Complexity management, as opposed to traditional practices, may be defined simply as the management of complex systems. So, before we can go much further, we really need to give some outline as to what exactly a complex system is. Thus, this article tries to give a clear outline to the differences between simple linear systems and complex systems.
Firstly, complex systems are a type of system, where a system is just a set of things that perform some collective function. So the human body is a system in that it consists of many individual organs that work together as a functioning entirety. A business is another example of a system – many different individuals and departments functioning as an integrated unit, to collectively produce some set of products or services. And of course, there are many other examples of systems such as transportation systems, ecosystems, information systems and so on.
Not everything is a system though. If we take a random collection of things, say a hard drive, a light, and a watch and put them together, this is not a system. It is simply a set of elements because they are not interconnected and interdependent in performing some collective function. It is because of the fact that the elements within a system perform some collective function that systems are said to be greater than the sum of their parts. That is to say, that the system as a whole has properties and functionalities that none of its constituent elements possess. A plant cell is an example of this. It is composed of many inanimate molecules, but it is only when we put these together that we get a cell with the properties of a living system. So it is not any element that has the properties of life, but it is the particular way that we arrange these molecules that gives rise to this emergent property of the living system as an entirety.
So that’s a very quick overview to the idea of a system. But systems can be defined as either simple linear systems or complex nonlinear systems. They are classified like this because each has very different properties and behavior and thus, how we approach trying to study, design or manage them changes fundamentally.
All systems start out simple and they evolve to become complex. Systems have a number of properties that make them more or less complex. These include the number of elements within the system, the level of interdependence between elements, the degree of connectivity between those elements and their degree of autonomy and diversity. Turning these properties to the system up or down makes them more or less complex, so we will go over each of these in more detail.
Firstly, the number of elements within a system – this is quite straightforward. The more parts there are to our organization, the more complex it will be. In true complex systems, the number of elements is for all intensive purposes virtually infinite. Think of the number of devices within a city – it is so many that we could not itemize them. Linear systems consist of a limited finite amount of components, where it is possible to know and itemize each element in the system, making it possible to define a closed system and say what it part of the organization and what is not. With complex systems, we typically can’t do this, which makes them open systems – we never have full knowledge or understanding of all the parts and all the interactions. The internet or financial markets are good examples.
Secondly, interdependence. Linear systems have a limited amount of elements interacting in a well defined linear fashion, meaning cause and effect are directly related. If I hit a ball with a bat, it will move off in the opposite direction – that is linear causality. It is very simple and intuitive to us, we are programmed to look for and try to understand the world through simple cause and effect interactions, which works well when dealing with simple systems. But, as we turn up the interdependency between the parts, this linear causality starts to break down.
Interdependence means the way we put things together and what we put together matter. Like putting two chemicals together, depending on which ones they are, we may get a very different outcome. They may work together giving us an outcome much greater than we expected or have no effect on each other at all. Coupled with this, events may be interdependent over time, feeding back on themselves to have a compounding effect over time, like compound interest. The result can be exponential growth or decay as things change very rapidly. With these feedback loops, small events can get amplified into large systemic effects and this is part of the idea of the butterfly effect and chaos.
Next, the degree of connectivity. Complex systems are known to be highly interconnected. This is why they are typically modeled as networks, that capture this high connectivity. Linear systems have a low level of interconnectivity between their parts. Thus, we define them in terms of these isolated parts and their properties. However, in the highly interconnected systems such as the internet or the global air transportation system, it is the structure of the network that comes to define the system. That is to say, what is connected to what and where you lie in the network. Are you a central hub like London in the financial markets, or are you on the periphery? What is flowing on the network? How evenly are things distributed out? These and similar other questions relate to the structure of the connections. Connectivity and access are central to understanding complex systems, and network theory is the language through which we do that.
Lastly, complexity is a product of the degree of autonomy and capacity for adaptation of the elements within the organization. A corollary to this is diversity. When all the elements within our system are very similar or homogeneous, then it is much simpler to model, design or manage. As opposed to dealing with a heterogeneous organization composed of many diverse parts – each with their own unique set of properties. When the elements have a very low level of autonomy and diversity, then the system can be designed, managed and controlled centrally in a top-down fashion. But, as we increase the autonomy of the elements within the system, this becomes no longer possible as control and organization become distributed out. Thus, it is the interactions on the local level that come to define how the system develops.
This gives rise to another important feature of complex systems, that is self-organization. When elements have the autonomy to adapt locally, they can self-organize to form global patterns of organization. The process through which this takes place is called emergence. Thus, as opposed to simple linear systems where order typically comes from some form of top-down, centralized coordination, patterns of order within complex systems emerge in a distributed fashion from the bottom up.