotides which determines a linear array of protein amino acids.” It was Seymour Benzer, said Stent, who established that definitively.
Yet Benzer himself had not been quite so sanguine. He argued as early as 1957 that the classical gene was dead. It was a concept trying to serve three purposes at once—as a unit of recombination, of mutation, and of function—and already he had strong reason to suspect that these were incompatible. A strand of DNA carries many base pairs, like beads on a string or letters in a sentence; as a physical object it could not be called an elementary unit. Benzer offered a batch of new particle names: “recon,” for the smallest unit that can be interchanged by recombination; “muton,” for the smallest unit of mutational change (a single base pair); and “cistron” for the unit of function—which in turn, he admitted, was difficult to define. “It depends upon what level of function is meant,” he wrote—perhaps just the specification of an amino acid, or perhaps a whole ensemble of steps “leading to one particular physiological end-effect.”♦ Gene was not going away, but that was a lot of weight for one little word to bear.
Part of what was happening was a collision between molecular biology and evolutionary biology, as studied in fields from botany to paleontology. It was as fruitful a collision as any in the history of science—before long, neither side could move forward without the other—but on the way some sparks flared. Quite of few of them were set off by a young zoologist at Oxford, Richard Dawkins. It seemed to Dawkins that many of his colleagues were looking at life the wrong way round.
As molecular biology perfected its knowledge of the details of DNA and grew more skillful in