Day 35 · Climate & the Fate of Civilizations

Climate Only Sends the Bill; Institutions Decide Who Pays

August 1, 2026 (Saturday) · BigCat's Time Machine
The same drought scatters one society and leaves another standing. So the thing to study is not the temperature curve but the moment of impact: who is asked to absorb it, who has an exit, which income stream breaks first. Four scenes, four different answers.
EVENT · 01

The Sahara Closes, and Egypt Forms Along the RiverThe Green Sahara Closes · c. 5300–3100 BCE

c. 5300–3100 BCEEastern Sahara · Nile ValleyKuper & Kröpelin

During the Holocene warm period the African monsoon belt sat further north, and today's desert was savanna and lakes. Nabta Playa in southern Egypt held seasonal cattle-herding camps from about 7500 BCE, and around 4800 BCE acquired aligned stone slabs and cattle burials. Rudolph Kuper and Stefan Kröpelin (Science, 2006) used more than a hundred radiocarbon dates to map human settlement retreating southward in step with the rain belt.

From roughly 5300 BCE the eastern Sahara dried fast, and herders moved south into Sudan or east into the Nile valley. The valley was not the preferred address: the floodplain was marsh, full of hippos and crocodiles, and living there meant first building dykes, draining land and organizing labour. But once people were squeezed into a narrow corridor, settlement density and extractable surplus rose together, and around 3100 BCE Narmer unified Upper and Lower Egypt. Fekri Hassan put it most economically: the Nile did not pull people in, the desert pushed them.

Had the rain belt retreated two thousand years later, the population pressure would not have concentrated at the same moment, and the early state would not have needed such strong hydraulic and redistributive functions — centralization was forced out by density, not selected by ideas. The dispute is over tempo: deMenocal reads Atlantic cores as evidence of an abrupt shift, Kröpelin reads Lake Yoa sediments in Chad as a gradual three-thousand-year drying. The gradual case is the more troubling one — slow change raises no alarm, so people simply edge toward the river, generation by generation.

Sahel rainfall variability and the runaway growth of Lagos and Kinshasa. Climate migration rarely looks like a refugee column; it is decades of household-level decisions accumulating.

Slow disasters do not generate alarms. They generate migration.
In your own system, which "only 3% worse each year" decay is being read as normal variation?
EVENT · 02

The Norse in Greenland Did Not Freeze to DeathNorse Greenland · 985–c. 1450

985–c. 1450Eastern & Western SettlementsErik the Red · McGovern

In 985 Erik the Red led a fleet west from Iceland and founded the Eastern and Western Settlements in southwest Greenland, peaking at perhaps two to three thousand people. They lived by European-style stock-raising plus seal hunting, and their hard currency was walrus ivory — traded for the iron and timber the island lacked. Around 1250 the North Atlantic cooled, sea ice increased and the sailing season shortened. When the Dano-Norwegian missionary Hans Egede arrived in 1721, he found only ruins.

Jared Diamond's account in Collapse (2005) is the best known: they clung to a European identity, kept cattle and refused fish, and would not learn Inuit ice hunting or the kayak. Later evidence overturned it — Jette Arneborg and Thomas McGovern measured carbon isotopes in human bone and found that marine sources made up 50–80% of the late population's diet. They had switched to seal long before the end. Andrew Dugmore and colleagues (PNAS, 2012) offer a combined explanation: crusader routes brought African elephant ivory back into Europe, walrus ivory collapsed in price, and the colony's only export income vanished; the Black Death then removed buyers; sea ice merely made resupply less certain. The Western Settlement was empty by about 1360 and the Eastern by about 1450, with no archaeological layer of warfare or mass starvation — it reads more like departure by boat, in batches.

Had walrus ivory held its value, Greenland might well have survived into the modern era: cold was not the cause of death, lost income was. The argument is not settled — Diamond's version retains defenders, and in four centuries the Norse genuinely never built a kayak. But the isotope evidence rules out climate as the single variable; what remains in dispute is how to weight cultural rigidity against the market shock.

The fragility of a single-export economy when demand substitutes away: output is fine, the buyer simply found another source. Lithium, rare earths, and any supplier with one dominant customer sit on the same curve.

What climate crushes is usually not productive capacity but the income structure.
Among the revenues you depend on, which one's substitute is getting cheaper somewhere else?
EVENT · 03

The Chongzhen Emperor's Two Ledgers: Post Stations Cut, Surtaxes AddedThe Ming Collapse · North China, 1628–1644

1628–1644North China · Shaanxi · BeijingChongzhen · Liu Mao · Li Zicheng

Zhu Kezhen's 1972 reconstruction of five thousand years of Chinese climate, built from phenological records and local gazetteers, identified the seventeenth century as one of the coldest troughs in that span. Geoffrey Parker's Global Crisis (2013) folds it into the "general crisis of the seventeenth century" — the same decades brought the English Civil War, the Fronde and Ottoman upheaval. North China suffered consecutive droughts from 1637 to 1643, with drought and locusts arriving together in Henan in 1640.

The treasury was already failing when Chongzhen took the throne. In 1629 the censor Liu Mao proposed abolishing the imperial post-station network, saving roughly 680,000 taels of silver a year; the proposal passed, and Li Zicheng, a post courier at Yinchuan in Shaanxi, was among those laid off (the details are disputed in the sources, but the mass layoffs are not). Meanwhile, to fund the Liaodong front and suppress rebellion, three successive surtaxes were levied, totalling some 20 million taels by 1639 — and their weight fell precisely on the driest north. The 680,000 taels were saved at the price of a cohort of unemployed young men who knew the roads and the courier network intimately. On 25 April 1644 Li Zicheng entered Beijing and Chongzhen hanged himself on Coal Hill.

The counterfactual has to be kept tight: keeping the post stations would not have saved the Ming, whose fiscal gap was structural. The real fork is the spatial distribution of the surtaxes — had they been apportioned by harvest, loading more onto an undamaged Jiangnan, mobilization in the north would have been far more expensive. The debate is about weights: Parker and Timothy Brook (The Troubled Empire, 2010) weight climate heavily; Ray Huang's 1587, a Year of No Significance and Li Bozhong emphasize fiscal technique and institutional rigidity — under the same cooling, Jiangnan's commercialization and rice markets absorbed the shock. The climate was uniform; the collapse was not.

Austerity cuts low-unit-cost, high-coverage frontline roles first: the accounting gain lands on one ledger, the risk exposure on another, and different people sign each.

Cost savings and risk exposure booked on separate ledgers eventually meet on a third.
Your last "obviously worthwhile" cut — whose account did the risk get booked to?
EVENT · 04

1876–1878: One El Niño, Three Kinds of ReliefOne El Niño, Three Responses · 1876–1878

1876–1878Deccan · Shanxi · Northeast BrazilLytton · Zeng Guoquan · Timothy Richard

1876–78 saw one of the strongest El Niño events on record, with the monsoon failing simultaneously across the Indian Deccan, North China and northeastern Brazil. China calls it the Dingwu Famine; Shanxi and Henan were worst hit, with deaths estimated at 9.5 to 13 million, against 5.5 to 10 million in India. One sea-surface anomaly, and mortality rates that differed several-fold between regions.

Viceroy Lytton staged the Delhi Durbar in January 1877 to proclaim Victoria Empress of India, a week of banquets. In the same months he insisted on non-interference in grain markets, opposed "excessive" relief, and imposed the Temple Wage — about a pound of grain a day on relief works, a ration whose calorie content Mike Davis compared to concentration-camp issue. Indian wheat continued to be exported through both years. In China, Shanxi's governor Zeng Guoquan and the British missionary Timothy Richard ran relief separately, Richard pressing for work-for-relief with published accounts; the Qing treasury was empty, but a gentry-merchant "righteous relief" network raising silver through Shanghai mobilized cross-provincial donations — the starting point of modern Chinese civil disaster relief.

Mike Davis's Late Victorian Holocausts (2001) charges the railway and the free market with converting local harvest failure into a nationwide price catastrophe — rails carried grain toward whoever could pay, not toward whoever was hungriest. Tirthankar Roy counters that fiscal and transport capacity were genuinely limited, and that judging "deliberate" by twentieth-century standards is too harsh. The middle ground is clear enough: the 1880 Famine Commission gave India a Famine Code, and the 1943 Bengal famine proved such rules can be suspended again for wartime reasons.

Through the 2023–24 El Niño, several countries restricted rice exports in turn. Price signals travel far faster than grain, so the shortfall usually erupts somewhere else.

Famine is rarely an absence of food. It is food having somewhere better to go.
In a crisis, does your system route scarce resources to whoever needs them most, or to whoever bids highest?

Four Shocks: What Climate Did, and What Actually Decided the Outcome

The same physical input, landing on different institutions, splits the result.
Place / Period
Climate shock
What actually decided it
E. Sahara · from 5300 BCE
Rain belt retreats, habitable zone shrinks
Hydraulic cooperation and extraction forced out by valley density
Greenland · 1250–1450
More sea ice, shorter sailing season
Walrus ivory collapses; the sole export income breaks
North China · 1637–1643
Repeated drought and locusts, coldest trough
Where the surtaxes fell — and the dismantled courier network
India / China / Brazil · 1876–78
One strong El Niño
Relief rules: laissez-faire vs work-relief and civil subscription

Going Deeper

1. Why does climate determinism always look so convincing in hindsight?
Because the climate series is continuous: near any year of collapse you can find an anomaly. The test is to look for the control group that did not collapse — seventeenth-century Jiangnan met the same cold and held; Iceland held while Greenland emptied. In complex systems the shock supplies the timing and the structure supplies the outcome — only when the same shock produces different results across different structures have you located a cause rather than a coincidence in time.
2. Why are slow variables more dangerous than sudden events?
In Holling's resilience framework, slow variables — soil, groundwater, the edge of the habitable zone — set the system's stability domain, while fast variables produce the fluctuations you actually watch. Three thousand years of Saharan drying never tripped an alarm in any single year; the Ogallala aquifer's water table and an organization's technical debt are structurally identical. There is only one remedy: give slow variables their own metric and review it on a schedule — not signalling is precisely their defining property.
3. If 1876 had today's El Niño forecasting, would the outcome have differed?
Probably not much. India's binding constraint in 1877 was not ignorance of the drought but an ideology of non-intervention and standing export contracts; a year's advance warning would only have moved the same political choice a year earlier. The same holds now: the resolution of climate models improves quickly, while allocation rules update slowly. The bottleneck was never the warning itself — it is who is authorized to release the stockpile afterwards, and who gets served first.