Collapse: How Societies Choose to Fail or Succeed

Name: Collapse: How Societies Choose to Fail or Succeed
Author: Jared Diamond

Jared Diamond, 2005 — risk.

The comparative autopsy suite: Easter Island, Norse Greenland, the Maya, alongside the societies that faced identical pressures and survived. Diamond’s five-factor framework — environmental damage, climate change, hostile neighbours, weakened trading partners, and the society’s response — carries one decisive asymmetry: the first four deal the hand, the fifth plays it. Collapse, in his data, is less often a resource problem than a perception problem: the societies that died could not see their own trend lines.

Why the engine keeps this on the shelf

Diamond is the shelf's counterweight to determinism: his fifth factor — response — is why the engine publishes forks rather than fates, and why the resonance index shows fifty-year endpoints side by side.

The record

Viking published Collapse in 2005 as the companion to Guns, Germs, and Steel (1997), which had won the 1998 Pulitzer Prize for General Nonfiction.
The framework is five-factor: environmental damage, climate change, hostile neighbours, loss of friendly trading partners, and the society's own response to its environmental problems.
Archaeologists answered in force with Questioning Collapse (Cambridge University Press, 2010), edited by Patricia McAnany and Norman Yoffee, which grew out of an American Anthropological Association symposium and disputed both his case selection and his 'choice' framing.
The Easter Island chapter is the most contested: Terry Hunt and Carl Lipo argue that introduced Polynesian rats drove much of the deforestation and that the population collapse followed European contact, disease and slave raiding rather than preceding it.

Marked passages

On perception failure, summarized: creeping normalcy and landscape amnesia — each generation resets its baseline to a slightly worse world, so a trend that would terrify anyone shown its endpoints side by side never alarms anyone living inside it.

The entire justification for plotting fifty-year series instead of quarterly ones. The Resonance Index exists to show the endpoints side by side.

His standing question, summarized: what was the person who cut down the last tree on Easter Island thinking as they did it — and the uncomfortable answer that they probably were not thinking about trees at all, but about that day’s entirely reasonable local incentives.

Terminal decisions rarely announce themselves. The engine watches incentive gradients, not stated intentions.

The core claims

What separates societies that survive environmental pressure from those that do not is usually the quality of the decision-making response, not the severity of the pressure.
Slow degradation escapes notice because each generation treats the degraded state it inherits as the baseline, so a severe trend is invisible to everyone living inside it.

Then and now

Landscape amnesia is measurable in the forest record: deforestation ran at 10.9 million hectares a year over 2015–2025, down from 17.6 million a year in 1990–2000 — improving, and still an Iceland-sized loss annually. Source: FAO, Global Forest Resources Assessment 2025 (published October 2025)

Baseline shift in climate terms: 2025 came in at 1.44°C above the 1850–1900 average, and the 2023–2025 three-year mean exceeded 1.5°C for the first time on record. Source: WMO consolidated analysis of eight global datasets, January 2026

The reset is visible in the ranking itself — Copernicus placed 2025 as only the third-warmest year, 0.01°C below 2023 and 0.13°C below 2024, a 'cool' year that would have been an all-time record in any earlier decade. Source: Copernicus Climate Change Service, Global Climate Highlights 2025

This text points at

TAINTER

The shelf exists because the engine reads it. See the Core, the projections, the sealed ledger, or all 81 texts.