by Nathan Ensmenger
The title of the book notwithstanding, the protagonist of George Dyson's Turing's Cathedral: The Origins of the Digital Universe is not the English mathematician Alan Turing, but rather the Hungarian polymath John von Neumann. Widely celebrated by historians of science for contributions to physics, mathematics, economics, and computer science, von Neumann is less well-known among the general public. Here, Dyson provides a concise and compelling biography that manages to humanize and, more impressively, make comprehensible the deep personal and intellectual connections among the wide-ranging interests and activities of this fascinating individual.
A Jewish émigré who resigned his position from a German university in the 1930s, von Neumann relocated to the newly founded Princeton Institute for Advanced Study (IAS) and threw himself into the war effort on behalf of the US, crisscrossing the country repeatedly as he circulated between Princeton, Los Alamos, Philadelphia (where he was involved with the Eniac project), Poughkeepsie, New York (where he served as a consultant to IBM), and seemingly every other major site of wartime techno-scientific development. Dyson makes excellent use of von Neumann's uncanny instinct to be "always willing to go where the action was" to navigate the reader through the complex web of academia, industry, and the military that characterized the early era of electronic computing.
The real subject of this book, however, is not a person at all, but rather a machine. Immediately following the end of the war, von Neumann leveraged his status and connections into a commitment by the IAS to build an electronic computer. The immediate motivation for doing so was von Neumann's ongoing collaboration with Los Alamos. During the war, his need to solve large numbers of hydrodynamic equations for blast calculations led him to the pioneering Eniac project and, in the early Cold War race, to build a hydrogen bomb, this military imperative became even more pressing.
But von Neumann also had a larger, more ambitious agenda in mind. The real significance of building computers, of "accelerating approximating and computing mathematics by factors like 10,000 or more," argued von Neumann, "lies not only in that one might do in 10,000 times less problems which one is now doing ... but rather in that one will be able to handle problems which are considered completely unassailable at present" (p. 85). It is this new mode of doing science through simulation that Dyson is most interested in, and much of the last third of the book attempts to make a connection between von Neumann's work on the IAS computer and later developments in computational biology. The IAS computer, Dyson argues, represents the true origins of modern computing, and von Neumann was first able to realize the vision of Alan Turing's Universal Machine.
Dyson's rich description of the history of the IAS computer is this book's most important contribution. The son of the physicist Freeman Dyson, George Dyson grew up in Princeton, and he makes great use of his insider knowledge and connections to paint a vivid and intimate portrait of Princeton in the late 1940s and early 1950s.... A curious mix of quaint, small-town isolation and cutting-edge cosmopolitan techno-science, Princeton in this period provided fertile ground for von Neumann to realize his vision.
In the last third of Turing's Cathedral, Dyson develops a thesis familiar to those who have encountered his earlier book Darwin Among the Machines.1 Building on the work of the biologist Nils Barricelli, Dyson posits a direct, literal equivalence between biological and computational processes. The machine language of the gene is the same as the machine language of the computer, Dyson argues, and Barricelli was the first to perceive this fundamental relationship. At the center of his argument is a series of computational experiments that Barricelli developed for and ran on the IAS computer between 1953 and 1956. According to Dyson, Barricelli's computer code represents the "Dead Sea Scrolls" that outline the "Genesis" of what we now know to be the digital universe, one in which iPods and the Internet are now effectively indistinguishable from biological organisms.
It is in developing his grand argument that Dyson is least convincing. He draws heavily on the work of fringe figures such as Barricelli without ever reflecting on why they were never taken seriously by more mainstream scientists. He stretches metaphors and analogies to the breaking point. This is particularly evident when he talks about the common language of "code" as mobilized by geneticists (as in DNA code) and computer scientists (as in machine code). The historian Lily Kay, among others, has revealed the way in which such shared metaphors traveled, were co-opted, and were transformed fundamentally by computationally minded biologists in the 1950s and 1960s.2 How seriously such metaphors and analogies were taken varied greatly, and their use continues to be controversial to this day. Dyson is either unaware of such arguments or has decided to ignore them.
The first half of this book is a beautifully written and compelling account of an important moment in the history of computing, but its overreaching and sensationalist conclusions diminish its value and effectiveness. As a historian of computing, there is much in this book that I found useful and interesting, but I would have serious reservations recommending it to a general audience. Dyson references almost none of the existing historical literature, and cherry-picks selectively from the history of computing to support his central thesis, imposing on the past a presentist and universalizing perspective. One would never know, from reading this book, that the IAS computer, although a pioneering and influential development, was but one of many stored-program computers built in this period, or that the computationalist turn in biology was resisted, and continues to be resisted, by many in the life sciences. This is a teleologically driven history in which the outcome is inevitable. We are living in a digital universe, according to Dyson, and it just happened to take us until recently to discover it. To paraphrase one of his numerous Biblical metaphors, God is a computer, and Turing and von Neumann were his prophets.
- G. Dyson, Darwin Among the Machines, Perseus Books, 1997.
- L.E. Kay, Who Wrote the Book of Life? A History of the Genetic Code, Stanford Univ. Press, 2000.
Nathan Ensmenger is an associate professor at Indiana University's School of Informatics and Computing. Contact him at firstname.lastname@example.org. This review excerpts a longer version that appeared in IEEE Annals in Computing (Jan.&38211;Mar. 2012, p. 6–7).