Sunday, July 23, 2006

Attributes of Complex Adaptive Systems (Part I)

The Simply Complex Law
By J.B. Ruhl
Post 3: Attributes of Complex Adaptive Systems (Part I)
July 23, 2006

In the previous post I suggested that the evolutionary path of a CAS is difficult to predict notwithstanding that all the agents in the CAS follow rules of behavior. Complexity theory uses a number of formal terms and concepts to describe what is happening in and to a complex adaptive system (CAS). This is the lexicon I will use in the series, with additional terms introduced as needed in later posts. The terms described in this post focus on the internal dynamics of a CAS.

: Because each agent in the system applies a set of rules to incoming information to decide what to do in the next system move, it is possible, even likely, that what any agent does in a move will eventually affect the information being received by that agent in a later move. This kind of feedback can either reinforce or mitigate the agent’s behavior.

: The micro effects of each agent’s actions aggregate with those of other agents at successively macroscopic scales to produce emergent behavior of the system at different scales. Feedback flows begin to move not only horizontally between agents, but also between system scales. The result is that the system behavior at “higher” scales cannot be understood through reductionist study of the individual agents

: With feedback flows and emergent behavior set into motion, the evolutionary path of the system becomes nonlinear over time. Large perturbations from the exogenous environment thus may have small effects on the system, and small perturbations may have large effects on the system.

Sensitivity and Path Dependence
: A CAS is sensitive to conditions at any particular move and its point on its evolutionary path thus is a product of all prior moves. Hence, if one could observe two absolutely identical CASs at time n and alter conditions in one infinitesimally compared to the other, over time the paths of the two CASs could depart substantially.

: The fitness of a CAS is governed in large part by its resilience—the ability to adapt so as to maintain its basic structural dynamics while taking advantage of its environmental conditions and responding to perturbations.

Conflicting Constraints
: Adaptation usually presents trade-offs, such that evolving a particular set of attributes to improve resilience to one type of environmental condition or perturbation can impair resilience to another type of condition or perturbation.

: The evolutionary path of the system may cross thresholds, or tipping points, past which system dynamics are massively and irreversibly altered. It is generally believed, however, that the most resilient CAS behavior exists at the edge of these critical thresholds. CASs, in other words, evolve toward self-critical states.

Stable Disequilibrium
: A CAS might “sit” in a self-critical state for quite some time, which an observer might mistakenly interpret as some form of climax equilibrium stage of system evolution. In fact, however, what keeps the system self-critical is its constantly dynamical behavior.

Next: Co-evolutionary properties of complex adaptive systems.


Post a Comment

<< Home

Web Jurisdynamics