|
|
single (and multiple) cause theories...
Gessler, Nicholas (gessler@ANTHRO.SSCNET.UCLA.EDU)
Thu, 20 Apr 1995 09:51:00 PDT
Diane King has raised an interesting point with regards to limitations of
single cause theories. While it is true that single cause theories can
explain many phenomena, those phenomena of the greatest puzzlement appear to
be complex systems in need of multiple cause theoretical explanations. But
do we embrace single cause theories because we take our cue from the harder
sciences? Not really!
The so-called sciences of complexity have emerged from mathematics, computer
sciences, and have been penetrating the physical, chemical, and biological
sciences. Certainly, much of the work in these sciences is still concerned
with single cause models, but these hard sciences also investigate complex
phenomena for which they've developed a new theory, epistemology, or as one
philosopher of science has put it, a new philosophy of science in itself.
If we look to the hard sciences for guidance, we need to know where to look.
If we look for hard science single cause theories to explain complex cultural
processes, it seems to me that our success will be limited. On the other
hand, if we look for hard science theories of complexity, and try to apply it
to anthropology, we may be developing a powerful explanatory tool for our own
discipline.
I think promise lays in the investigation of computational approaches to
simulation, rather than formulaic approaches. I am not proposing simple
systems simulations which resemble factory process flow diagrams. I am
referring to what are variously called "complex adaptive systems" and
"artificial life." These are computational systems which create a population
of individual agents, which interact with one another, and with their
environment, according to a set of rules which the programmer can either
specify, or which he can allow to evolve on their own. These simulations
generally instantiate the processes of *emergence* of global behavior from
local rules of interaction, *multi-agency* of populations of individual
agents interacting, and *evolution* through natural selection within the
system unguided by the programmer. This kind of theory of complexity is
beginning to be applied to anthropology, in projects which I call "Artificial
Culture," and which others call "Artificial Society" and "Cultural
Algorithms."
Perhaps the reason we often look for simple theories from the hard sciences
is the hope that we can reduce some complex human behavior to simpler
processes: We look for a *different* simple theory to explain our puzzles
when our own simple theories don't work. This may be a fruitful approach
in many cases. But I think we do a dis-service to the hard sciences when we
characterize them as the source of our single-cause limitations. And
ultimately we do a dis-service to ourselves.
There is one nasty thing about complex theoretical models. While they may
accurately capture the events we're trying to understand, the models can be
as conceptually intractable as the phenomena themselves. We may have to give
up the notion that our explanations should provide us with a conscious
palpable grasp or understanding. We may have to settle for understanding how
the pieces work and fit together, and understanding that a computer can track
all the interactions to give us a view of the behavior of the whole set of
processes. We may have to settle for the realization that we still can't
"fathom" or "visualize" the process any more than we can "understand" the
real-time complexity of an automobile racing down the highway.
Nick Gessler
UCLA - Anthropology
|