Complicated systems are systems that have been designed by human beings. Assuming they work, they do what they were designed to do. They are predictable. When they go wrong, because we built the system, we can find the fault because we know how to analyse the system to find the problem. Once we find the problem, there the solution will be. We tend to call such systems ‘machines’ because that is what they are – mechanical things. There is beauty in their simplicity, even when it is very complicated like a jet-engine.
Complex systems are different. They were not deliberately designed. Parts of them may have been, but other factors have crept in – often in the form of human beings, or circumstances or chance. These systems are about relationships, not things. They may do what we expect, but they are not predictable. Unpredictable events emerge. When things go wrong, there may be many causes. To put things right, there may be many possible ways to proceed. Such systems are messy, complex and uncertain. There is beauty in their complexity, like a City, but this also makes it hard for us to know what to do when such systems are not working.
The prevailing way of thinking, based on the machine/thing view of the world, assumes that analysis will always find the ‘one true cause’ which in turn will lead to ‘the one true answer’. Measurement, analysis, tools, predictability – these are the assumptions of the prevailing mindset.
Many of us recognise that the complexity of the world cannot be so simply understood or corrected. What we are dealing with may be too complex to ever be fully understood. We know it doesn’t make sense to pretend that a complex, multi-faceted problem can be simplified and solved. We need different assumptions and beliefs, different approaches to make sense of this complexity, to make decisions, to solve problems.
We need to free our thinking – liberate ourselves from this mistake of the intellect by learning new ways of seeing and acting which respond effectively to complexity, messiness and uncertainty.