INSIGHTS FROM COMPLEX SYSTEMS
Scientists and non-scientists alike are in the business of trying to make sense of the world. And in so doing, both scientists and non-scientists make use of conceptual frameworks, habitual ways of thinking which influence both how one tries to make sense of new observations and the new questions one asks (and doesn't ask).
These conceptual frameworks are themselves a product of trying to make sense of the world, and hence themselves reflect the kinds of observations that have been made and can be made. Computers, like telescopes and microscopes, have opened a whole new world of possible observations. Because of the rapidity with which they can do well-defined calculations, computers have made it possible to explore the consequences of relatively simple interactions of relatively simple things in a way never before possible. And, in a variety of different disciplines, it is similarly emerging that this new capability for observations makes possible significant insights into phenomena long believed to be too complex for serious analysis.
Perhaps even more importantly, there is emerging a new and quite general conceptual framework, one equally useable in a variety of different sciences ... and by non-scientists as well. The following is an effort to begin characterizing that framework, in a form which can be used by everyone interested in trying to make sense of the world.
This is a work in progress (see November 2002 notes), so if you're interested, please return from time to time. And if you'd like to help, drop us a line
- Many (all?) interesting phenomena can usefully be described as "orderly ensemble properties" and productively understood in terms of the properties and interactions of sub-phenomena ("elements").
- Ensemble properties are permitted by but not determined by element properties
WHOLES ARE MORE THAN THE SUM OF THEIR PARTS
- The behavior of ensembles is both influenced by and influences the behavior of elements
THERE IS A RECIPROCAL CAUSAL RELATIONSHIP BETWEEN PARTS AND WHOLES
- Orderly ensemble properties can and do arise in the absence of blueprints, plans, or discrete organizers.
INTERESTING WHOLES CAN ARISE SIMPLY FROM INTERACTING PARTS
- Ensemble properties may be largely unaffected by variations in the properties and behavior of elements
HOLISTIC PROPERTIES MAY APPEAR RESISTANT TO CHANGES IN PARTS
- Ensemble properties may be highly sensitive to variations in the properties and behavior of elements
HOLISTIC PROPERTIES MAY SUDDENLY AND APPARENTLY MYSTERIOUSLY CHANGE
- Ensemble properties can be dramatically changed by modifying the nature of the interaction among elements
ENUMERATION OF PARTS CANNOT ACCOUNT FOR WHOLES
- Ensemble properties may be dynamic for reasons entirely internal to the ensemble
CHANGE DOES NOT NECESSARILY INDICATE THE EXISTENCE OF AN OUTSIDE AGENT OR FORCE
- The same change in element property or behavior may have a small effect on ensemble order at one time and a large effect at another time
THE RELATION BETWEEN PARTS AND WHOLE MAY ITSELF CHANGE FOR A GIVEN WHOLE.
- Disorderly variations in element properties or behavior may be the driving force for ensemble order
INTERESTING WHOLES CAN ARISE FROM CHAOS OR RANDOMNESS
- Deterministic systems will not explore all possible ensemble states
RANDOMNESS PLAYS AN IMPORTANT ROLE IN THE EXPLORATION OF POSSIBLE WHOLES
Try The Voyage to Serendip, and remember the Serendip principle on future voyages of your own (requires a Java compatible browser)
by Paul Grobstein, in consultation with colleagues in Bryn Mawr Biology 367, Computational Models of Biological Organization: Sarah Blankenship, Jane Lui, Jeff Oristaglio, Jennifer Santos, Beth Tinker. Original, spring 1995. Updated 7/28/97.
Updating notes (from August, 1998):
Additional complex system resources