5 Lessons from Systems Thinking
Without a change of perspective, we might never become aware of the new properties that emerge in the higher level networks in which we participate, yet these might be crucial to promote stability in the system.
Understanding the risks and benefits from living in an interdependent world can be daunting. In recent decades, new fields such as systems thinking, complexity theory, or ecological modeling have tried to tackle this challenge. In economics, agent based models inspired from physics show how apparently benign low level interactions can lead to rapid collapse at the macro level. Ecological models measure the impact of environmental changes on different species. Network analysis predicts how phenomenons, from economic shocks to fake news, can propagate in a network and who to target to stop the spread of risk. Systems biology led to drug discovery by explaining how functional regulation emerges from complex low level molecular interactions. These fields all share a similar agenda: making sense of low level interactions and how they affect the system as a whole. What lessons can we draw from these emerging fields?
Networks are formed when mutually beneficial connections establish between pairs of individuals, often without any understanding of the immense superstructure in which the connections sit. Given the range of domains in our lives which are part of a globally interconnected system, here are five ideas we can take away from systems thinking.
- Phenomenon spread faster in highly interdependent networks
As connections multiply in a network, the speed at which a phenomenon can spread increases, sometimes dramatically. Imagine you went to a big party and found a lost sweatshirt. You ask each one of your friends if they lost a sweatshirt. Your friends in turn ask their friends whether someone lost anything, etc. Clearly, the more friends you have, and the more friends your friends have, the faster you will find the person who lost the sweatshirt. This is also true for the spread of economic shocks, ideas, or diseases. This is why the “stay home!” message promoted during COVID-19 is so important. Only through drastically reducing the number of contacts we have with each other will we collectively be able to slow down the spread of the disease to a point where it becomes manageable for health systems.
Whether fast speed is good or bad depends on the context. We may want scientific discoveries, good news, or positive shocks to spread as fast as possible in a system. However, we may want diseases, fake news, or economic crises to spread as slowly as possible such that they remain manageable. The crucial point is that we should at least be able to deploy solutions as fast as problems arise. For this ambitious goal to be met, we need to better anticipate the spread of risk and cooperate more closely to develop timely responses.
2. Network formation incentivizes homogeneity
True connection often requires speaking the same language. And the more widespread a language becomes, the higher are the incentives to adopt that language. This is why many individuals worldwide are learning English or Mandarin, and why every two weeks a language dies. Networks push towards homogeneity. This is true about internet protocols, sanitary standards, media formats, etc. Although beneficial at the individual level, this unifying force comes with a cost: the more similar we all become, the less insurance against risk we collectively hold. Insurance only works if individuals are affected by uncorrelated shocks. Otherwise, when hard times hit, no one is there to lend a helping hand. These concerns have already been raised in agricultural practices where the number of different crops being grown has diminished dramatically in recent years. With the loss in diversity, the global food system becomes increasingly susceptible to pests or diseases.
Promoting diversity and redundancy at all scales can help us diminish the risks of global collapse. If one system fails, another one can take over. Resilient systems, from the one controlling the 100,000 daily beats of our heart to fuel management in spaceship all rely on this principle.
3. Systems become more resilient through frequent small shocks
Unlike many physical structures, dynamic self-organizing networks get stronger with frequent small challenges. If you are exposed to small germs, especially during childhood, it will boost your immune response throughout adulthood. Because systems can adapt, small shocks will lead to a reorganization of the system that will make it more resilient. As a result of individuals growing up in sanitized environments, there is a rise in allergies and autoimmune conditions around the world. Similarly, the use of pesticides and antibiotics in farming practices is leading to one of the world’s biggest challenges. Instead of allowing plants and animals to be exposed to smaller shocks, we are creating potential huge shocks such as antibiotic resistant pathogens. The absence of small challenges can lead to rapid degradation. After a short time spent in microgravity, the lack of micro stress on the skeleton leads to osteoporosis, affecting the astronaut’s ability to move and walk upon return to Earth’s gravity.
Often, we build systems that are permeable to small shocks, which may seem beneficial in the short term, but leaves them vulnerable to bigger shocks. Allowing for continuous reorganization through small shocks is beneficial in the long run.
4. The shape of a network matters a lot to the individual
The Medici became the most influential family during the Italian Renaissance. Yet the Medici initially were not as rich as the Strozzi and did not have the same prestige as other aristocratic families. What allowed them to gain so much power? A study of the Florentine marriage network in 1430 shows that the Medici had ties to families who didn’t know each other, making them an indispensable intermediary in many business and political transactions. Their central position in the network allowed the Medici to progressively gain influence and eventually become the most powerful family of their time. This example illustrates the importance of the shape of a network. Networks can be evenly distributed, with every node having similar numbers of ties, or very centralized, one node being connected to all others.
Depending on the structure of a network, who has influence may differ a lot. Centralized networks (sometimes referred to as hub and spoke), although they are often more efficient, allow only for a phenomenon to spread in one direction. Distributed networks on the other end give a similar weight to each node. A well structured company will allow for smooth transfer of information and feedback loops, while a highly hierarchical company with many silos will give power to only a few individuals.
5. New properties emerge from low level interactions
The study of how low level interactions lead to higher level network properties has spurred the development of the field of emergence. There are countless illustrations of emerging phenomena in nature. Individual neurons link together to produce the emergent state of consciousness. In business, a company’s profile and brand is an emergent property of the actions of its individual employees. The key, remarkable property of emergence is that very simple low level interactions or structures can give rise to enormously complex emergent phenomena. The sandpile simulation shows how, simply by sprinkling grains at random on a surface, a huge complexity of pattern can emerge.
Without a change of perspective, we might never become aware of the new properties that emerge in the higher level networks in which we participate, yet an understanding of these are crucial in promoting stability of the overall system.