Science in Society

Bryn Mawr College

Emergent Systems Working Group

November 15 and December 6, 2002
Paul Grobstein
From "Complex Systems" to "Emergence": A General Conceptual and Philosophical Perspective?

Prepared by Ted Wong
Additions, revisions, extensions are encouraged in the Forum and/or at
Participants 15 Nov

10 years separate Waldorf's book "Complexity" and Johnson's book "Emergence." 10 years changed our thinking (though subtly) form complexity to emergence, but complexity is the product of 150 years of thought, at least. That is, the field expanded in the 1980s, but the ideas that gave rise to the expansion are old.

What happened in the 1980s? Computers: easy-to-use desktop computers appeared. Computers are like telescopes: they expanded our range of perception -- in the case of computers, into the world of iterative models in discrete time. Innovation: treat continuous time as discrete. Apply a rule at t, reapply it to the results at t+1, and so on.

The difference between emergence and complexity is that with emergence we look at the order that comes from non-equilibrium systems. This order is at all levels of reality. Systems create order in flux. So, in the solar system, we have a stable system arising from disorder.

The similarity between emergence and Darwinian evoluton is not accidental. We have simple interactions giving rise to complex behavior. [PAUL, interesting that it's so natural to use up-down metaphor for the LEVELS of organization. I wonder if everyone naturally knows that UP is for more aggregated levels. Maybe it only happens that way because people transmit the metaphor by using it, and that's the way they use it.] In emergence and in evolution, we have:

Conway's Game of Life. Order from initially random configuration. There is no planner, no global rule. We have context: in this case, the number of cells and the topology of the world.

Per Bak's pile of sand. Dribble sand from your fist into a conical pile. The shape of the pile is stable, and it arises not from a plan or a global shaping mechanism, but by the local interactions among the grains. The shape is a stable state, but it is always changing, fluctuating. But what is the context??

Paul Grobstein

"Complexity" was an exciting new interdisciplinary area of inquiry that emerged in the 1980's (cf. M. Mitchell Waldrop, Complexity, Touchstone, 1992), in large part because computers made it possible for the first time to make observations on the behavior of "iterative" systems, systems of interacting relatively simple but non-linear elements whose overall behavior can be determined only by successive application of rules which take the current state as input to establish the subsequent state. The hope/expectation was that there would emerge general rules of such iterative systems which would be applicable to phenomena ranging from physics through chemistry and biology to social systems. A paper in 1988 tried to deduce some of these general principles on the basis of similarities between brain function and developmental processes. In a course in 1995, an effort was made to summarize in general terms the emerging Insights from Complex Systems.

The current rubric of "emergent systems" (cf. Steven Johnson, Emergence, Touchstone, 2001) is a noteworthy extension of this area of inquiry, one which helps to link a variety of relatively independent lines of investigation in recent years (cf. Forum) with a likely origin in Darwinian thought. It also helps to define new questions for continuing exploration (see Complexity -> Emergence: Notes, Nov 2002).

A general problem, apparent in all sciences but particularly dramatically in biology is that of the origin of "levels of organization" (eg subatomic particles to atoms to molecules to prokaryotic cells to eukaryotic cells to multicellular organisms to groups of individuals to societies). Existing complex systems/emergence models poorly replicate this kind of progressive emergence. And should, if the basic concept of interacting simple elements yielding progressive complexity is indeed quite generally applicable.

Several characteristics, over and above the basic insight of interactive, distributed systems (cf. Swarm Intelligence), may be necessary to create full blown emergent systems:

Broader implications, if successful:

Additional issues:

Home | Calendar | About | Getting Involved | Groups | Initiatives | Bryn Mawr Home | Serendip Home

Director: Liz McCormack - | Faculty Steering Committee | Secretary: Lisa Kolonay
© 1994- , by Center for Science in Society, Bryn Mawr College and Serendip

Last Modified: Wednesday, 02-May-2018 11:57:03 CDT