Wednesday, November 14, 2007

Myerson Not A Historian

Philip Mirowski has been quite critical of an essay by Roger Myerson:
"Some of the most bizarre and outlandish statements about the history of economics have recently been made about the role and accomplishments of the Nash equilibrium concept in game theory. Much of this cynosure is due to the fact that a broad regiment of economic theorists have recently sought to write von Neumann out of the history of game theory to the maximum extent possible, striving to supplant his legacy with the work of John Forbes Nash...
At the end of the twentieth century, this quest to doctor the record with regard to Nash and his equilibrium has attained the status of a public relations campaign with ... a commissioned survey in the Journal of Economic Literature (Myerson, 1999). Although this is not a treatise on the contemporary sociology of the economics profession, a small caveat needs to be inserted here about this unusual effusion of infotainment concerning what must seem, to even the most avid follower of economics, a topic of rather arid compass appealing to only the most limited circle of cognoscenti. Incongruity turns into downright incredulity when one learns ... that the commissioned JEL survey was never vetted by any historian familiar with the relevant events, and consequently the only 'history' to grace the text appears as a word in its title. The reader should approach these texts forewarned that John Nash and his equilibrium, through no effort of his own, have become the object of a vast ceremonial purification exercise." -- Philip Mirowski, Machine Dreams: Economics Becomes a Cyborg Science. Cambridge University Press, 2002: pp. 332-333
And again:
"...the account in Myerson, 1999 of how von Neumann 'failed' has absolutely no relationship to the historical record. The idea that Nash possessed acute insights into human psychology and institutions that von Neumann somehow missed would be ludicrous were it not persistently found in conjunction with the obvious motivation to rewrite von Neumann out of history." -- ibid, p. 347
For example, one would never know from Myerson that Morgenstern and von Neumann analyzed non constant sum games:
"In 1928 and again in his 1944 book with Morgenstern, von Neumann tried to justify this cardinal utility assumption by identifying all payoffs with monetary transfer payments, which led him to the restriction that payoff is transferable and all games are zero-sum...

In 1947 (in their book's second edition), von Neumann and Morgenstern published their third great contribution to game theory: the axiomatic derivation of expected-utility maximization from a substitution argument. This new justification for measurable utility should have prompted them to consider dropping their restrictive assumption that payoffs must be transferable and zero-sum in all games, but they did not." -- Roger B. Myerson (1999). "Nash Equilibrium and the History of Economic Theory", Journal of Economic Literature, V. 37, N. 3 (Sep.): 1067-1082

1 comment:

Gabriel M. said...

I don't understand... is history the monopoly of historians?

We can all disagree with Myerson's claims or text or whatever. But to accuse a survey of "never [being] vetted by any historian" and inferring that it lacks history for that reason.. that's crazy.