Did a volcano doom the Roman empire? (Eric Lafforgue/Art in All of Us/Corbis/Getty Images)


May 22, 2024   8 mins

In the Fifties, the science fiction author Isaac Asimov published perhaps the most optimistic vision ever of history as an exact science. In his Foundation series (recently adapted for television), he imagined a distant future in which a cadre of “psycho-historians” developed historical methods so precise that they could not only explain the past, but accurately predict political and social change centuries into the future. Could history ever become a truly scientific discipline? In the past few years, a number of thinkers have found new ways to say yes. I am not so sure. Visions of history as science go back long before Asimov: during the Enlightenment, philosophers impressed by the advances being made in natural science wondered if they could discover laws of social change equivalent to the physical laws governing such things as motion or gas pressure. Looking at the indigenous peoples of the Americas, they saw similarities to the ancient Greeks and Romans and concluded that all human societies followed the same basic path of historical evolution from “savagery” to civilised modernity (exemplified, of course, by themselves). In the 19th century, Karl Marx and Friedrich Engels claimed to have discovered a different universal pattern of historical change: “The history of all hitherto existing society is the history of class struggle.” Engels, in particular, did not hesitate to call history a science. Throughout the 20th century, both Marxist and non-Marxist social scientists developed elaborate quantitative models to explain why, for instance, revolutions broke out in some times and places but not in others. All these approaches had in common the assumption that throughout history impersonal forces greatly circumscribed the scope of human free will. As Marx wrote: “Men make their own history, but they do not make it as they please.” But in the late 20th century, social scientific model-making lost its attraction for most historians. Following the so-called “cultural turn”, the discipline increasingly looked to literary studies and cultural anthropology for inspiration and experimented with “microhistories” of single individuals or incidents, such as Carlo Ginzburg’s The Cheese and the Worms, or Natalie Zemon Davis’s The Return of Martin Guerre. History-writing of this sort, grounded in a close reading of texts, did not easily lend itself to sweeping theories of change over the centuries. Historical sociologists continued to pursue the work of model-building, but increasingly in isolation from their colleagues in history departments. More recently, the field’s turn towards “global history” did partially revive interest in the large-scale, quantitative modelling of change. Kenneth Pomeranz’s influential 2000 book The Great Divergence, for instance, developed a theory of why certain countries successfully industrialised that placed heavy emphasis on the availability of coal and other natural resources. But much of the new “global” work is deeply bound up with issues of race and has a distinctly moralising character. The authors tend to hold European imperialism and white supremacy responsible for the ills of the modern world (not without reason, of course). But moral responsibility and blame can only be imputed to those who had the ability to choose — not to those who acted as the blind instruments of impersonal historical forces. At the same time, new attempts at a scientific history have begun to appear. A field calling itself “cliodynamics”, spearheaded by the polymath biologist Peter Turchin, has attracted considerable attention, in part because in 2010 Turchin predicted that the United States was heading for massive instability exactly 10 years later. Turchin argues that history can become a “mathematised science” and explicitly compares himself to Asimov’s “psycho-historians”. Meanwhile, an increasing number of historians have been using data compiled by actual scientists about past physical changes in the world (especially climate change) to offer new explanations for political and social events such as the fall of the Roman Empire. For the moment, most historians remain either ignorant or sceptical of these new approaches. But should they once again embrace the banner of science? [su_unherd_related fttitle="More from this author " author="David A. Bell"]https://staging.unherd.com/2024/01/should-nations-have-a-right-to-conquest/[/su_unherd_related] Cliodynamics is strongly reminiscent of the social scientific approaches of the mid-20th century. Peter Turchin’s complex mathematical models of social change put great weight on what he calls “elite overproduction” as one warning sign of coming social strife. As he puts it in his new book End Times, when the “social pyramid has gotten top-heavy”, with “too many ‘elite aspirants’ competing for a fixed number of positions in the upper levels of politics and business”, civic cohesion weakens and social fracture grows. The sociologists Seymour Martin Lipset and Reinhard Bendix came to some of the same conclusions in their classic study Social Mobility in Industrial Society, published in 1959, which in turn inspired many historians of revolutions. More generally, Turchin puts far greater weight on material factors — above all, the competition for wealth and power — than on ideology, culture or sheer accident in explaining why societies undergo extreme stress. For example, he argues that the American Civil War was not “fought over slavery”, but rather “over ‘slavocracy’” — that is to say, to challenge the wealth and power of slave-owning elites. In a previous book, he and a collaborator offered a long and detailed explanation of why France fell into chaos during the Wars of Religion of the 16th century without discussing… religion. Their assumption — shared by French Marxist historians of a previous generation — is that the Reformation would not have caused such strife in the absence of underlying social pressures. To quote the Marxist Janine Estèbe, writing in 1975, “social antagonisms” were “covered by a religious cloak”. But as Natalie Zemon Davis replied to her, simply pointing to such antagonisms tells us very little about how and why people resorted to violence. “We must stretch our definition of ‘social tensions’ well beyond the issue of wealth and poverty,” Davis wrote. “Rather than being ‘covered by a religious cloak,’ the social face of the Reformation is as real as its obverse, the spiritual face, different sides of the same coin.” [su_pullquote]"The idea that human history, in all its fantastic variation, can be reduced to a science remains a chimera."[/su_pullquote] Turchin makes great claims for the mathematical rigour of cliodynamics, but in fact for his model to work he has to put the math aside and resort to some very traditional — indeed, commonplace — historical interpretations. For instance, he argues that the same combination of material factors that led to the French Wars of Religion and the American Civil War also threatened Great Britain with civic strife in the 1830s. But Britain avoided revolution. Why? Because emigration served as a safety valve; because of intelligent institutional reforms by a series of flexible governments; and because workers managed to claim new rights for themselves. These points can be found in most textbooks. In addition, like his social scientific predecessors, Turchin gives too little attention to the dynamism of historical events. The American Civil War may not have started as a war to end slavery, but at a certain point, as the fighting intensified, it became one. Why? The answer can’t be found simply by measuring the forms of social stress that led to the initial crisis. Another problem for cliodynamics is that so much historical data is simply unreliable. In the earlier book, in a chapter on medieval France, Turchin and his collaborator repeatedly acknowledged that “demographic data are hard to come by for the medieval period… the data are very crude… there are no systematic wage data… the Forez and Bar-sur-Seine data cannot be directly compared”, and so forth. Nonetheless, on the basis of a plausible but by no means certain data set, they ventured confident conclusions about the way a large decline in the population of French nobles solved the problem of “elite overproduction” and helped end a period a severe political instability. Despite the exaggerated claims, Turchin’s work can be serious and thought-provoking, and follows in a distinguished line of social scientific analysis about the origins of social crises. But it tells us much less about how such crises subsequently develop, how they can lead to such different outcomes and how they generate such wildly different ideas and ideologies. And the crises are themselves only a part of the hugely rich and complex tapestry of the human past. To reduce “history” to the mathematical modelling of social crisis would be impoverishing in the extreme. [su_unherd_related fttitle="Suggested reading" author="Adam Smith"]https://staging.unherd.com/2023/03/america-is-showing-its-age/[/su_unherd_related] If few practising historians have yet embraced the cliodynamic version of scientific history, more and more are approaching the past with techniques borrowed from the natural sciences, often in collaboration with actual scientists. A fascinating recent article in The Guardian by Jacob Mikanowski surveys this new work: examples range from genetic testing of leather to see if the ancient Scythians tanned human skin (they did), to measuring air pollution as recorded in ice cores to see when Roman silver production crashed (the third century CE), to discovering that a nun who died in 1100 had traces of expensive lapis lazuli on her teeth — probably transferred while painting an illuminated manuscript. Mikanowski might also have cited the innovative use of historical linguistics by historians of medieval Africa, who use reconstructions of past vocabulary to trace the movement of peoples across the continent. Overall, this new work falls into two broad categories. One is principally empirical and aims at simply determining such things as past temperature variations, drought conditions or the movement of peoples, animals, and commodities. Especially when the historians in question also draw on archaeological evidence, they are here following in a distinguished, long-established tradition. After all, to study the history of peoples who left no written records we have no choice but to turn to various forms of physical evidence. But as techniques have improved and evidence has accumulated, some historians have grown more ambitious — often by introducing climate change as a key driver of historical change. In perhaps the best-known recent example, the historian Kyle Harper suggested that the Roman Empire flourished and expanded in large part thanks to “a phase of warm, wet and stable climate”, only to run out of luck when the climate turned unstable, and a set of deadly infectious diseases rampaged across the Mediterranean world. Harper’s dissertation advisor, Harvard’s Michael McCormick, has offered similar arguments, and enthusiastically embraced the new marriage of history and science. The partnership, he told Mikanowski, is currently at the same stage as astronomy “when Galileo first looked up with his telescope”. [su_unherd_related fttitle="Suggested reading" author="Samuel Rubinstein"]https://staging.unherd.com/2024/02/e-p-thompson-marxist-rebel/[/su_unherd_related] The Georgetown historian Dagomar Degroot has offered a particularly engaging example of how “proxy data” collected by scientists can enrich our understanding of even well-documented parts of human history. In his book The Frigid Golden Age, he showed that while the early modern “Little Ice Age” posed massive challenges for all European societies, some adapted and thrived. The Dutch in particular demonstrated considerable resilience in devising techniques for dealing with record cold (perfecting ice skates, for instance). They also mastered the period’s new Atlantic wind patterns to shorten voyages for their commercial fleet and to devise effective naval tactics against enemy fleets. But while Degroot collaborated with scientists to ascertain past temperatures (for instance through a study of tree rings — narrower rings indicate colder winters), he took most of his evidence from traditional documentary sources. Where such sources are less plentiful, or more ambiguous, the work of interpretation becomes harder, and the ability to draw “scientific” conclusions more dubious. Historians of late antiquity, for instance, have devoted enormous attention to the year 536, which Science magazine in 2018 labelled “the worst year to be alive”. Evidence from tree rings and ice cores indicates that as a result of large-scale volcanic eruptions the world experienced the coldest summer of the past two millennia. Some written records, meanwhile, suggest that at the same time the deadly “Justinianic plague” devastated Mediterranean societies. McCormick, Harper and others speculate that this one-two punch shattered the late antique world, marking the real fall of the Roman Empire. Its surviving, eastern portion, ruled from Constantinople, moved away from the Roman heritage, embracing a fervent, zealous Christianity — in a word, turning Byzantine. But other historians have countered that the plague did not in fact have world-changing effects, and that 536 may not have been particularly disastrous. One of them, Lee Mordechai of the Hebrew University, is currently writing a book arguing that the “536 event” has served as a sort of blank screen onto which the present day projects its own fears of climate disaster. [su_pullquote]"There are more things in heaven and earth than are dreamt of in the most capacious scientific history writing."[/su_pullquote] But in any case, even as these new partnerships open up new perspectives, they can still only illuminate a relatively limited range of subjects. Climate historians define the “Little Ice Age” as having lasted from the late Middle Ages to the early 19th century: a period that includes the Renaissance, the Reformation, the European encounter with the Americas, the creation of the slave plantation, the Enlightenment, the American, French and Haitian Revolutions, Britain’s rise to global supremacy, the birth of Romanticism, the start of the Industrial Revolution, and much, much more. Climate change may have served as a contributing factor to many of these things, but it cannot serve as the principal explanatory factor for any of them. Human history is simply too gloriously, too colourfully and often too tragically complex to be reducible to single variables of this sort. Climatic variation, like “elite overproduction”, can help illuminate some parts of the canvas, but not others. Any explanation of change requires a willingness to bring together very different, and often incommensurable phenomena: social, economic, political, cultural, religious, environmental. While scientific methods can cast light on some of them, the idea that human history itself, in all its fantastic variation, can be reduced to a science remains a chimera. Even Isaac Asimov recognised that sheer chance could play an outsized role in history. In his Foundation series, the “psycho-historians” forecast the fall of a galactic empire, and for centuries events turned out exactly as they had predicted. But then came along an unexpected development: the rise of a mutant human with sinister telepathic powers known as “the Mule”, who started carving out an empire of his own. Suddenly, none of the predictions panned out. There are more things in heaven and earth than are dreamt of in the most capacious scientific history writing.

David A. Bell is a history professor at Princeton with a particular interest in the political culture of Enlightenment and revolutionary France. His latest book is Men on Horseback: The Power of Charisma in the Age of Revolution.

DavidAvromBell