This is the first in a series of posts that discuss Complex Adaptive System theory, research, features, and other interesting aspects. This post focuses on the basics of complex adaptive systems, how they exist in reality, and why their study is valuable.

Complex adaptive systems exist in every facet of life. From the more foundational aspects such as biological and chemical systems, to human made constructs such as the global economy and terrorist networks, and more recent technological creations such as online social networks and smart cities. Complex adaptive systems, or CAS, permeate everything. With a concept so massively ubiquitous, research and analysis of these systems is essential. However, the study of CAS is relatively recent and goes mostly hand-in-hand with the rise of computational modelling and simulation.

How is the concept of CAS so ubiquitous and widely applicable? The short definition of a CAS is “a system that contains many complex interacting entities, agents and environments, whose interactions lead to non-linear, aggregate, or self-similar properties”. This covers the “complex” and “system” part of CAS, but how are these systems adaptive? Firstly, entities may possess adaptive functions. Secondly, agents can, through aggregate processes (emergence or self-organisation), cause an environment to adapt or evolve. Finally, both environments and agents, can create aggregate processes that cause the overall system to adapt. It is important to note that these final two points, as well as the environments, distinguish CAS from classic complex systems.

The above definition can easily be applied to a wide variety of systems. As an example, we’ll apply this definition to a very popular CAS: an online social network. Social networks such as Twitter, or Facebook, contain large numbers of interacting entities, through human (and some bots) users commenting, sharing, or ‘liking’ other user’s posts. The users interact with the environment by making original posts on the service. These posts, shares, and ‘likes’ can contribute to emergent behaviours which may lead to communities forming. Studying a social network as a CAS allows us to focus on how and which aggregate properties occur, and how users’ adaptability contributes to changes in communities.

So, why do we study CAS and what are the benefits? There are two main reasons. Firstly, understanding how these systems operate under certain conditions. Secondly, determining how properties, such as aggregation or self-similarity, form and under which conditions these can occur or not. For certain real-world systems, being able to determine how and when an emergent behaviour is likely to occur can lead to creation of methods which may dissuade or encourage the entities in the system from achieving this. Translated into real-world systems, these methods may result in more efficient transportation in smart cities, detection of disreputable sources on social media, or prevention of cascading failures in distributed systems.

In the next post, I will be focusing on the history of CAS research and some key challenges that have yet to be resolved.