The concept of self-organisation is a very misunderstood topic when it comes to applying it to organisations.
Bas Reus is exploring what it means to say that humans are self-organising, over at http://basreus.nl/2009/07/27/self-organization-defined/#comment-51. His post outlines the development of his thinking in attempting to define self-orgnisation.
Arising from the study of complexity, the important thing about self-organisation is that the ordering of society (or people in organisations) occurs through local interaction in the absence of an overall blueprint or plan. As any top manager will tell you, you can’t just make a plan, tell others and then confidently expect that the plan will be followed. Instead, all sorts of unexpected things happen – people interpret things differently, they react to things in surprising ways and there are unintended consequences. This is what is meant by saying there is no overall blueprint or plan.
Many people, including me, when they first learn of this idea of self-organisation, immediately think of questions like "How can we empower employees to be self-organising?" or "How can we manage our people so that the emergence can take place?"
In attempting to answer these questions during my doctoral thesis, thanks to my supervisor Ralph Stacey’s empathetic guidance, I came to realise that it doesn’t make much sense to talk in this way because humans already always are self organising, even when they work in organisations with top down management approaches. If they are working in an organisation with restrictive management approaches, then they are still self-organising, with a given mix of constraints, power relations and so on that is determined in part by the management approach. This is because top down or highly directive management approaches give a certain combination of constraints and power relations.
A more useful question might be something like "How can we change the constraints and power relating so that different patterns will emerge from the self-organisation?"
The challenge with thinking about self-organisation for people in organisations is to avoid falling into the trap of thinking that self-organisation means random – or that just anything can happen. Obviously, in an organisation, people cannot just do what they like. And it doesn’t make much sense to me for a manager to allow people just to do anything to respond to the environment – the chances of achieving management goals would be very low.
So self-organisation means something much more subtle than "anything goes." The challenge for managers that is presented by the concept of self-organisation is not "How can I empower my people to be self organising?" They are already self organising (in spite of management directives). The challenge is "How can I influence the constraints and power relationships so that different (hopefully more desirable) patterns of social interaction emerge." If different patterns of social interaction emerge, then along will come innovation and different results – the actions of the manager will play a big part in whether those results are more desirable or less desirable – so we cannot just say "anything goes."
The big insights of self-organisation are 1) the recognition that managers have a lot less control than the dominant managerial literature would have you believe, and 2) that managers themselves are also part of this self-organising dynamic of local interaction.
This means that as a manager you can only influence your organisation from within your own local interaction with others. So you must pay attention to your own interaction, observe what results and adjust as you go along.