Why did the very first civilizations arise in areas such as Mesopotamia, Egypt and northern China? It sounds a bit contradictory, but many historians see war as an important catalyst for civilisation. In the journal PNAS, American scientists substantiated this idea with a mathematical model. Can human history be captured in this way?
Anyone who plays the computer game Civilization knows, know that a nascent civilization is constantly engaged in warfare. In the real world too, armed conflict has been one of the major drivers of the development of civilizations. Virtually all historians recognize the importance of warfare in the development and spread of civilizations. The fiercest wars were once fought on the fringes of the great Eurasian steppes, where formidable hordes of steppe nomads attacked small farming communities.
These nomads were initially feared for their extraordinary military innovations, the most important of which is the tamed horse. Weak societies were regularly overrun by steppe people. Some farmers on the edge of the steppes managed to resist the nomads and adopted their military innovations (horses, chariots, values, etc.) and became stronger themselves. In the competition with the nomads new ways of waging war arose. Technological progress and social bonding are both important for the development of civilization. And it is precisely war that leads to those technological innovations and is an important binding agent for a society.
Warring Squares
American scientists wanted to test the importance of war in the development of civilizations. Not by digging into the ground in search of clay tablets or remains of weaponry, but with a mathematical model. That model is a bit like the game Civilization. It breaks up the map of Africa and the Eurasian landmass into squares measuring 100 by 100 kilometers. For each square, the scientists filled in data about mountains, climate and the presence of agricultural opportunities. Military technologies were initially distributed over the squares on the edges of the steppes.
The hypothesis was simple:the more intense the warfare, the more likely it is to create something known as civilization. In the model, an increase in military technology intensified warfare. This in turn leads to the emergence of 'institutions' that belong to a civilization, for example laws, justice, religion and an administrative hierarchy. Just like in Civilization squares can conquer each other over time and thus transfer and spread their own culture.
To check the model, the researchers made another model based on data from historical atlases. That shows how civilizations developed according to the historians. The researchers ran both computer models – the historical data and the mathematical model – simultaneously for the period 1500 BC. up to 1500 A.D. BC, a time when many great empires arose and then disappeared.
Big Matches
Broadly speaking, the two models showed some striking similarities. In the mathematical model, the first serious civilizations arose in Mesopotamia, northern China and Egypt. These are areas on the fringes of the steppe that struggled with steppe peoples and benefited first from innovations in military technology. Egypt begins around 1400, a period that (very roughly, the model is a few hundred years off) corresponds, according to the authors, to the invasions of the Hyksos, a horse-riding steppe people from Anatolia. Later still, in the first millennium BC, the "civilization squares" spread to Persia, India and the Mediterranean area.
The similarity between the model and the historical data is due to the influence of military innovation and intensified warfare, the researchers write. Disabling military innovations as a variable or randomly distributing military technology (ie not along the edges of the steppe) meant that the models no longer ran parallel. Disabling the influence of the landscape had a much smaller effect on the reliability of the model. The degree of warfare thus seems to be more important for the development of civilizations than geographical factors.
Too rude
The paper's lead author, Peter Turchin, is not a historian, but an evolutionary biologist. The kind of mathematical models he uses have already proven their worth in biology, for example to describe the evolution of species. He knows that his method is quite different from how historians usually work, and recognizes that the rise and fall of empires is a complex process. “But our analysis also supports the idea that there are universal mechanisms at work in history,” he writes. But is that really so? You can question whether history can be captured in this way.
Wijnand Mijnhardt, professor of comparative history of science and director of the Descartes Center for History of Science and Philosophy of Science, is not convinced. “Historians and philosophers have often tried to discover universal mechanisms behind history,” he says over the phone. “Take, for example, the work of Edward Gibbon or Oswald Spengler. But the same problem keeps cropping up.”
“At a very high level of abstraction, models are always correct, but the patterns are far too general and too coarse. Historians try to explain a specific context. In addition, such a model usually does not lead to something that we did not already know, not in this case either. Maybe they'll laugh at me in about fifty years, but for now I don't see the added value.”
Universal mechanisms?
Professor Bert Theunissen, who is involved in the history of science as a biologist and historian, is also skeptical about the usefulness of these types of models. “I think it's good that they are made and that they are experimented with. Although there are problems with it, it is always better to have something to provide insight into developments than to have nothing.”
“But it goes too far to claim that you could discover universal mechanisms behind history in this way. That is really spielerei. Our society, for example, has changed quite a bit in the last fifty years. Warfare has become far from self-evident. Let alone that it still has a function as a civilization spreader. Such models therefore have no predictive value for the future.”
“If you make such a model, you can only focus on a few things. You should omit everything that you, as a researcher, do not consider important for your model. And if you're talking about the entire history of civilization, that's quite a bit.”
Historians are not so concerned with finding laws. "Beta's simply don't know exactly how historians work, says Theunissen. "That's why it sometimes happens that they want to try to make their work a little more scientific by throwing in some more figures."