The Cambridge Quintet: A Work of Scientific Speculation. John Casti. 181 pp. Helix Books/Addison Wesley, 1998. $23.
It is a stormy night in Cambridge, England, in 1947, and C. P. Snow has assembled a brilliant guest list of intellectuals and scientists for a lavish dinner party at his old college, Christ's. The purpose of the occasion is to get his guests talking about artificial intelligence and, specifically, about the likelihood that in the foreseeable future anyone will make a machine that can think for itself. Snow's bosses at the defense ministry want to know if something of such evident strategic military importance is a possibility. The dinner party is, of course, pure fiction. Snow, Erwin Schrödinger, J. B. S. Haldane, Ludwig Wittgenstein and Alan Turing never gathered to discuss this or any other topic. But when John Casti chooses this form for his Cambridge Quintet: A Work of Scientific Speculation, we know in what tradition we are supposed to be.
The intellectual dialogue has a long and distinguished history. In 5th century b.c. Athens, Plato orchestrated a philosophical discussion into dinner-table talk in his Symposium; Cicero emulated him in a series of Latin philosophical dialogues. Galileo's final word on planetary motion, The Dialogue of Two World Systems, which led to his arrest by the Inquisition, was cast in the form of a debate between three individuals taking distinct intellectual positions. The format allows the author to present a range of competing opinions on a topic, with each protagonist speaking lucidly and passionately from his distinctive point of view. The author astutely orchestrates the engagement between the speakers, to press the readers toward one line of thought over another. But the format itself ensures that the conclusion reached is tempered by the need on all the speakers' parts to respond to those whose arguments do not in the end win the day.
While Snow fusses around in the background, seeing to the cigars and brandy, there is no doubt who is the dominating figure in Casti's imaginary dialogue. Alan Turing, rising above his chronic shyness and diffidence, lucidly explains his Turing machine to the assembled guests and defends it with ardor against all comers. Wittgenstein is the dinner party's villain—the equivalent here of Galileo's Aristotelian puppet of the Vatican, Simplicio, whose every interjection arouses the reader's annoyance. He shouts, he loses his temper, he refuses even to countenance as meaningful the idea that without socially learned language a machine could think. Having produced his "hieroglyphic room" as a philosophical counterargument to Turing's contention that the sequences of operations of his machine might as readily produce thinking as the synapses in the human brain, he retreats into sulking punctuated by outbursts of incredulous indignation.
Haldane is the genial facilitator, whose occasional remarks on genetics and on the primeval soup from which he believes we all came are less important than his readiness, when prompted by Snow, to recapitulate areas of medical and linguistic theory in which he protests he is "not an expert."
The Cambridge Quintet engages readers and holds our attention. It allows us to eavesdrop on clear explanations of the issues and to follow the interlocutors through carefully sign-posted stages in their disagreements. Casti also peppers the text with witticisms for the initiate: Schrödinger, at one point, "puts the cat among the pigeons"; it is Haldane, later to become a passionate vegetarian, who exclaims, "Absolutely first-rate roast beef, Snow."
Schrödinger contributes the points of view familiar from his 1944 series of lectures on molecular biology, "What is Life?" which in published form has been a cult work ever since Francis Crick told us that Schrödinger's thoughts on the possibilities for mathematizing the workings of the human body had inspired his search for the double helix of DNA.
As for verisimilitude, I am prepared to suspend my predictably British disbelief over one or two of Casti's fictional details. Would the Christ's kitchens really have served a separate salad course rather than one of those horrible overcooked high-table savories? Would a group with such distinctive and varied sexual preferences really have contented themselves for an entire evening with high-minded intellectual inquiry? (In 20 years at Cambridge, I never attended such a gathering that did not degenerate into personalized gossip and innuendo.)
None of this, however, will matter to most readers, who will be gripped and intrigued. Those readers will also, thanks to the dinner-party device, be able to follow an argument that is well above many heads, even in the 1990s.
Does the fictionalized form of The Cambridge Quintet in the end communicate difficult ideas to the general reader in understandable form? This is, I think, a serious question. To satisfy my own curiosity I went back to the equivalent book I read in the 1960s, John von Neumann's The Computer and the Brain, and asked myself how its pellucid, carefully argued prose compared today with Casti's more accessible packaging of the same kinds of ideas.
Von Neumann's account of the Turing machine is succinct, with no diagrams, no witty repartee. But it does state clearly a part of the argument fundamental for the mind-machine discussion, which seems to get lost at Snow's dinner table: "Turing showed that it is possible to develop code instruction systems for a computing machine which cause it to behave as if it were another, specified, computing machine." The point of the Turing machine is not just that it is an imaginary computer; it is that, logically, its short code instructions allow it to imitate the integrated processes of the human mind—in Turing's terms, to think.
The user-friendly format of Casti's book will, I hope, encourage large numbers of ordinary readers to try to understand the ideas he presents. But the real test is whether ordinary readers have, by the time they finish it, understood the force of Turing's logical argument and of Wittgenstein's philosophical objections to it.—Lisa Jardine , Renaissance Studies, Queen Mary and Westfield College, University of London