Flocking

For many years those who observe the behavior of birds were trying to discover how birds so beautifully flock. The tight coordination of rapid movement among many birds must mean that a very clever and powerful bird is “in charge” and is choreographing the complex flocking. This remarkable leader was never discovered. It seems that birds flock without any one bird being in charge. Rather, each bird moves in coordination with the bird(s) right next to them.

A bird at the outskirts of the flock notices a hawk swoping in and engages in diversionary action. The neighboring birds follow suite and quickly the entire flock is taking this diversionary action. If the birds were operating through a hierarchy, then information about the incoming hawk would be relayed through a set of channels to the lead bird who would then send out a message to all birds in the flock about the diversionary action to take so that the hawk can be avoided. By this time, the hawk would have already hit the birds at the edge of the flock. Delay caused by multi-step communication in the hierarchical structure would reduce the agility of the flock and lead to its vulnerability.

This startling finding regarding the absence of hierarchy among flocking birds soon expanded to the study of many other animals that operate in herds (e.g. Bison), schools (e.g. fish) and other forms of clustering and swarming. Apparently, hierarchical control is rarely found among these animals. In fact, hierarchical control is rarely adaptive in social systems. Furthermore, those who study complex systems find that they are LEAST amenable to hierarchical control is they are large and complex—such as human corporations and most health care systems. We must study and appreciate complex adaptive systems precisely because they point us in directions that lead us away from hierarchy.

Feedback and Delay

The first, fundamental building block in any description of complex adaptive systems concerns the feedback that occurs among components of this system. We refer to each component as an “agent.” An agent, in turn, can be a person, an organization or a specific event. Feedback specifically concerns the influence which one agent in a system has over other agents in the system. This influence typically is engaged through the flow of information (though it can take place through the physical impact of one agent on another agent – the “billiard ball” metaphor is appropriate in this regard).

Feedback can be either positive or negative. These terms don’t refer to the “goodness” of the feedback. Rather, they refer to the nature of influence that the feedback has on the second agent. Positive feedback means that the second agent is more likely to perform in a particular manner as a result of the influence. Negative feedback means that the second agent is less likely to perform in a particular manner as a result of the influence.

The critical factor to note is that a complex adaptive system will only endure if it contains both positive and negative feedback relationships between agents operating within this system. A system with only positive feedback relationships between its agents will soon explode from the exponential growth and pace of action within the system – or this system will collapse as all of its resources are exhausted as a result of its unbridled growth. Some negative feedback is needed to apply breaks and corrections.