American Scientist

Who Wrote the Book of Life?: A History of the Genetic Code. Lily E. Kay. x + 441 pp. Stanford University Press, 2000.

Giving the subtitle A History of the Genetic Code to a book published in the year 2000 suggests that its topic is the unraveling of the human genome. Instead, in Who Wrote the Book of Life? Lily E. Kay chronicles the exciting period from 1953 to 1967, during which biologists formulated, addressed and answered the question of how the sequences of bases (nucleotides) that constitute DNA molecules specify the sequences of amino acids that form proteins.

Do we need a new book on the subject? The definitive, standard history is Horace Freeland Judson's Eighth Day of Creation, which was first published in 1979 and revised and expanded in 1996. It remains the book I would recommend first to anyone wishing to learn who did what when in molecular biology from its beginnings through the "deciphering" of the genetic "code." Kay's book describes the same events as Judson's, but from a different and somewhat unconventional perspective: In the preface she says that her vantage point is "the dawn of the information age and its impact on representations of nature and society." Thus Who Wrote the Book of Life? might be considered revisionist history.

I would recommend this book only to those who wish to see how well it accomplishes Kay's stated objective of examining critically the supposed influence of information theory—with its use of terms such as code, message and noise—on empirical research in genetics. My own assessment is that the book, which reads like a thoroughly researched legal brief (more than 100 pages are devoted to notes, references and a very detailed index), makes the best possible case for the highly dubious proposition that the ideas of information theory influenced the substance, rather than merely the rhetoric, of research in molecular biology in the 1950s and 1960s.

Kay attempts to interpret developments in molecular biology in their historical context. The period she covers has been described as the post–World War II era, the Cold War period, the Information Age and the early Computer Age, and to each of these characterizations she ascribes significant impact on the course of research in molecular biology. I think she goes too far in doing so. During the Cold War, the United States was engaged in arms, missile and space races with the Union of Soviet Socialist Republics—but certainly not a molecular biology race. (Trofim Lysenko prevented any Soviet progress in molecular biology during the Stalin and Khrushchev years.) That the U.S. Office of Naval Research (ONR) funded some U.S. scientists scarcely meant that the ONR saw the early, basic work in molecular biology as directly relevant to the Cold War; and researchers grab funding wherever they can. Claude Shannon and Norbert Wiener, who never worked in molecular biology, are each referenced far more often than many of the major contributors to the deciphering of the genetic code, even though none of the theorems or formulas of Shannon or Wiener ever played any role in molecular biology. Computers during this period were far too primitive to have the kind of major role they played decades later in elucidating the human genome. What really drove progress in molecular biology was the steady improvement in the technology of biochemistry, which is insufficiently emphasized in this book.

What goes on in the outside world certainly affects the lives of scientists but not necessarily the content of their science. Isaac Newton's life was significantly influenced by the plague and the Puritan revolution, but only an extreme social constructivist would maintain that those events had an impact on the scientific content of his laws of motion and gravitation or his formulation of the calculus. Similarly, the correspondence between nucleic acids (DNA and RNA) and polypeptides (proteins, made up of amino acids) existed undiscovered long before the post–World War II era, the Information Age, the Computer Age and the Cold War period, and those eras had minimal effect on the process of discovering that correspondence.

According to the dust jacket, "the author persuasively argues that technically speaking, the genetic code is not a code, DNA is not a language, and the genome is not an information system." Kay seems obsessed with what words ought to mean and takes a far more restrictive view than most of us would. Consider the word code, which in fact has a long list of accepted meanings, as suggested by its use in phrases such as criminal code, computer code, Morse code, error-correcting code, building code and cryptographic code. No mid–20th-century ideas were required in order to refer to decoding the genetic code. In the 1820s, Jean-François Champollion undertook the decoding, indeed the deciphering, of ancient Egyptian hieroglyphics. The scribes of the Pharaohs had no intention of engaging in cryptography when they carved their inscriptions, yet to most people it does not seem unreasonable to apply the words decoding or deciphering to figuring out how to read forgotten scripts.

The book belabors the fact that in Shannon's information theory, information is devoid of meaning and semantic content; it is merely a logarithmic measure of the number of possible messages that could be constructed or sent. Although Shannon worked for Bell Laboratories and viewed information from the perspective of a communications utility company, Kay tries to link his view of information to World War II military notions.

It is, perhaps, unfair of a reviewer to use firsthand knowledge of events to nitpick at minor errors of fact. But it seems worth pointing out that in the following passage from chapter 4 (in which I am mentioned), the many minor errors do lend support to Kay's dubious notion that there was a significant military role in the work on the genetic code:

Comma-free codes attracted considerable attention. . . . It seemed natural . . . to draw on the expertise of mathematicians and communication engineers to analyze and elaborate on the general properties of comma-free codes. Such expertise was readily available at Caltech's Jet Propulsion Laboratory (JPL), which sat only a stone's throw from [Max] Delbruck's group and which since the end of World War II had been sponsored by the United States Army.
. . . JPL researchers developed some of the key technologies of the missile age. . . . And their role in the arms race intensifed after the apssage of the National Defense Security Act (1958) and the militarization of NASA, both reactions to the launching of Sputnik I on October 4, 1957.
At Caltech, Delbruck cultivated close working relations with physicists and mathematicians, in and outside JPL. . . .
Three JPL scientists responded to the challenge: Solomon W. Golomb, a young mathematician and communication engineer; mathematician Basil Gordon, specialist in combinatorial analysis; and mathematician and electrical engineer Lloyd R. Welch, who two years later joined the Institute for Defense Analysis [sic]. In the summer of 1957 they tackled the challenge of providing a mathematical generalization of the coding problem: What was the maximum size of a comma-free dictionary . . .?

Although it is true that Welch and I were working at JPL at the time, our recruitment to work with Delbrück had nothing to do with that. In fact, Delbrück merely contacted the California Institute of Technology math department, where Gordon and Welch were advanced doctoral students (Welch was working part-time at JPL while completing his thesis). I don't believe that Delbrück ever visited JPL at any time in his life—and neither did Gordon, who wasn't a "JPL scientist" in any possible sense. I had known Gordon since high school, and it was he who recruited me to think about the "comma-free code" problem. Although I became a friend and disciple of Delbrück, I am not aware of anyone else at JPL with whom he maintained contact. Our initial work on comma-free codes (which was not supported or sponsored by JPL) began in the fall of 1956, shortly after I arrived at JPL—not the summer of 1957, by which time it had already been completed.

I consider Kay's claim that the role of JPL researchers in the arms race intensified in 1958 to be a gross distortion of history. The National Aeronautics and Space Administration (NASA) was created by the Eisenhower Administration to run a nonmilitary space program for the United States in July 1958, at which time funding and administrative responsibility for JPL were transferred from the U.S. Army (the military) to the civilian space agency NASA. At that point, my own work at JPL transitioned from developing secure electronic guidance systems for Army missiles to designing the communications and tracking systems for peaceful communication satellites and lunar and planetary probes.

What Gordon, Welch and I actually had in common in 1956 was that we were all Ph.D. candidates in mathematics (my doctoral defense at Harvard was in the spring of 1957) and were interested in combinatorial problems. Contrary to Kay's suggestions, what we did required no knowledge of information theory and could certainly have been done by Leonhard Euler in the 18th century, had anyone asked him to help.

Who Wrote the Book of Life? is, in general, carefully researched and technically accurate. It is a veritable treasure trove of quotations, citations and interesting information relating to its historical period. The author's perspective gives a certain coherence to the story that some readers may appreciate. But as an information theorist who observed much of this drama close at hand (primarily from 1956 to 1964), I formed my own conceptual model, which differs significantly from that put forth by Kay.