It’s an old, old dream. Four hundred years ago, Newton discovered mathematical laws of motion that precisely described how the physical world behaved. Those laws were so accurate, they also allowed us to predict the future behavior of world.
Given that predictive power, it wasn’t long before others began searching for the “laws of motions” for human beings in aggregate. If nature could be predicted with math, then maybe society and its future history could be too. The stakes on the endeavor are high. If there are precise laws of history, then we might come to understand how to avoid the mess and waste of our conflicts. On the other hand, discovering those laws might provide an irresistible tool of control, locking us into a new and terrible form of tyranny.
These questions are no longer just abstractions because as the third decade of the third millennium approaches, we may have turned a corner in our search for a mathematics of history.
“Psychohistory” is the name Isaac Asimov gave to his fictional version of Newton’s Laws of society. In his famous Foundation Trilogy, Asimov made the discovery of psychohistory and its ability to predict the behavior of large, and therefore statistically relevant, numbers of people central to a story playing out over tens of thousands of years. Psychohistory made an impression on a lot of people, including Nobel Prize-winning economist Paul Krugman, who saw his discipline as a potential contender for the ultimate theory.
It also made a huge impression on me, and that’s why my mind was so fully blown when I first began learning about what is called network theory. Finally, it seemed, we might have a path forward.
Complicated, complex networks
Networks are everywhere. They are composed of any set of basic “actors” and the relations between them. Your brain is network of linked neurons. Your cells are a network of proteins and the links between them. Your life is a network of friends, family, co-workers, and the links between them. By their very nature, networks are not just complicated, but complex.
It’s an important difference.
Look at a “map” of a social network of, say, the relationships within an average high school (who has relationships with whom), and you’ll see what looks like a bunch of overlapping spiderwebs. That’s the complicated part. But the complex part comes because, taken as a whole, every network has the capacity to self-organize into a remarkable variety of internal patterns.
The spontaneous formation of cliques and hubs with different levels of dense connections is a feature of many networks, and these profoundly affect the way information flows around the network. These patterns are said to be “emergent,” which means you could not have anticipated them from a study of just the individual actors. It’s only once they come together in the network that these new possibilities become, well, possible.
When we add time, i.e. evolution, into the mix, we find that networks are not just complex but are also adaptive. They respond to changes in their external or internal environments by changing the architecture of the links. Think about how Yahoo! was once a dominant player on the World Wide Web, but has now shrunk in comparison with Google. The evolution of networks makes them an example of what scientists now call complex adaptive systems, which now define one of the most exciting frontiers of physics (and all science). The self-organizing emergent dynamics of complex adaptive systems also makes them our best hope yet for getting to laws of history.
What is human society if not a network of actors and the links? There are many layers to these networks. There are networks of trade between nations. There are networks of energy distribution and networks of transportation. Winding through them all are social networks between individuals. Network theory, as an example of complex adaptive systems, promises to unveil the general and generic laws for how these networks form, behave and evolve. On its own, this a remarkable advance. But network science is now strongly coupled with the advance of data science, where researchers have trillions of pieces of information about all aspects of human interaction to work with. Being able to look deeply into the networks of human civilization—that’s what holds the true promise of a science of society, a true psychohistory.
We are most definitely on the cusp of something. Whether or not we find predictive laws of history is still up for grabs. How those laws, or partial laws, will be used is also up for grabs. Either way, we should all be paying attention to network science.
The future is still open. For now.