Complexity & Nonlinear Dynamics

Gessler, Nicholas (gessler@ANTHRO.SSCNET.UCLA.EDU)
Mon, 20 Jun 1994 10:42:00 PDT

It's timely to see O'Brien's posts on NONLINEAR ATTRACTORS, Garson's post on
the Social Science Computing Association, and Robinson's post on computer
usage in anthropology.

I just returned from a workshop on Multi Agent Simulation Systems (MASS or
SWARM) at the Santa Fe Institute, which many of you know is dedicated to
scientific research into non-linear research and complexity and is highly
respected for its contributions in the field. The goal of the meeting was
to survey multidisciplinary progress in software design to assist the SFI in
developing a general purpose software platform (called SWARM) for models in
which "intelligent" autonomous agents interact in some kind of space with a
given physics. There were many informative discussions.

My own presentation focused on modeling culture from the bottom-up, by
beginning with individuals in a geographic physics and space operated either
with declared or evolving rules and environments. I saw promise for this
technique in testing the characteristics of the (causal or
constraining) linkeages in Harris's "Universal Pattern of Culture"
(infrastructure, structure, superstructure), in assessing the normative
simplifications/assumptions of participant homogeneity, temporal consistency,
and referential accuracy of "rules," and in exploring the methodological
problems of doing ethnographic fieldwork by placing an embodied
"ethnographer" into the system to subject the simulation to the natural world
problems of sampling error and Heisenberg-like uncertainty. I concluded
by saying that in the long term, the technique should help us to formulate
problems and solutions in understanding hominidization, culture change, and
cultural evolution.

One question (among dozens)all participants were concerned with was where to
concentrate these efforts. Many wanted the platform to run on massively
parallel machines, and many (even some of the same) people wanted scaled-down
versions to run on PCs and Macs. I was felt that if outreach to oter
disciplines was important, particularly anthropology, the software platform
needed to be scaled down to run on PCs and Macs.

So, please consider this a formal request for feedback from the discipline:
Anyone interested in working with a general purpose platform such as the one
described, please send me a PRIVATE e-mail and I will pass that information

Subject: non-linear dynamics


What computer(s) do you presently use? PC, Mac, etc.

What hardware and perating systems do they have? DOS, Windows, etc.

What computer languages (if any) do you program in? Basic, Fortran, Pascal,
Prolog, C, C++, etc.?

What anthropological problems do you wish to investigate?