By Simon DeDeo, Research Fellow, Santa Fe Institute

This article appeared in the Santa Fe New Mexican on October 14, 2013

One of the basic principles of science is that good theories extrapolate well. The laws of gravity hold both here and on the Moon, and this universality enables us to land a spacecraft on Mars. Darwin drew the theory of evolution from studies of a remote island in the Pacific, but today we use it to explain the emergence of drug-resistant TB in a city hospital.

The theories of gravity and of evolution are two of our greatest scientific achievements. But in contrast to the universal nature of these laws, our understanding of the human world – the messy realm of newspapers and cafes, traffic jams and gossip, governments and social movements – is remarkably limited.

Take, for example, some of the most basic questions in politics. How do societies resolve conflict? How do new methods for resolving conflict emerge? A newspaper story can give us incredible detail on a particular fight – the war in Syria, say, or the political brinksmanship over Obamacare.

Yet we have very little idea of how the lessons of a previous conflict might apply in the future, and the lessons we do learn often tend to be the wrong ones: generals, it is said, always fight the last war – especially if they won it. Can science do any better?

The computer revolution has helped, drawing away the veil by allowing us to capture data on conflict and cooperation at unprecedented scales. With my colleagues at the Santa Fe Institute, and the larger network of intellectuals and scientists who come here from around the world, we can pan across centuries of recorded human experience – and zoom in, literally to the second – to see how cooperation fails and conflict unfolds.

One of the best places to observe this interplay of conflict and cooperation is in a long-running example of the promises and pitfalls of our networked world: Wikipedia, the online encyclopedia. It's a fascinating case because we have a complete record, down to the second, of how every article was created and then revised by volunteer authors and editors. By studying an article’s edit history, we can classify how and when cooperative contributions are made – and how and when cooperation falls apart as people undo each other's work in struggles over differing facts and competing points of view.

In recent work we developed a model of how cooperation and conflict balance each other within the Wikipedia system. Like many systems both on the screen and off, it is indeed a balance, rather than a utopia, because while cooperation benefits all participants, gaining the initiative in a conflict can benefit one’s own point of view – at the expense, of course, of another’s.

We were surprised to discover that the vast majority of the most-edited articles in Wikipedia – on a range of contentious topics ranging from George W. Bush to climate change and the Israel-Palestine conflict – could be described by nearly identical mathematical models. Despite the differing particulars of the arguments in play, the odds of resolution, and the way these odds changed over time, were remarkably consistent.

In a follow-up collaboration with Seth Lloyd at MIT and Drew Cabaniss of the University of North Carolina Chapel Hill, we turned our lens to a seemingly very different system: the world of ancient Greek city-states. Much like contemporary Wikipedians, the Greeks were prone to conflict: unlike Wikipedians, this conflict showed itself not in rude comments about global warming but through violent revolution. Remarkably, the patterns of their revolutions seemed to follow the same dynamical principles we observed in Wikipedia.

It’s a long way from Wikipedia to Ancient Greece, but these kinds of universal principles suggest that we may be able to capture fundamental laws of cooperation, laws that hold from the second-by-second evolution of Wikipedia to the decade-by-decade history of the Mediterranean. (Cabaniss asked me recently whether we could model the conflict driving the government shutdown with the same mathematics; we're crunching the data on our machines now.)

One of the longest timescales we work with in detail involves the Old Bailey, the main courthouse in London, England. More than two centuries of courtroom records – names, verdicts, and even transcripts, rescued from file cabinets and government libraries and then digitized – now live on our hard drives. With Sara Klingenstein here at the Institute, and our collaborator Tim Hitchcock in the United Kingdom, we have analyzed this trove of data to understand exactly how notions of “modern justice” – including the ideals of fairness and even forgiveness – emerged from what was essentially a harsh medieval world. We model these thousands of courtroom conflicts “from within” the data, trying to determine what information might have been available to the actual participants.

Our most moving result concerns the years between the American Revolution and the end of our data just before World War I. Our analyses detect – and the signal that jumps out is remarkably strong – the emergence of a new distinction, both practical and moral, between violent and non-violent crimes. In the 1770s, trials for non-violent crimes were often indistinguishable from trials for violent offenses: the ways people talked about stealing and murder, and the ways the justice system dealt with them, were similar.

Over the next hundred and fifty years, however, the English – meaning not just judges and legislators, but also witnesses, defendants and juries – changed. Imperceptibly, but systematically and irreversibly over the decades, the way they spoke about these two kinds of crimes changed. By the 1900s, we can say that a modern view of "civil society" had emerged: the system acts as though violence such as assault, rape or murder is a distinct problem, requiring different approaches, from non-violent offenses such as petty theft and fraud.

From these hundred-year timespans we can zoom in, to the beginning of our data in the 1600s, to read about "incorrigible" pickpockets condemned to death – or to the end of our data in 1913, where we can read how Emmeline Pankhurst, a suffragette "animated by a sincere desire to get political power into the hands of women,” tried to sway her jury.

I began my scientific life as a cosmologist, studying the origins of galaxies and stars; our unit of measurement was the gigaparsec, a length so unimaginably large that the universe (as so far revealed) can only fit about three of them on a side. The work I do today with my colleagues at the Santa Fe Institute provokes a similar kind of vertigo. Our Old Bailey data – like much of the large-scale data we need to understand the human world – contains the evidence of hundreds of thousands of moments of evil and goodness, failure and courage.

Martin Luther King once wrote that “the arc of the moral universe is long, but it bends towards justice.” It’s a different kind of arc from those the planets describe, but one whose bends and turns science might finally be able to study. Such work may give us one more guide as we try to understand – and change – the arc we see ourselves on today.



The Santa Fe Institute is a private, not-for-profit, independent research and education center founded in 1984 where top researchers from around the world gather to study and understand the theoretical foundations and patterns underlying the complex systems that are most critical to human society – economies, ecosystems, conflict, disease, human social institutions, and the global condition. This column is part of a series written by researchers at the Santa Fe Institute and published in The Santa Fe New Mexican.

Read the article in the Santa Fe New Mexican (October 14, 2013)