As I missed his presentation, I checked out Axelrod's paper on his website.
I must admit I'm having difficulty getting my head round his philosophical
approach and I'd appreciate some comments.
In "The Evolution of Ethonocentric Behavior", Axelrod makes some big claims
- "Ethnocentrism may have a more complex cognitive basis - but our model
shows that it needn't", ". the present model demonstrates that none [of six
quoted mechanisms] are necessary for ethnocentric behavior to have evolved
or sustained." In my view, claims like this entail a substantial burden of
proof and I don't see how Axelrod has met it.
In the first place, the bulk of the studies he quotes in his literature
review are anthropological/sociological studies of real people and real
communities, not virtual critters moving round a checker board. Axelrod
makes an implicit equivalence between the two that I don't think he supports
convincingly. This surprise me, considering his criticism of game theory: ".
its unrealistic assumptions may undermine much of its value as a basis for
advice" (see "Advancing the Art of Simulation in the Social Sciences" - also
on his website).
Secondly, even assuming this people/critter analogy is reasonable, I don't
see how his actual results support his conclusions. The fact that
ethnocentric behaviour only emerges if you have local reproduction and local
interaction (see p14 and table 2) suggests that the observed ethnocentricism
is a modelling artefact: it's a result of fixing these critters immobile on
a checkerboard rather than occupying some (more realistic) non-local or
small-world network.
These may seem like quibbles, but it strikes me that if an author is making
big claims, his conceptual or philosophical foundation needs to be pretty
solid. I just don't get that feeling with these papers. Or am I missing
something?
Robert
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://constantinople.hostgo.com/pipermail/friam_redfish.com/attachments/20040216/c015910b/attachment.htm
Locked
