Paul Erickson, The World The Game Theorists Made (Chicago: University of Chicago Press 2015) 390pp. $ 35.00 Pb, $105.00 Hb. ISBN: 9780226097176
In Robert Coover’s lamentably obscure 1968 novel, The Universal Baseball Association, J Henry Waugh, Prop., a socially-retiring accountant becomes increasingly absorbed into the private world of his fantasy baseball league. Each night after work, J Henry Waugh takes out his dice and plays his game, tabulating the outcomes in a vast ledger, 'The Book.' Henry layers over these numbers a rich universe: he creates generations of players, maps the narrative of their careers, but the decisions are all driven by the rules of the game and decided by the roll of his dice. Henry can populate his teams, make them villains or heroes, but he’s not free to alter the results. Random fate – those dice – meant that while he had created this world, he couldn’t fully control it, and ultimately, it comes to control him.
At first, the man plays the game, but in the end, the game plays the man: Coover’s trope is familiar – it’s Dostoevsky’s The Gambler, or Luke Rhinehart’s The Dice Man, or Iain Banks’s The Player of Games. By generating complexity through the complication of simple rules, games recreate the human condition in miniature, and, metafictionally, point to how the particularities of the novel map outward to the universal.
But with the publication in 1944 of Theory of Games and Economic Behaviour, by mathematician John von Neumann and economist Oskar Morgenstern, that metaphorical idea tipped over into a more literal sense: perhaps, they suggested, human interactions really could be perspicaciously understood in the same terms as a game? After all, if a game, broadly understood, was the competition between interested actors within the boundaries of a particular set of rules or laws, then almost every human activity was susceptible to a game-description.
Building on work von Neumann had done in 1928 on the mathematics of parlour games – which was itself a sort of mathematical parlour game – they now approached the mathematics of games more robustly: schematising strategies for 'zero-sum' (win-or-lose) games, and drawing up the now-iconic matrix to locate the strategy by which a player could balance maximum gain against minimum loss.
Like Henry’s fantasy baseball, the mathematical games were given elaborate illustrative settings. By far the most famous of these was Merrill Flood’s 'Prisoner’s Dilemma', which conjured a scenario of trust and betrayal, the four cells of the matrix becoming prison cells. The analogies to life were suggestive: a heavily embroidered version of the Prisoner’s Dilemma would form the opening chapter of Ian McEwan’s Enduring Love; for Richard Powers it was an extended metaphor for family commitments in a novel called Prisoner’s Dilemma.
Layering narratives over the numbers made them intoxicating, but in the 1940s, the life-as-game metaphor was ripe. The aleatory dimension of the new physics had prompted Einstein’s famous riposte that 'God does not play dice with the universe' (on which Coover’s novel – where the action in J-H-W’s Book is entirely decided by dice – was an extended satire). Meanwhile, the looming dread of nuclear apocalypse added another dimension: games offered an abstract mode through which the madness of nuclear war might be refracted into more manageable terms. Thinking of nuclear war as a 'game' was nauseating, but if everything was a game, then games were worth taking seriously. Game Theory threatened to trivialise life even as it valorised play.
Paul Erickson’s The World The Game Theorists Made takes on the history of game theory’s diffusion – from its origins in mathematics and economics, through to its use by the US military in planning and logistics, to RAND and the nascent behavioural sciences, and then to the Dr-Strangelove-world of nuclear strategizing, and finally to evolutionary biology, where it would come to underpin the logic of neo-Darwinism and the selfish gene.
Yet as Erickson’s narrative makes clear, the apparent success of Game Theory conceals a persistent dissatisfaction with its actual usefulness. As it flits between disciplines, Game Theory looks very much like an intellectual fad – a butterfly progress replete with a butterfly lifespan. While never being entirely abandoned, Game Theory certainly seems to have reliably disappointed its numerous disciplinary hosts. Erickson’s history of the idea sees Game Theory enthusiastically adopted only to be supplanted or rejected soon after. The value of such a history is that it reveals that the persistence of an idea can involve a kind of disciplinary crowd-surfing, where its visibility is a function of its transience, being constantly passed on to the next pair of hands.
While much of Erickson’s book will certainly be more detailed than scholars working at the intersection of literature and science will require, his dogged tracking of the theory from research centre to corporate think-tank to political war-room, and of the disputes between the various proponents and users of the method, is a reminder that theories are rarely as homogenous as they seem from the outside. Consensus is a trick of distance, and there have been subsequent calls to speak not of 'game theory' but 'game theories' to acknowledge this.
Some of the most interesting disputes involve the extent to which Game Theory might act as a guide to rational behaviour. For a certain type of thinker, what made Game Theory attractive was that it obviated moral decision-making. A matured Game Theory might supplant our existing ethical theories entirely – just as modern medicine had supplanted traditional medicine. One needn’t ask if something was good or bad, simply whether it was rational – 'rational' by the lights of game theory, at least.
This amoral calculus finds its most chilling embodiment in the 'Doomsday Machine' proposed by Cold War strategist Hermann Kahn. If or when nuclear conflict looked inevitable, the plan was to hand control over to a computer that would finish the job, triggering the 'spasm war' that would guarantee global annihilation. A machine could reliably do what a sentient person might lack the stomach for; computers put the assurance in mutually assured destruction.
Among a fairly enormous cast, one of the heroes of Erickson’s book is the mathematical psychologist Anatol Rappaport, who felt that it was a mistake to read game theory as offering a prescriptive account of rationality. When Rappaport challenged Kahn on his Doomsday proposal by asking how he would defend himself in a genocide trial, Erickson records that the audience baulked: 'academic decorum was breached: […] questions of ‘morality’ were judged impertinent in such a conversation' (197).
Ultimately, however, it was 'rationality”'that would be driven out of game theory. Evolutionary biologists and geneticists had long argued over the level at which selection operated: if you were a population biologist, it seemed as if species were in competition with one another, acting in concert for a common good. Nature exhibited homeostasis – a convergence on steady ratios of predator to prey or male to female. But the geneticists objected that there was no means by which 'species' could make choices – from what Richard Dawkins’ would memorably call the 'gene’s eye view', every transaction occurred at the individual level. Unless you wanted to introduce some higher power steering events along (and Dawkins obviously didn’t), you’d need a way of explaining how this strategic looking behaviour was emerging.
Via early computer simulations in which automata obeying simple rules generated surprisingly complex behaviours, Game Theory – or yet another new version of it – provided just such a mechanism. The question of whether games were a good way to think about life now took on a rather different meaning: competition between human players involved strategy and choice – the very 'rationality' that had been at the centre of (and a stumbling block for) so many of the attempts to apply Von Neumann and Morgenstern’s work. But competition between less cognizant organisms – fruit flies, for example – could make no such appeal to intentionality.
What had begun as an analysis of human rationality had within a few decades become a means to explain it away. The Game-Theory inspired view of evolution that W D Hamilton, Robert Trivers, and the tragic George Price (who became obsessed with recovering the altruism his equations had stripped out, gave away all his possessions, and died homeless), drops us into a world where our behaviour, whatever the proximate motive, is also a cryptic move in a much longer game. We thought we were playing the game of life, but it was playing us all along.
The World The Game Theorists Made is not always an easy read, and although nearly a third of the book is endnotes and bibliography, there are times when even more of the text might have been stored out of sight. Be that as it may, Erickson marshals his huge archive with exemplary confidence, producing what is likely to be a definitive history of a hugely influential idea.
Jon Adams, London School of Economics
