Two quick notes in this thread --
I am about to start talking with some people who do complex social systems
analysis using some of the economic models ala Santa Fe, esp. game theory
approaches. Will be interesting to see how the traditions of these
disciplines play out in relation to my own views. I think that a lot of
what complex systems theory has been used to do is more to critique
classical economic theory rather than fully present an alternative (e.g.
Arrow), but of course a lot of people are working on this. Sometimes you
can't make an omelet without breaking some eggs first! :)
In conversational and related sorts of improvisation, I think that there is
again a bit too much emphasis on the originality of what is emergent. That
is what tends to privilege the bottom-up aspects. The top-down perspective
should help us to understand how it happens that people wind up improvising
the same sorts of culturally and historically specific patterns over and
over again!
Typical outcomes do NOT always result from mechanical, algorithmic
procedures. They can and often do also result from more improvisational
convergences of the apparently independent choices of many actors (or at
least two in a classic CA phone conversation). But this is because, so my
view of multi-level relations in complex dynamics says, there are always
already existing longer-term patterns in the community, which have an
impact on each actor, often in ways that we do not pay attention to (as in
the case of the social contexts for learning to walk) because we are so
fixed on individual action as the motor of outcomes.
Emergence normally happens subject to contraint from above (at least for
complex, evolved, developing eco-social systems). It would not happen at
all without such constraint. The term "constraint" should be taken in a
purely formal sense. Functionally such constraints are clearly enabling of
the emergence. Order does not come from nowhere ... it always comes from
prior order (except perhaps at the beginning of the universe, and even
then, who knows?) ... but when new levels are emergent BETWEEN prior levels
of organization, the KIND of order that emerges may be sui generis ... but
the origins of that new order can be traced to the ways in which the higher
level order (N+1) sets the bounds on the combinatorial patterns among units
at the lower (N-1), to 'catalyze' the new emergent order that _is_ level N.
We do not have good metaphors for what I just termed "catalyze" ... the
relationship is not quite the same as "produce" (which is too like cause
and does not apply across levels), or "afford" (which is closer but seems
more like the relation from below, as with 'constitute") ..... words are
wanted!
Finally, as to the issue of whether emergence is in the eye of the
beholder, there is a certain sense in which it is. There is a question as
to what counts as "order" ... and I have found some interesting patterns in
what counts as emergent for us (basically an alternation between
quantitative and qualitative modes of order from one level to the next)
that suggests to me that we operate with two principles of order and that
we make sense of all new order by juggling these two principles. (For the
basics of this, see
http://academic.brooklyn.cuny.edu/education/jlemke/papers/gent.htm , esp.
the appendix.)
Even in the thermdynamics of entropy, there is a certain need for the
analyst to construct a measure which is sensitive to a kind of order, and
otherwise one may not see it .... which is a way of saying that from
"nature's" point of view, it is all the same, whether we see it as
increasing order or not.
Emergence is, after all, an explanatory framework, not an absolute reality
outside of our interests and participation in/with the systems we explain.
It is important, however, not to confuse NEWNESS or the perception that
something new has come along with EMERGENCE of a new levels of organization
and a corresponding new mode of order in a system. Not everything new
represents a new level of organization. It is not enough that something
looks new to us (and maybe does not to the natives of a community); it must
require explanation in terms of a new level of organization, and in my
model that means a distinct timescale and a specific set of relationships
to phenomena at levels/timescales an order of magnitude or more below and
above it.
Conversely, as I said at the beginning of this message, not everything
which is emergent has to be new! ... frequently it is the same old pattern
re-emerging once again, though perhaps each time with slightly different
details in the trajectory that gets us there.
JAY.
---------------------------
JAY L. LEMKE
PROFESSOR OF EDUCATION
CITY UNIVERSITY OF NEW YORK
JLLBC@CUNYVM.CUNY.EDU
<http://academic.brooklyn.cuny.edu/education/jlemke/index.htm>
---------------------------
This archive was generated by hypermail 2b29 : Mon Jul 01 2002 - 01:00:06 PDT