Do societies becoming increasingly brittle and prone to disintegrating?

Do societies becoming increasingly brittle and prone to disintegrating?

They do, according to an examination of hundreds of pre-modern governments; this trend has implications for the aging global powers of today.

It's a historical cliché that great powers rise and fall. It is a widely held belief that civilizations, governments, or communities rise and fall. Is it real, though?

We, a team of complexity scientists, archaeologists, and historians, made the decision to investigate this theory. In order to determine if the historical record reflects the aging of society, we conducted the biggest research to date. Published in the Proceedings of the National Academy of Sciences, our data indicate that states do age and eventually become more likely to end. Is there anything to learn from this for the modern era?

The states' mortality

It is difficult to define civilizations or societies, and the former sometimes has a bad reputation. Instead, we limited our research to pre-modern "states"—centralized institutions that impose laws on a specific area and populace, akin to the contemporary nation-states of the US and China.

We used two distinct datasets and a statistical method. The Greek God of Doom inspired the name of our own "mortality of states" dataset, which we called Moro, and which spans 3,000 years, from 2000 BC to AD 1800. This came from a plethora of different databases, an encyclopedia on empires, and several other sources. We also consulted the Sehat databank, which included 291 polities and is the biggest online repository of historical data vetted by historians and archaeologists worldwide.

"Survival analysis" is the term for the method we used. We collated these states' lifespans and examined how far apart they were. We may anticipate a "ageless" distribution in which the probability of a state ending at year one and year 100 is the same if there is no ageing impact. That's precisely what a prior study of forty-two empires discovered. But with our bigger sample, we saw a distinct trend. The probability of termination increased throughout the first two centuries in both datasets, then peaked at a high level after that. Our results corroborate those of a previous investigation that looked at more than 168 historical crisis episodes. In their crisis database, polities had an average lifespan of almost 201 years.

Even after ruling out dynasties, the pattern of aging persisted. Family lineages provide the foundation of dynasties, which have a shorter lifespan and often disintegrate owing to conflicts over succession or the loss of authority within the family.

Promising investigations on "critical slowing down" corroborate our results. A complex system often starts to recover from disruptions more slowly before experiencing a significant structural change, or a "tipping point". The human body ages similarly, thus injuries may take longer to heal as you become older.

For two distinct historical groups—the Pueblo communities of the southwest United States and the early farmers of Neolithic Europe—we now have evidence of such significant slowing down. Neolithic farmers first appeared between 4,000 and 8,000 years ago, spreading over present-day Turkey and Europe. Subsequently, they experienced recurring crises marked by the outbreak of strife and warfare, population declines, abandonment of grain growing, and changes in agricultural locations. Before Chicago's metal-framed skyscrapers of the 1800s, the Pueblo societies, who farmed maize, built the biggest non-earth structures in the US and Canada. The Puebloans also saw many cycles of expansion and collapse, culminating in crisis moments approximately around the years AD 700, 890, 1145, and 1285. Violence increased as population, maize, and urbanization decreased throughout each of these occurrences. These cycles required two centuries on average, which is consistent with the larger pattern we discovered. Before their collapses, populations of the Puebloans and the earliest farmers in Europe recovered more slowly from shocks like droughts.  

There are a lot of warnings to be mindful of. First, there are several forms of state terminations. They could just be the result of a warlord coup or some other change in the ruling class. Alternatively, they might represent a collapse of society similar to that which happened in Mycenaean Greece, with the permanent loss of writing, governance, colossal constructions, and population reduction. Many villages among those that did experience a complete collapse managed to survive and even thrive. These layoffs aren't always negative things. Premodern governments were rapacious and wildly unequal. One estimate put the late western Roman Empire three-quarters of the way towards the theoretical limit of wealth disparity (one person owning all the excess money).

Second, our figures are based on historical and archeological records that generally agree on start and finish dates. These are often contested. For example, did the Byzantine Empire collapse in 1453 with the fall of its capital, Constantinople, or did the Crusaders divide its lands after the Fall of Constantinople in 1204 or did the Islamic Caliphates take over a significant portion of the country in the 7th century? We used both higher and lower estimations for the start and finish of a state to aid with this.

Notwithstanding these drawbacks, this is the biggest research to date, and the results from the two large datasets were comparable. Right now, it's the most complete response we have.

Investigating what promotes social longevity and what leads to increasing susceptibility will be the following stages. States may become less resilient over time for a number of reasons. Social friction may eventually increase due to growing inequality, exploitative institutions, and conflicts amongst elites. Pollution might jeopardize the ecosystems upon which polities rely. Maybe as cities become denser, there's a greater chance of illness and conflict? Or a variety of factors may combine to induce a loss of resilience. (Read more: Are we headed for the breakdown of civilization?)

Does the contemporary world become older?

Are aging trends seen in pre-modern nations relevant to the contemporary era? We believe that they could. It is uncertain whether the current global system as a whole is susceptible to the same tendencies we found. But the globe is not immune to variables that have been suggested as antecedents to collapse in the past, such as environmental deterioration, elite competitiveness, and rising inequality.

The wealthiest 1% of people on the planet have about half of the total wealth, while the lowest half only have around 0.75%. Today's extraordinary climate change is occurring an order of magnitude quicker than the warming that led to the greatest mass extinction catastrophe in recorded world history. The world is dependent on nine major Earth systems, yet six of them have entered a high-risk area. Disagreement between the wealthiest classes has contributed to polarization and mistrust in several nations.

The globe is now globalized and hyperconnected, unlike the nations we examined, but this should not be reason for comfort. While the collapse of a single weak state would often have little influence on the rest of the globe, the unrest of a superpower like the US may have far-reaching effects. Covid-19 and the global financial crisis of 2007–2008 both demonstrate how interconnectedness may intensify shocks in times of crisis. This is seen in a lot of other intricate systems. Coral reefs and other densely linked ecosystems are better at absorbing tiny shocks but also have a tendency to supercharge and disperse large hits.

The majority of nations now vary significantly from the empires that existed centuries ago. States with more technical prowess, industrialized industry, and professional police and bureaucracy are likely to be more robust and stable. But modern technology also carries with it new vulnerabilities and risks, such nuclear weapons and the capacity for infections to spread more quickly. Additionally, we must exercise caution while applauding or supporting the establishment of corrupt or dictatorial governments. Longevity and resilience are not inherently beneficial.

However, we believe that a knowledge of long-term history may aid in avoiding previous errors, including possible causes of social aging.

* Luke Kemp is a research affiliate of the University of Cambridge's Centre for the Study of Existential Risk and a research associate of the Notre Dame Institute for Advanced Study. Penguin Random House will publish his first book, "Goliath's Curse: A Deep History of Societal Collapse and What it Means for our Future," in May 2025.  

Marten Scheffer, Egbert van Nes, Luke Kemp, Timothy Kohler, Timothy Lenton, and Chi Xu's "The vulnerability of aging states: A survival analysis across premodern societies" served as the basis for this essay.