> I would argue just the opposite -
> that because in a social system agents are embedded, the macro behavior of
> the collective is MORE predictable. Embedding serves as a form of
> constraint and access net that channels behavior into specific patterns.
> Kathleen
This is the hope of many modellers of social systems (esp. economists),
but I think that this may be true only in very special circumstances.
The presence of constraints can certainly introduce irreducible
complexity when they are not completely constrictive.
Take an example: the movement of agents in a large space. If all the
agents were completely free to move in any direction they _may_ be
modellable in a way similar to a gas (assuming independance of goals as
well) - where the assumption is that individuals interaction with each
other is minimal (lots of space, few molecules).
Now add some constraints, e.g:
(a) they are crowded, so there are mutual constraints on movement
(b) their goals are linked (so they all want to get to one of a number
of]
hot-dog stands)
(c) they want to avoid dense crowding
The result is less predictable behviour with increased constraints.
Thus _unless_ there are other reasons for this, the macro behaviour will
not (generally) become more predictable with more constraints.
After all, any interaction can be viewed as a constraint!
Regards.
PS. My approach to social embeddedness aims to diferentiate between
systems that will be modellable at such macro levels and those that can
not be.
--------------------------------------------------
Bruce Edmonds,
Centre for Policy Modelling,
Manchester Metropolitan University, Aytoun Bldg.,
Aytoun St., Manchester, M1 3GH. UK.
Tel: +44 161 247 6479 Fax: +44 161 247 6802
http://www.cpm.mmu.ac.uk/~bruce
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|