Let me provide my entry in this very interesting kickoff. Wondering about
the nature of complexity is a great starting issue, I think.
A reminder of where I am coming from: I believe that we will soon begin
serious work on creating a new science that leverages information as a
primitive and emergent behavior as a mechanism or effect. It will be a
universal science, covering chemistry and physics as well as social
behavior but leverage (human) notions of introspection.
I say this, Joe, so you know where I am coming from, and why I chose to
complicate the discussion by wondering about how complexity would fit into
such a framework.
Joe suggests that we look at what we mean by complexity, and he suggests
three approaches: number of different elements, the number of different
ways that they are organized (which I assume includes the number of nodes
that are related to others), or the increased "capability" of structure (or
network). Stanley adds the idea of hierarchy which is a way of structuring
causality. Igor adds a cognitive aspect, internal structure of reflexivity
and assumes that is unique to human systems. Loet adds that old standby:
entropy.
Rather than speculate on whether these are "they ways things are," I would
like to propose a different approach. Let's assume that all these
definitions are elective, that we can choose our definitions of complexity
and subsequently the associated metrics. That's fair enough. Even if you
are a realist, you can accept that the abstractions we choose to model the
world with can be chosen, or even designed.
So if we were going to invent a new science. And it was going to apply at
the social level as well as, say the chemical and biological...
And it was going to apply to synthetic systems as well as natural systems
(allowing that you may consider human systems partly synthetic)...
And it was going to make sense "inside" a system as well as outside a
system looking in...
Given all that (which I suppose to be the FIS agenda) what would the notion
of complexity look like?
I suppose it wouldn't be quantitative in its most natural form. I suppose
it would be a matter of causal relationships in a network sense. I suppose
the relationships would be defined in terms of information somehow. (You
see my own bias in assuming causality here as the basic mechanism - but if
science is anything it is predictive models of cause and effect and only
secondarily measurement.)
I further suppose that hierarchies could be found but only by interpretive
filters - and that these would be easier to justify in the human case.
And here is a speculation I have made before here: I suppose that there is
an upper limit on the complexity that any system can maintain before it
creates a new level of abstractions in creating a new layer. So chemistry
breeds biology as a way of managing complexity. Human systems breed deities
for the same reason.
And another speculation: why should the mechanics of the model be separate
from the model? If the stuff of our new science is information and causal
relationships, is not there a complexity of the model as well? Does not it
have emergent behavior as well? Why should science be fixed dogma, except
to help us pretend the explanations are simple, numeric, static and involve
mystical things like fields, and conditional states?
Just wondering. Complexity is a good place to start in my mind, I repeat.
-Ted
_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Mon Dec 18 16:29:00 2006
