Member-only story
Is there a “sweet spot” of structural effectiveness?
Both in natural systems and in systems designed by us humans (which are really the same, because we are natural systems ourselves), there are trade-offs between network structures and hierarchical structures, which correspond in many respects to trade-offs between systems’ flexibility and our ability to comprehend and control them.
Many natural systems, especially ones that we describe as “living” ones, grow as network structures. Our brains are a prominent example, being extremely complicated networks of neurons and other cells. But as they grow, they also develop hierarchical structures within them, such as the layers of neural networks in the cerebral cortex and the stages of sub-networks in our visual systems that assemble primary visual inputs from our retinas into secondary and tertiary recognitions of visual patterns of shapes and colors.
In systems that we design intentionally, from computer programs, to cities, to political systems, there are advantages to network structures that can grow “organically” by making arbitrary new connections, and there are advantages to more rigid, hierarchical structures over which small numbers of people can maintain comprehensibility and control.
During my long career teaching computer programming and software design, early methods…
