Why Civilizations Collapse
The ruins that draw us — the empty streets of Pompeii, the collapsed pylons of Akhenaten's Amarna, the jungle-swallowed temples of the Classic Maya — all carry the same implicit question: what went wrong? The fascination with civilisational collapse is not morbid curiosity alone. It is the recognition that the people who built these places were, in many respects, capable and organised societies, and that their ending was not written in advance. Understanding why civilisations collapse is one of the central problems of historical scholarship, and the answers resist easy summary. Collapse is rarely a single event with a single cause. It is usually a process — sometimes gradual over a century or more, sometimes compressed into a generation — driven by the interaction of multiple stresses that a society's political, economic, and ecological systems were not designed to absorb simultaneously.
What Collapse Actually Means
Before examining the causes, it is worth questioning what 'collapse' means in this context. Most academic historians today are sceptical of the dramatic model in which a thriving civilisation suddenly ceases to exist. What actually ends, in most cases, is a particular form of political complexity: centralised states, long-distance trade networks, monumental construction programmes, literacy and record-keeping, and the hierarchical social structures that sustain all of these. The people do not disappear. The territory does not become empty. What changes is how they organise themselves, what kinds of surplus they produce, how far their exchange networks reach, and what kinds of physical structures they build.
The Late Bronze Age collapse in the eastern Mediterranean around 1200 BC is a good example. Within roughly fifty years, the palace-based states of the Aegean, Anatolia, the Levant, and Egypt either collapsed entirely (Mycenae, Hatti) or contracted dramatically (Egypt). Populations declined sharply in some areas and shifted in others. Writing systems disappeared from regions where they had been in continuous use. But people continued to farm, herd, produce pottery, and trade. The material record from the following Dark Age is simpler and poorer, but it is not absent. The collapse was of a particular, highly integrated form of Bronze Age palatial economy, not of human life in the region.
Climate and Ecological Stress
Environmental change is now among the best-supported drivers of historical collapse, thanks to advances in palaeoclimatology — the reconstruction of past climates from ice cores, lake sediments, tree rings, cave formations, and marine sequences. Extended droughts, shifts in monsoon patterns, volcanic cooling events, and sea-level changes can all place stress on agricultural systems that depend on predictable rainfall or reliable river flooding. When food production falters for multiple consecutive years, the surplus that supports non-farming specialists — soldiers, scribes, artisans, priests, palace administrators — contracts or disappears, putting the entire edifice of state organisation under pressure.
The collapse of the Akkadian Empire in Mesopotamia around 2150 BC appears to have coincided with a severe 200-year drought documented in geological records across the Near East. The Maya collapse of the Terminal Classic period (roughly 800–900 AD) in the southern lowlands correlates with a series of extended droughts identified in the Belize and Yucatán lake sediment records, including a particularly severe dry interval around 820–870 AD. The Justinianic Plague of 541–549 AD, which devastated the Byzantine Empire and contributed to its long-term weakening, was preceded by a volcanic cooling event (the Extreme Cold Period of 536–545 AD) that may have stressed agricultural systems already under pressure from decades of war.
None of these correlations is simple causation. Droughts and cooling events are common in the geological record, and most do not produce collapse. The question is always why a particular society was unable to absorb a particular stress.
Complexity, Fragility, and the Cost of Scale
Joseph Tainter's influential 1988 analysis argued that collapse is predictable when complex societies reach diminishing returns on their investments in complexity. A state that expands its military, administrative, and infrastructure systems to manage growing territories and populations achieves genuine benefits early in the process: security, economic integration, specialisation of labour, irrigation management. But as the system grows, the marginal return on each additional unit of complexity diminishes while the cost of maintaining the existing system grows. The state must extract more surplus to sustain itself, which stresses the agricultural and commercial base. When an external shock arrives — drought, invasion, epidemic — the complex system lacks the reserve capacity to absorb it and begins to unravel.
The Roman Empire's trajectory broadly fits this model. The cost of maintaining the Rhine-Danube frontier, suppressing internal revolts, and sustaining the administrative apparatus grew through the 3rd and 4th centuries AD even as the economic base of the western provinces was declining. The debasement of silver coinage, the third-century military crisis, the fragmentation of the empire, and the eventual withdrawal of Roman administration from Britain and the western provinces can all be read as consequences of a system that could no longer sustain its own complexity. The eastern half of the empire — wealthier, more urbanised, and in a more defensible geographic position — survived as Byzantium for another thousand years.
Conflict, Migration, and External Pressure
Military defeat and population movement have always been among the most visible explanations for collapse, and they are real. The fall of the Western Roman Empire was not purely internal: the Hunnic incursions from Central Asia in the 370s AD pushed Gothic, Vandal, and other Germanic peoples across the Rhine and Danube in ways that Roman frontier management could not contain. The Bronze Age collapse in the eastern Mediterranean was associated with — and may partly have been caused by — movements of the 'Sea Peoples', composite groupings of migrants and raiders who attacked Egyptian, Ugaritic, and Cypriot coastal settlements in the late 13th and early 12th centuries BC.
But military pressure alone rarely explains collapse. The Roman Empire had absorbed external military pressure for centuries before the 5th century. The Mongol devastation of the Abbasid Caliphate in 1258 (the Siege of Baghdad, the sack of the House of Wisdom) was catastrophic and terminal for a particular form of Islamic urban civilisation in Mesopotamia, but Mongol expansion alone cannot explain why the Abbasid system was vulnerable at that moment. Political fragmentation, internal sectarian conflict, and economic stress had been accumulating for a century before the Mongol advance.
Elite Overextraction and Internal Conflict
A pattern that appears across multiple collapse sequences is the overextraction of surplus by elites, leading to the impoverishment of the productive agricultural base, the alienation of regional powers, and eventually the breakdown of the central authority's legitimacy. The Classic Maya collapse provides a well-studied example: archaeological evidence from the southern lowland cities indicates that competition between ruling dynasties intensified sharply in the 8th century, fuelling constant warfare, the expansion of elite monuments at the expense of agricultural infrastructure (terracing, reservoir maintenance), and the disruption of the regional exchange networks that had sustained urban populations. The drought evidence from lake sediment cores adds environmental stress to the political dysfunction; neither alone appears sufficient.
Similar patterns — elite competition undermining the productive base, combined with environmental stress and external pressure — can be identified in the collapse of the Indus Valley Civilisation around 1900 BC, the end of the Minoan palace culture after 1450 BC, and the decline of several early Chinese dynasties. The specific mechanisms vary, but the underlying dynamic of systems stressed beyond their adaptive capacity recurs.
What Follows Collapse
The archaeological and historical record is consistent on one point: collapse is followed by reorganisation, not by permanent void. After the Bronze Age collapse, new forms of Iron Age states emerged within two to three centuries across the same territories. After the western Roman collapse, medieval kingdoms gradually reconstructed long-distance trade and literacy. After the Classic Maya collapse in the southern lowlands, the Postclassic northern cities — Chichen Itza, Uxmal, Mayapán — flourished. The Great Zimbabwe state and the Aztec Triple Alliance both rose in the aftermath of the collapse of earlier regional powers (Mapungubwe and Tula, respectively).
The ruins that we visit today are not endpoints but transitions — the physical evidence of one form of complexity giving way to another. The stone that remains in a collapsed Roman city or an abandoned Maya pyramid is the residue of a society that overreached, or was unlucky, or faced stresses it could not coordinate a response to. The specifics differ in every case. Open the map to connect the physical traces of these societies to the historical processes that brought them to an end — and to the landscapes where their successors built again.