Our thinking about development has been dominated by the metaphor of the machine. But, as deterministic systems, machines must be predictable to run smoothly, and their behavior is dictated by their structure and parts, along with the instructions of an external operator. The smallest change or variation can bring a machine crashing to a stop.
Using the same mechanistic model, project managers apply multiple controls to reduce variety in the system, with the goal of trying to make input-output relations more predictable and efficient.
When this approach is applied to complex development problems, it falls woefully short. It can even be dangerous. The realities we try to change — the big questions of development — are not machinelike at all. They involve exceedingly complex systems of interaction, coevolution and emergence, which are more like ecologies than machines. In ecosystems, as well as economies, variation or innovation provide the very possibility for change.
Read the remainder of the blog here.