Numbers illuminate the shape of history.
That’s a core belief of the Stanford archaeologist and historian Ian Morris, and his book Why the West Rules—for Now puts it into practice by quantifying the development of different societies.
But he’s hardly the only one. A growing crop of scholars share Morris’s science-steeped, big-picture approach to the past. And while his book relies on simple mathematics, others go further. They aspire to capture history in equations. And they want to use math to predict how the future might play out, given different assumptions about factors like population growth or climate change.
“There are people much weirder than me out there,” says Morris.
Among the most prominent of these is Peter Turchin, a professor of ecology and evolutionary biology at the University of Connecticut, who once published a famous paper about lemming outbreaks.
A midlife crisis shifted Turchin’s research agenda from rodents to empires. His focus had been the population cycles of organisms. But he came to believe that the major questions in his field, population dynamics, had been answered. He itched for a new challenge, and found one in applying math to the past.
“Instead of divorcing my wife and marrying a graduate student,” Turchin says, “I divorced my biology and married history.”
A different kind of history. In 2008, Turchin issued a manifesto in Nature that called for turning history into an “analytical” and “predictive” science.
He argued against “trying to reform the historical profession.” Instead, he christened a new discipline: “cliodynamics.” The name comes from Clio, the mythological muse of history, and “dynamics,” the study of temporally varying processes. The fledgling research area has generated growing attention and its own publication, Cliodynamics: The Journal of Theoretical and Mathematical History.
As Turchin sees it, historians generally neglect the scientific method. The example he likes to use is the Roman Empire. More than 200 explanations have been proposed to explain its collapse, he says, and new ones keep coming.
Turchin contrasts that with science. Biology had Lamarckism and Darwinism, two different theories of evolution. Biologists “did experiments, collected data, and basically rejected Lamarckism in favor of Darwinism,” he says. “In natural sciences, theories get rejected. But in history, that hasn’t been happening.”
Cliodynamics aims to change this. One belief unites the scholars who fall under its banner: Looking at the big picture reveals patterns that play out over millennia, says Morris, “without historical actors really knowing what’s going on.”
Many historians are less smitten with the big picture.
“What historians believe in is contingency,” says Edward L. Ayers, a prize-winning historian of 19th-century America. “That means that things can always be different, and that to explain things by talking about grand laws and patterns is not to explain them at all.”
He adds, “When it’s all over, you can look back and see a pattern that has been carved by this contingency. But it doesn’t mean that the pattern determined what happened within the lives of individuals.”
Turchin, however, argues that societies as varied as the Roman Empire, medieval France, and Han China share common dynamics.
To test their theories, Turchin and his colleagues are trying to code “basically the whole world,” building a historical database of social-cultural evolution going back to the Neolithic Era. Input ranges from standard indicators, like population, to more unusual ones, like what kind of literature a society has.
Much of Turchin’s work focuses on how preindustrial societies experienced cyclical waves of political instability, like urban riots or civil wars. He has sought to explain these cycles with a theory that examines the effect of population growth on social structures. In testing his theory against eight historical case studies, he wrote in Nature, he found one factor always preceded crisis: “elite overproduction,” meaning that the number of elites exceeded society’s ability to provide them positions.
He applies similar ideas to modern society. “The next decade is likely to be a period of growing instability in the United States and Western Europe,” he wrote in 2010. The indicators? Mounting public debt. Stagnating or declining wages. A rising gap between rich and poor. And an excess of young advanced-degree holders.
“We should not expand our system of higher education beyond the ability of the economy to absorb university graduates,” he warned in Nature. “An excess of young people with advanced degrees has been one of the chief causes of instability in the past.”
If all this talk of scientific history and giant databases sounds familiar, it should.
Beginning in the 1960s, says Ayers, historians had “a brief love affair with quantification.” They adopted quantitative techniques from economics, sociology, and political science to explore the roots of contemporary issues like race and social mobility.
Computers were new then, and the idea was that big assumptions about history could be tested with data.
That wave crested with the 1974 book Time on the Cross: The Economics of American Negro Slavery, which provoked a heated debate that would have serious repercussions for the future of quantitative history. Throughout the 20th century, says Ayers, the consensus of historians had been that slavery didn’t pay, and that it was therefore a social system rather than a profitable economic one. Time on the Cross, by Robert William Fogel and Stanley L. Engerman, told a much different story. Slaves were productive laborers contributing to an agricultural system that was 35 percent more efficient than family farms of the North. What’s more, the book argued, slave owners furnished their chattel with living conditions that compared well with those of free industrial workers.
Fogel and Engerman based their findings on quantitative evidence about slave sales, whippings, family life, food, medical care, housing, and more. They crunched the data with the mathematical methods of economics. They called their work “cliometrics,” and its stunning results got Fogel on the Today show. The New York Times gushed that Time on the Cross “exposed the frailty of history done without science.”
But critics were merciless. Several books assaulted the authors’ assertions, evidence, and methods. For a taste of the criticism, consider Fogel and Engerman’s claim that slaves on one plantation received 0.7 whippings per year, which might not seem that many. The figure turned out to be incorrectly low. But it also failed to capture the more important point that whipping was a tool of social control meant to affect people who saw it and heard about it, not just the immediate victim, as the historian Thomas L. Haskell wrote in 1975.
A joke began to circulate: “If a cliometrician were to write the history of the crucifixion, he would begin by counting the nails.” When Haskell surveyed the criticism in The New York Review of Books, he reached the very serious conclusion that Time on the Cross “now appears at least to be severely flawed and possibly not even worth further attention by serious scholars.”
“From the viewpoint of most historians,” says Ayers, “Fogel and Engerman had been bewitched by an economist’s vision of a smoothly functioning system in which inputs were optimized.” The irony, he adds, is that “they did change our understanding of slavery. People now believe that slavery was a profitable system.”
What happened next depends on who is telling the story. Some say quantitative history faded. Others say it wasn’t that big to begin with, and was carried on beneath different institutional roofs, like departments of economics and political science.
“The lesson that was learned from Time on the Cross in history departments was the wrong lesson,” says Stephen Haber, an economic-historian-turned-political-scientist at Stanford. Their conclusion: “We don’t have to do this quantitative stuff.”
That’s the intellectual backdrop, as science once again mates with the muse of history. So what’s the difference between the old cliometrics and the new cliodynamics?
Think of it this way: Earlier quantitative history was like electrical engineering. “The idea was you had a lot of noise, and you were hoping to find a signal in there,” explains Kenneth Pomeranz, a University of Chicago historian who has contributed to Turchin’s journal. Say you had a large set of data representing income fluctuations over time. You would take a bunch of variables, run regressions, and try to figure out which correlations seemed robust. Then you’d come up with a model: 32 percent of the variation in income is due to trade liberalization; 25 percent to factors beyond human control, like the weather affecting harvests; and so on.
The problem was, cliometrics didn’t jibe with many historians’ views of how the social world works. The world, says Pomeranz, has all sorts of feedback loops. What cliodynamics aspires to do is to be less like electrical engineering, he says, and more like meteorology: It acknowledges that phenomena interact in ways that are contingent. “But that doesn’t mean that you throw up your hands and say there’s no system here you can study,” Pomeranz says. “Meteorologists haven’t thrown up their hands.” In fact, he says, scientists in fields like evolutionary biology have come up with ways to think mathematically about things that are constantly interacting.
“After a generation in which a lot of people said, Let’s give up on quantitative analysis because it just doesn’t work for a world in which the variables can’t be neatly separated from each other, people are saying, Hey, wait a minute, look again.”
