War is the subject of detailed study among historians, reflecting a general hope that by learning from the past, we can avoid similar mistakes in future.
Many historians study war in terms of the actors involved and the decisions they make. It is often possible to describe how wars emerge from these stresses and to identify patterns of behavior that should be avoided in future.
But in recent years another, more powerful way to think about war has emerged. In this way of thinking, war is a simple but unavoidable network phenomenon that is hard-wired into the structure of society.
The thinking goes like this. Society is a complex web of social, political, and economic forces that depend on the network of links between individuals and the countries they represent. These links are constantly rearranging, sometimes because of violence and death. When the level of rearrangement and associated violence rises above a threshold level, we describe the resulting pattern as war.
This network science approach is providing a new way to think about how to avoid the causes of war. But it raises important questions, too. Not least of these is whether this new approach is evidence-based at all: does the historical record provide good evidence that war is a network phenomenon?
Today, we get an answer thanks to the work of Ugo Bardi at the University of Florence in Italy and a couple of colleagues, who have analyzed one of the largest historical databases of violent conflict and say its statistical properties are entirely consistent with the network theory of war. “Our result tends to support the idea that war is a statistical phenomenon related to the network structure of the human society,” they say.
Bardi and co begin with a data set compiled by Peter Breche at Georgia Tech University in Atlanta, which consists of the number of war fatalities each year between 1400 and 2000.
The analysis is straightforward. Bardi and co consider various kinds of trends over time, both in the raw data and in the data normalized to the world population. They then examine the statistical characteristics of this data.
Network phenomena generally show a clear signature: events follow a power law distribution. This kind of signature crops up in all kinds of network studies—for example, the size of websites on the internet, which connect to each other across a complex network.
Most websites are linked to by small number of other sites. But a small number of websites are linked to by a huge number of other sites. Indeed, the difference in popularity varies by many orders of magnitude. That’s a power law distribution.
The size of disease epidemics, which spread through social networks, follows a similar pattern over many orders of magnitude. The vast majority of disease occurrences are small, but a small number are huge, affecting many millions of people. Even the size of forest fires, which spread via the network of physical connections between trees, follows this power law distribution.
Bardi and co’s key finding is that the data on violent conflict clearly displays this power law signature. Most violent conflicts involve a small number of deaths, but a small number involve many millions of deaths. “War seems to follow the same statistical laws as other catastrophic phenomena, such as hurricanes, earthquakes, tsunamis, floods and landslides, whose distribution follows approximate power law,” they say.
That’s important because it allows network theorists to study war using the same mathematical tools developed for a wide range of other network phenomena. It also provides new insights into the nature and causes of war.
For example, historians often focus on the specific events that trigger war. But this new approach suggests that the trigger does not determine the eventual size of a war.
A good analogy is with forest fires. The size of these fires has little to do with the spark that starts them but instead depends on the network of connections between trees, which varies over time.
Similarly, the size of a war has little to do with the triggering incident but instead depends on the network of political, social, and economic tensions that exist at the time. These are notoriously hard to measure. That’s why claims that a war can be fought on limited terms must always be greeted with skepticism.
Bardi and co use this approach to explore the idea that humanity is becoming more peaceful and say the evidence is not persuasive on this point. “There is little evidence supporting the idea that humankind is progressing toward a more peaceful world,” they conclude. That’s because wars have become less common but at the same time more destructive.
Bardi and co’s approach is by no means unique or new. Various other researchers have begun to look at war in the same way in the last 20 years or so. However, the new work is important because it backs up earlier work by applying it to one of the largest databases of violent conflict for the first time.
This kind of work also throws into perspective the significance of the period of relative peace since the Second World War. Last year, Aaron Clauset at the University of Colorado in Boulder carried out a similar study on a smaller database of violent conflict and concluded that the current peace would have to last for over 100 years before it could be considered a trend reflecting meaningful change.
By the same token, this makes the possibility of future major conflict uncomfortably high. As Clauset put it: “The historical patterns of war seem to imply that the long peace may be substantially more fragile than proponents believe.” A sobering conclusion.
Ref: arxiv.org/abs/1812.08071 : Pattern Analysis of World Conflicts Over the Past 600 Years
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Driving companywide efficiencies with AI
Advanced AI and ML capabilities revolutionize how administrative and operations tasks are done.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.