I have received the following snippet off-list:
>There are two Herbert Simon awards. One is quite standard: seven years in <http://www.themercury.co.za/index.php?fSectionId=283&fArticleId=2875466>
>existence, given regularly by a committee of the Society for Design and Process
>Science, consisting of a gold medal and $1000. The other is the two-year-old
>brainchild of a one-man organization called the New England Complex Systems
>Institute and consists of a wooden plaque
>
So I thought there could be several Gandhi awards, too. See, for example,
http://www.themercury.co.za/index.php?fSectionId=283&fArticleId=2875466 <http://www.themercury.co.za/index.php?fSectionId=283&fArticleId=2875466>
From the number of off-list emails I have had applauding my posting
about the Herbert Simon award and the nature of the on-list responses,
it seems that there are at least two quite distinct camps subscribing to
this list. And it is a long time since we have had any substantive
discussions here. So I offer the following in hopes of starting another
thread. With sufficient contributions to the thread, the subject line
might make Google and turn into an urban myth. :-)
A brief statement of the difference between Simon and those of us who,
unlike game theorists, were influenced by him might go something like this:
Simon wasn't just a bottom-up modeller, he used computational
methods to try to capture how individuals actually behave. Building
on that approach, others (Epstein and Axtell come naturally to mind)
developed computational representations of interaction among agents
meant in some way accurately to describe actual individuals. Nash
and those he has influenced are indeed top-down modellers who design
agents not to describe observed behaviour but to fit it with game
theoretic constructs and to be consistent with the existence of Nash
equilibrium.
This is not the old bottom-up/top-down issue we had previously. It is
that other hoary old chestnut about whether social analysis should be
evidence/observation-driven or theory driven. For example, Jonathan
Bender suggested on-list
some of his [Simon's] ideas are CONSISTENT w/ strategic analysis.for
ex., one can analyze under what conditions a set of satisficing
agents will converge to Nash.
This is only an interesting question if one accepts a very narrow
definistion of strategic analysis. For myself, I cannot imagine the
circumstances under which it would be remotely interesting to "analyze
under what conditions a set of satisficing agents will converge to
Nash." But I am absolutely certain that the representation of
satisficing required for such an analysis could hardly be more remote
from anything Simon wrote or believed.
Another suggestion was
I seem to remember having read in one of Weibull's papers that Nash had ideas
similar to the evolutionary stability argument in his thesis
Maybe Nash did, but surely Simon never contemplated evolutinary stability.
I think the most judicious response I have had was the following:
I think that you could make a link in a very weak way by saying that
Nash brought attention to the importance of agent interactions, esp.
expectations of others' behavior, in economics. But I can't take
it much further.
I would hate to start a textual exegesis of Simon's writings. Perhaps
the substantive issue here is whether those of us who develop simulation
models to describe observed social behaviour and interaction without any
consideration of equilibrium social outcomes (and, in so doing, claim
Simon as an intellectual forebear) have anything in common (whether
through Simon or otherwise) with those who start from theoretical
structures such as game theory (whether top-down or bottom-up) in the
designs of the agents and mechanisms.
--
Professor Scott Moss
Centre for Policy Modelling
Manchester Metropolitan University
Aytoun Building
Manchester M1 3GH
UNITED KINGDOM
telephone: +44 (0)161 247 3886
mobile: +44 (0)7740 942564
fax: +44 (0)161 247 6802
http://cfpm.org/~scott
|