This is the first in a series of short blog posts where I will discuss some of the major themes in complex systems science in an informal, and hopefully accessible, way. The perspective is my own, and might differ from others’ takes on these concepts.
From these concepts and themes I will distill implications for real-world decision-making, and rather than applying specific results to a specific problem, I will emphasize the big picture and ‘lessons learned’. So, let’s get started with a small dose of philosophy.
To get a sense of the essence of complex systems science, it is important to understand the philosophy of reductionism. Reductionism asserts that to understand something, it is sufficient to (1) break it down into its ‘natural’ parts, (2) study the properties of those parts.
Said differently, a reductionistic philosophy assumes that any property of a whole is inherited directly from a corresponding property of a part of the system.
Complex systems science recognizes the insufficiency of reductionism to understand and describe our world — especially the most interesting and relevant phenomena: living, social, civilizational, systems.
To be clear, this is not a claim the reductionist methodologies are never appropriate, just that by themselves they are insufficient.
Reductionism has become so ingrained in our way of thinking it is sometimes difficult to imagine what the alternative could be. Recall that reductionism assumes that if a system, call it S, has a property, call it P, then P must be present in at least one of the parts that compose S.
Clearly, the alternative is that property P of S is not present in any part of S. How can this be?
The answer is that P emerges from interactions of component parts of S. Let’s make this concrete with an example.
We all know that the brain is involved in recognizing patterns. For example, we see a pattern of light and know it is a face, or we might even recognize the pattern as our Grandmother. So it is safe to say ‘pattern-recognition’ is a property of the system ‘brain’.
The brain is composed of parts called neurons. In a simplified description, a neuron is either ON or OFF. When it is ON, it sends signals to other neurons it is connected to that it is ON. Each neuron, at each moment in time, adds up the signals being sent to it, and makes a simple decision whether to be ON or OFF. That is essentially it. No pattern-recognition property to be found in any neuron.
However, when these neurons are connected into networks, they are able to recognize complex patterns, such as human faces. How those networks come to be and how they function to do cognitive work can be saved for another time, just trust me here that it works.
The important point here is that the property, pattern-recognition, emerges at the scale of the system, a network of neurons (brain), without any part of the system having such a property.