Critical Approach to Complexity
Conventional reductionist methods fail considerably while dealing with Complex systems (Morin 1992). By modelling a complex system, a process of reductionism is undertaken but there is a need to make the models more inclusive to be successful. In order to do so, a critical analysis of modeling complexity is undertaken.
Inspiration from nature is not uncommon in most fields. Though it’s a common understanding that scientific research has all the answers, Science is still trying to understand complex behaviours in nature. But Science has progressed in leaps and bounds and has dealt with modelling complexity for more than a century now. Though modelling complexity can never be complete, science has taught us that what is modelled and what is not should be comprehended for a successful outcome.
Complex systems occur in nature and should be studied as a system. Complex systems have the ‘property to change, grow and perish and thus adapt’ (Kitchley, L.J., Aravamuthan, S.2014, pg636).
Paul Cilliers in his 1998 paper ‘Complexity and Post Modernism’ offers a list of characteristics of a complex system…
- ●
Complex systems have many interdependent components that influence each other.
- ●
The complex and dynamic interactions between the components have the properties of Nonlinearity. This means that the size of the cause and effect need not be proportional.
- ●
The interaction between some components constantly creates feedback loops that could either enhance or inhibit the interactions.
- ●
The components do not know or understand the whole system - they only interact with the nearest neighbour components. Though the whole complexity does not reside in the single component, the interactions are rich and have multiple routes of enhancement /suppression / alteration which promote wide ranging influence.
- ●
Emergence is a property of complex systems that cannot be ‘pre-established but is born over time and can be explained by the organisational structure or the algorithm governing the complex system’.
- ●
They are not stable and the extent of the system is usually determined by the position and purpose of the observer. This phenomenon is referred to as Framing.
- ●
The dimension of time is very active in complex systems and the past history and the lessons learnt determine the present behaviour.
On Critically analyzing the above characteristics of a complex system, the following insights are made by Paul Cilliers (2005), The degrees of freedom established in the system, i.e., the looser the structure of the system is, determine the randomness of behaviour of the structure. This results in less functionality. Therefore, complex systems need to have a balance between flexibility and constraint to be more effective.
- ●
Description of a complex system depends on the perspective in which it is looked at and the number and type of characteristics that is taken into account. All of its characteristics cannot be taken into account and the results depend on which of them is taken into account.
- ●
Not all the micro features can be taken into account for describing the macro behaviour (Richardson 2004). The micro activities and their mutual interactions, the interaction of the system with the environment, influence the macro behaviour.
To fully understand a complex system, it needs to be understood it in all its complexity (Cilliers 2002). The first step is to understand the environment before understanding the system. This in itself is complex and cannot be understood fully. Reductionism is invariably applied while modelling and with time and changing scenarios the left-out aspects may become more prominent and could affect the behaviour of the system (Allen et al 2010).
This implies that the framework for the modelling of complex systems will have to be continuously revised. The models are always provisional and the modeler and the model will have to be modest and flexible about the claims they make (Cillers 2013). The claims of a model may be neither realistic nor vague (Allen 2001), may only be relative, and this responsibility has to be taken by the users of the model. Claiming that the process is well defined and scientific may not be an accountable position.