"The notion of self-organization in human systems may suggest to some that the manager has no real role. The reality is that management has a different role; rather than controller, the manager becomes a liberator."
L. Douglas Kiel, Managing Chaos and Complexity in Government

    Complex systems are comprised of many independent agents interacting with each other in a great many ways. It is the very richness of these interactions that allow the system as a whole to undergo spontaneous self-organization: Atoms form chemical bonds and become molecules, birds adapt to the behaviors of their neighbors and organize themselves into flocks. Teams transcend themselves to become more than the sum of their individual parts. In every case, it happens without a blueprint or conscious plan. The realm of computer networks is a created world, built on an intentionally organic, anarchy-inspiring skeleton. "Chat groups" form around their special interests, linking people from every conceivable background and demographic segment. Just as there is no master neuron in the brain, no computer in the Internet "web" means more than another. In organizations, ties form every time colleagues communicate -- and solidify over time into surprisingly stable networks. To create a enterprise around a living system model is to give an organization the freedom to realize its organic capacity to self-organize without any externally imposed plan or direction. This means that leaders have to understand that the work force has the innate ability to respond to continuous change, to create temporary structures and relationship as required, to experiment, and to find simplified and more effective ways to get results.

    Xerox Corporation found the value of self-organization in the 1980s when it was looking for a way to boost the productivity of its field service staff. An anthropologist from the Xerox Palo Alto Research Center (PARC) traveled with a group of tech reps to observe how they actually did their jobs -- not how they described what they did, or what their managers assumed they did. The anthropologist discovered that the reps spent more time with each other than with customers. They'd gather in common areas like the local parts warehouse or around the coffee pot and swap stories from the field. An old model company manager would have viewed the time spent socializing as a "gap" to be eliminated for higher productivity, but the anthropologist saw the exact opposite.

    For Xerox, the informal gatherings didn't represent time wasted, but rather money in the bank. For it was here, within these self-organized communities of practice, that the reps asked each other questions, identified problems, and shared new solutions as they devised them. And it was through conversations at the warehouse --conversations that weren't part of any formal "business process" or reflected in any official organizational chart -- that knowledge transfer took place. Thanks to the anthropologists, Xerox is now experimenting with "elegantly minimal processes" that under prescribe formal procedures and create more room for local interpretations and innovations, so that more effective business practices can evolve naturally from the grassroots upward, as they should, rather than being imposed from above.

    In 1996, a vice president in the research and development department of Nynex set out to find why some groups were quick to adopt new technologies and others were not. For example, some groups needed 17 days to set up data serves for customers. It was discovered that the departments involved in the process never communicated informally and, as a result, didn't understand one another's roles and needs. When the workers were put together in the same room in an environment that allowed informal groups to form around various tasks, the result was a reduction in the time to provision data services from seventeen to three days.

    Fritjof Capra writes: "Self-organizing systems.., tend to establish their size according to internal principles of organization, independent of environmental influences. This does not mean that living systems are isolated from their environment; on the contrary, they interact with it continually, but this interaction does not determine their organization."

    Self-managed teams, encouraged to keep their membership "fluid" -- to form, adjust the size and composition of the group, even to decide when to disband -- follow these same principles. A 1993 problem-solving study divided management staff into several teams and assigned each to prepare simulated bidding on a government bond auction. Some teams were set up with the same people working together throughout the experiment. In other teams, the membership was fluid; people left or were added as seen necessary by the group.

    The fluid group of teams produced a higher error rate initially, but over time made fewer mistaken decisions. The teams with rigid membership became inhibited by their more limited and structured view of the problem, with the result that fewer alternatives and solutions were produced. The "unstable" teams promoted more divergent outcomes and were more readily adaptive to changing situations.