News

“Complex adaptive systems are characterised by perpetual novelty” — John H. Holland

Complexity science is a logical domain for researching post-bureaucratic organisational design and development. By studying the nature of complex adaptive systems we can obtain rich insights into the nature of an organisation. Many future transmissions will refer to complex adaptive systems research as it applies to organisational development. Here is a short primer on the basics.

What are they?

Complex Adaptive Systems (CAS) are systems comprising of many interacting components (agents) that can learn or adapt.

Examples include cities, economies, civilizations, the immune system, animal species, ecosystems, and organisations. Complexity science is the study of such phenomena.

Complex adaptive systems display four common properties:

1. They are comprised of relatively simple components (agents)
2. The agents interact non-linearly (non-linearity)
3. There is no central control (autonomy)
4. They exhibit emergent behaviour (emergence).
   
   

Agents

The key to complex adaptive systems is the agent. These are the autonomous subcomponents of a system. Agents are entities that are ‘simple’ relative to the system to which they belong. For example, cells are simple relative to the complexity of the bodies in which they live. Companies are simple relative to the complexity of the economies in which they exist.

Agents do not need to be biological. In organisations, agents can refer to people, technologies and even policies.

Organisations can be referred to as agents as well, when viewed as components of a larger system, such as a supply chain or a market.

Non-linearity

Non-linearity refers to the multiple interacting variables at play in complex systems.

Imagine a population of rabbits in an ecosystem. If the only variable controlling their population is birth rate then the plot of their population over time is a straight line — a linear equation that we learnt to plot at school. But there are lots of variables that affect the population of rabbits: the availability of food and water, the presence of disease, the atmospheric temperature, the number of predators, etc. To forecast the population of rabbits over time you need to use non-linear mathematics (e.g., agent-based modelling).

Most of what happens in organisations is non-linear. There are very few simple cause-and-effect relationships.

Autonomy

Complex adaptive systems do not have any central controller. They do not have anything telling the agents what to do, how to behave. The agents interact with the system autonomously.

Consider, for example, our immune system. It operates completely autonomously, without direct control from our brains, yet exhibits highly complex behaviour. The immune system has learnt which organisms to exclude and kill, and which organisms to encourage, allow entry and support (1).

The key problem with bureaucracy is that it suppresses autonomy. Post-bureaucratic forms of an organisation should aim to increase the autonomy of its agents.

Emergence

Roughly speaking, emergence means “the action of the whole is more than the sum of the actions of the parts” (2). Emergence is distinct from determinism. In deterministic systems, similar starting conditions yield similar trajectories; in complex systems, behaviour emerges from the interaction of agents.

For example, in some primate populations, when there are particular distributions of power, policing emerges. This is where a powerful monkey stands between two combatants to impartially resolve a conflict. Policing behaviour is an emergent property of some groups. It only emerges under certain conditions (3).

In organisations, it is often futile to approach challenges deterministically. Solutions should be allowed to emerge.

Why it matters

The decisions that we make about organisations are heavily influenced by the models, metaphors and interpretive frameworks that we possess for them.

The managerial hierarchy model, that most organisations are based on, emerged in the late 19th century, early 20th century, and was derived from mechanistic paradigms (4). It is basically an ‘organisation as a machine' metaphor (5).

Longstanding research suggests managerial hierarchies function more effectively in stable conditions but face serious challenges in dynamic conditions. They work to ensure reliable execution of known tasks but inhibit complex non-routine problem-solving (6).

Complexity science provides an extensive body of research that we can use to inform new models and interpretative frameworks, and in turn utilize those to design organisations that are better suited to a VUCA world.

Posted by John Dobbin.
Post Bureaucracy 

Learn more:

Visit the Santa Fe Institute for a treasure trove of articles, courses and publications and/or subscribe to their terrific podcast.

References:

1. Scott F. Gilbert, Jan Sapp, and Alfred I. Tauber. 2012. A Symbiotic View of Life: We Have Never Been Individuals, The Quarterly Review of Biology 
   
   
2. Holland, John H. 2014. Complexity: a very short introduction. Oxford University Press
   
   
3. Flack, Jessica C., Editor: Jonathan B. Losos, Associate Editor: Peter Nonacs, Frans B. M. de Waal, and David C. Krakauer. 2005. Social Structure, Robustness, and Policing Cost in a Cognitively Sophisticated Species. The American Naturalist. 165 (5): E126-E139.
   
   
4. Liening, Andreas. 2013. The Breakdown of the Traditional Mechanistic Worldview, the Development of Complexity Sciences and the Pretence of Knowledge in Economics. Modern Economy.
   
   
5. Morgan, Gareth. 1986. Images of organization. Beverly Hills: Sage Publications.
   
   
6. Lee, Michael & Edmondson, Amy (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Research in Organizational Behavior.