The source code of political power

Part of a continuing series about complexity science by the Santa Fe Institute and The Christian Science Monitor, generously supported by Arizona State University.

People search for information with Google and Yahoo, but they often find it on Wikipedia, the sixth most-visited site on the Internet. Everyone brings their questions there: students use Wikipedia to cram for exams; journalists, to check their sources; scientists, to broaden their view of a field.

Famously, Wikipedia isn’t a well-planned operation. Its salaried employees are massively outnumbered by tens of thousands of “editors” who, attracted by Wikipedia’s vision or irritated by its inaccuracies, take it upon themselves to contribute. And contribute they do: researchers at the University of California, Berkeley and the University of Minnesota estimate that volunteer editors have invested more than 40 million hours of labor into Wikipedia, comparable to the effort required to build the underwater tunnel between Britain and France.

These contributing editors not only write articles, they also argue with each other. How should they describe controversial issues or track down hoaxes and errors? What should they do with ill-intentioned or chronically ill-behaved editors? How should they punctuate the movie title “Star Trek Into Darkness”?

Debates on these questions spill out well beyond any particular conflict, as editors write essays, propose policies, and build support for guiding principles – some might say rationalizations – to bring order to the chaos. What happens when they try?

More is different

Questions such as these are more than just occasions to tell stories: they’re also questions that scientists try to answer. In the 20th century, social scientists – impressed by the success of physics and chemistry – tried to provide general answers by writing down mathematical laws to describe and explain social behaviors. When they did so, they imagined a world in which everyone was the same. This assumption made calculations possible, and occasionally it even predicted what people would do.

But these tools lagged behind what astute observers already knew: when people interact with each other, unexpected, unanticipated patterns emerge. Society is not one man writ large. When enough people gather together, what they create has its own, autonomous life.

The Nobel laureate (and Santa Fe Institute affiliate) Phil Anderson made this principle a fundamental law of complex systems, succinctly expressed in the title of his famous article, “More is Different.” Dr. Anderson, a physicist, had won the Nobel Prize for the study of how crystals conduct electricity when they have imperfections. His essay went far beyond physics, however – and it inspired many of us to strive to understand the imperfect crystals of society, made of human minds.

It was in that spirit that my Indiana University colleague Bradi Heaberlin and I trawled the hidden side of Wikipedia – a human society in which every editor’s moves are documented down to the second.

We drew up nearly 2,000 pages, a dense network of interconnected policies, expectations, and social norms: accounts of how a good editor ought to behave. As it was added to and woven together over time, this “norm network” tracked the hopes and fears and struggles of tens of thousands of users since the project began in 2001.

Ecosystems of ideas

When we turned our mathematical tools on this hidden world, we didn’t find a handbook or a For Dummies guide, the product of a single mind. She and I found something far more interesting: an evolving ecosystem of ideas.

Some pages urged users to be civil or to be neutral, for example, while others, written later, tried to understand what being civil, or being neutral, really meant. Some pages were concerned with truth, others with proper formatting (in case you’re wondering: Wikipedia is formally neutral on the Oxford Comma). Some talked about the importance of being polite, but others warned about how a preoccupation with politeness can undermine excellence.

The patterns that emerged also changed over time. As Wikipedia’s network grew, halos of commentary emerged to feed off the core policies. An unexpected analogy to what we see in Wikipedia is to the Talmud, the book of Jewish law, whose basic texts underwent a similar pattern of growth through commentary – albeit over the course of centuries, not years.

In Wikipedia, commentary often signaled the authority of what it commented upon: the more commented-on a page, the more often it was accessed, edited, and talked about. At the same time, however, commentary had a second, somewhat unexpected effect: it pulled the network apart. Debates on how to be neutral, for example, tended not to connect to debates on how to be civil. We could track the emergence of distinct communities of interpretation, as if watching the emergence of different forms of law over time.

Most interesting yet, we found a strong signal of what scientists call a founder effect: in many cases, the enduring norms were first written years before Wikipedia grew to fame, and when the population was a small fraction of what it would one day become. Just as the foundational precepts of the United States still govern us today and generate much commentary in the form of Supreme Court opinions, Wikipedia is governed by rules that originated among a small minority.

None of this was planned, of course: no super-user wired these networks together, or later pulled them apart. Instead, dynamic patterns emerged from the actions of thousands of individuals, each with their own idiosyncratic beliefs, working with and against each other to define what it meant to be a good Wikipedian.

And while a small minority may have drafted the rules, it was the actions of those who came later that gave them the status they enjoy today. None of this could have been predicted by studying the motivations of individuals in isolation: the patterns we see are a fundamentally social affair. More, as Anderson said, is different.

Re-running the tape of history

What could Wikipedia have been like if that early group had chosen differently? If the group who drafted the influential policy “Neutral Point of View,” for example, had chosen a different title, or changed their minds? Would Wikipedia have failed? Would it have just been different? Or would others have stepped in to write the very same thing?

The biologist Stephen J. Gould once asked this question about life itself: what would happen if we re-ran the tape of evolution? Basic problems in how societies change and adapt, thrive, or fail to prosper depend on the answer.

It’s a question we ask about societies much larger and grander than Wikipedia. With our colleagues in political science, for example, we’re now looking at the development of the speeches in the United States Congress. And our tools allow us to compare how the United States runs itself (or doesn’t) to what happens today when new democracies emerge from authoritarian control.

Our goal is to make maps: not of the Earth, but of the landscape of ideas that people use to govern themselves. And we benefit from a new wave of digitization that allows us to look back as well as forward.

Centuries-old records from the United Kingdom, for example, have allowed us to track how the courts of London managed crime over the course of decades. We can see the emergence of new patterns of talking about crime, as courtroom participants learned to talk about violence in increasingly distinctive ways, marking a desire to manage violence differently from property crimes and disorder. In fact, in 18th century England, pick-pocketing was a capital offense, and a thief could receive as harsh a punishment as a murderer. We see the creation of new “genres” of criminal justice, new ways to talk about the crimes people commit — a pattern that continues today.

We venture into worlds that the mathematically-minded have rarely been able to go – to the other side of the English Channel, for example, where the French government put the archives of the French Revolution online. Now digitized by the libraries at Stanford University, these texts give us the chance to look in at one of the turning points of the modern world, as a small group – roughly the size of Wikipedia’s early editing population – abolish feudalism, slavery, and hereditary aristocracy; separate church and state; and draft the declaration of the rights of man.

Vive la différance

On the broadest level, we see similar patterns in the history of London, the emergence of new democracies, and the evolution of Wikipedia: toward increasing complexity, refinement, and — often — bureaucracy. Systems make distinctions and create differences, whether they be between ideas, parties, or people. And as our colleagues Tim Hitchcock and Robert Shoemaker find in their studies of London, these patterns coexist with the tactics ordinary people use, as they leverage big ideas to get on with their lives.

It’s exciting to see. But we’re wary: even the best can stumble when they look on the very largest scales. In the early 1990s, the great political scientist Francis Fukuyama cast a learned gaze over the violence of the 20th century. He declared that the millennia-long struggle for recognition – the innate human desire to be something in the eyes of others – had finally ended, with the invention of liberal democracy. One idea had triumphed. He titled his essay “The End of History.” But history didn’t end, and new ideas continue to appear. One need to look only as far as the Internet, where its offspring social media have now democratized the seeking and granting of attention. Would Professor Fukuyama have come to a different conclusion having seen the Islamic State on Twitter?

Will my colleagues and I do any better? I’m sure that in a few decades our theories will be easy to criticize as blind, even prejudiced. But one mistake I think we will avoid is the idea that social worlds evolve toward a stationary state. Whether we look inside parliament houses or web servers, we see dynamism and change: new ideas and unexpected logics of development. Some we may find silly, outdated, or even abhorrent. Others exciting, new, or perplexing.

When we study the complex patterns that emerge from human interaction, we find laws of invention, turmoil, and creation – not stability. For Americans watching the primary season, that’s not hard to believe at all.

Simon DeDeo is an external professor of the Santa Fe Institute and professor of Complex Systems and Cognitive Science at Indiana University, where he runs the Laboratory for Social Minds.