Desirable Properties
-
Resilience - the ability of the system to survive and persist within a variable environment.
- It arises from a rich structure of many feedback loops that can work in different ways to restore a system even after a large perturbation
- The system should have many, different, and potentially redundant balancing loops.
- Meta-resilience comes from the ability to restore or rebuild feedback loops.
- Meta-meta-resilience comes from the ability of feedback loops to learn, create and evolve complex, restorative, self-organizing structures.
- Resilience does not imply constancy, in fact, rigidity proves to be the opposite of resilience.
- Resilience is not so obvious without a whole-systems view. It is often sacrificed for something short-term.
- Resilience expands the possibilities afforded by the system, allowing it to explore more states without destabilizing.
-
Self-organization - the capacity of a system to make its own structure more complex.
- It produces heterogeneity and unpredictability. It can make new structures, but it requires the ability to experiment.
- These systems diversify and complexify, often from just simple rules.
-
Hierarchy - as a consequence of self-organization, a hierarchy emerges.
- A complex system can evolve from a simple system only if there are stable intermediate forms.
- They also serve to reduce the amount of information that any part of the system has to keep track of.
- They can be taken apart, and the subsystems formed can still function in their own right.
- When a system breaks down, it splits along its subsystem hierarchies. Thus, a reductionist point of view has merits in system science (although remember that subsystems also interact with each other)
- When a subsystem’s goals dominate at the expense of the total system’s goals, the resulting behavior is called suboptimization 1. Of course, there is a balance that must be struck between centralized and decentralized approaches
- Hierarchical systems evolve from the bottom up. The purpose of the upper layers is to serve the purposes of the lower layers.
Links
Footnotes
-
In other words, it is not Pareto optimal ↩