Emergence and self-organization

Many people say that open source software developers have the most efficient ecosystems for learning that have ever existed. What is it, then, that is so special about the way developers do things? Is there something that could act as a model for the future of work, or the future of education?

What takes place in open source projects is typically not the result of choices made by a few (powerful) people that others blindly implement. Instead, what emerges is the consequence of the choices of all involved in the whole interconnected network, “the connective“, as Stowe Boyd puts it. What happens does not follow exactly a plan or a design, what happens emerges. It is about the hard to understand process of self-organization.

We still don’t quite understand what emergence and self-organization mean. The problem is that we believe that the unit of work is the independent individual. Self-organization is then thought to mean that individuals organize themselves without the direction of others. People think that it is a form of empowerment, or a do-whatever-you-like environment, in which anybody can choose freely what to do. But connected people can never simply do what they like. Cooperating individuals are not, and cannot be, independent. People are interdependent.  Interdependence means that individuals constrain and enable each other all the time. What happens, happens always in interaction and as a result of interaction.

According to the present approach to management, planning and enactment of the plans are two separate domains that follow a linear causality from plans to actions. From the perspective of open source development, organizational outcomes explicitly emerge in a way that is never just determined by a few people, but arises in the ongoing local interaction of all the people taking part. For example GitHub “encourages individuals to fix things and own those fixes just as much as they own the projects they start”.

What emerges is, paradoxically, predictable and unpredictable, knowable and unknowable at the same time. This does not mean dismissing planning, or management, as pointless, but means that the future always contains surprises that the managers cannot control. The future cannot be predicted just by looking at the plans.

Emergence is often understood as things which just happen and there is nothing we can do about it. But emergence means the exact opposite. The patterns that emerge do so precisely because of what everybody is doing, and not doing. It is what many, many local interactions produce. This is what self-organization means. Each of us is forming plans and making decisions about our next steps all the time. “What each of us does affects others and what they do affects each of us.”

No one can step outside this interaction to design interaction for others.

An organization is not a whole consisting of parts, but an emergent pattern in time that is formed in those local interactions. It is a movement that cannot be understood just by looking at the parts. The time of reductionism as a sense-making mechanism is over.

What we can learn from the open source ecosystems is that organizational sustainability requires the same kind of learning that these software developers already practice: “All work and learning is open and public, leaving tracks that others can follow. Doing and learning mean the same thing.”

The biggest change in thinking that is now needed is that the unit of work and learning is not the independent individual, but interdependent people in interaction.


Thank you David Weinberger, Ken Gergen, Ralph Stacey and Doug Griffin

More on the subject: the GitHub generation, Sugata Mitra. Video: “Knowledge in a MOOC” Steve Denning on complexity. The mundanity of excellence.