Collapse: How societies choose to fail or Succeed

Penguin, New York, 2005

Collapse: How Societies Choose to Fail or Succeed is my fourth book written for a broad public.  It was first published in 2005, a revised edition with an added chapter on Angkor appeared in 2011, and National Geographic released a documentary film based on the book.  Collapse has achieved almost as wide an international distribution as has Guns, Germs, and Steel, appearing in 31 translations, including Mongolian, Indonesian, and Bulgarian, as well as in all of the major languages of book publishing.

Content.  Collapse arose as an attempt to understand why so many past societies collapsed, leaving behind ruined or abandoned temples, pyramids, and monuments as romantic mysteries to baffle subsequent visitors and modern tourists.  Why did societies that were as powerful as the Khmer Empire, and as brilliantly creative as the Maya, abandon the sites into which they had invested such enormous effort for so many centuries?  Archaeological and paleoclimatic studies of recent decades have documented a role of environmental factors in many of these collapses.

But this subject is much more complicated than a simple story of “It got cold, and they chopped down all their trees, so they starved to death.”  One has to begin by defining how drastic a decline must be to merit being termed a collapse.  A definition requiring that all members of a society end up dead, while that may have actually happened to the Greenland Norse and the Henderson Islanders, would be too extreme, because it would exclude most of the cases that most scholars and lay people would recognize as collapses.  At the opposite extreme, most people wouldn’t consider the temporary economic downturn of the U.S. in the year 2008, or the temporary population decline of Ireland during the potato famine of 1846 – 1851, as meriting the term “collapse.”  For the purposes of my book, I adopted a common-sense working definition: “A drastic decrease in human population size and/or political/economic/social complexity, over a considerable area, for an extended time.”

The next complication is that it’s obvious from the outset that different societies may collapse for different reasons, and that there are likely to be multiple factors contributing to the collapse of any given individual society.  Even in apparently simple cases such as those of Easter Island or Henderson Island, whose societies declined or disappeared in isolation from other societies, a purely environmental explanation wouldn’t be sufficient: one also has to understand the political, economic, and social factors that prevented the society from solving its environmental problems.  Even if the main problems were environmental, one has to ask whether the environmental problems were endogenous (self-inflicted) ones such as cutting down too many trees, or exogenous (e.g., natural climate change), or a combination of endogenous and exogenous.  In addition, the great majority of human societies haven’t been totally isolated, but have been involved on a friendly or unfriendly basis with neighboring societies, which may thus cause or contribute to collapse by being unable to continue essential trade in the case of friendly neighbors, or else by conquering and destroying in the case of unfriendly neighbors.  The collapse of Carthage was surely for military reasons (defeat and destruction by Rome) rather than for environmental reasons, but it remains debated whether the subsequent collapse of the Western Roman Empire itself was similarly for military reasons (defeat and destruction by barbarians) or for ultimate reasons of Rome’s own endogenous problems.  Hence in order to analyze the collapse of any given society, I apply a five-point checklist: self-inflicted environmental problems, exogenous climate change, problems with or of neighboring friendly trade partners, problems caused by neighboring enemies, and the society’s own political, economic, and social institutions.

As still a further complication, it’s obvious that not all societies collapse quickly.  Some societies (such as those of Japan and of New Guinea) have had no documented collapses for over 10,000 years, while other societies (such as those of northern Europe) have had only rare and/or local collapses.  Why have some societies succeeded in solving problems that defeated other societies, or why have some societies faced less severe problems than have other societies?

Finally, no modern reader would be satisfied with a book that discussed only the successes and failures of past societies, and that failed to discuss how our modern societies are dealing with similar problems today.  Our motive for being fascinated by collapses in ancient societies isn’t just curiosity about romantic mysteries.  We also want to know whether past collapses hold lessons for us, that might help us deal with our own problems and avoid collapse.

Thus, my book includes four sets of studies.  Seven chapters discuss some of the clearest, most familiar, most striking examples of past collapses: the ends of Polynesian societies on Henderson and Pitcairn Islands, where everybody either did abandon the island or else ended up dead; the end of the Viking settlements on Greenland, which similarly disappeared completely;  the disappearance of Anasazi settlements in desert areas of the U.S. Southwest; the decline and abandonment of Classic Maya cities in the Southern Maya lowlands, while Maya cities survived outside those southern lowlands; and the decline of Easter Island’s Polynesian society, famous for erecting giant stone statues.  If you are inclined to object that I drew those case studies disproportionately from isolated or peripheral or poor societies, I’d respond that the Maya were among the most central and advanced Native American societies of their era; and that the peripheral location of the other societies is advantageous for the purposes of my book, by letting us examine either endogenous declines divorced from influences of neighbors (Easter Island), or else equally clear illustrations of effects of identifiable neighbors and trade partners (Henderson, Pitcairn, and the Greenland Vikings).

My next set of studies concerns three societies that have flourished for several millennia, for over 10,000 years, and for over 40,000 years: Tikopia Island, Japan, and Highland New Guinea respectively.  While all three of those regions enjoyed environmental advantages, we can also identify policies or conscious decisions that enabled the societies of all three regions to solve the environmental problems that they did encounter.

Next come five chapters examining gripping and instructive courses of events in the modern world.  Montana, seemingly the most pristine and underpopulated U.S. mainland state south of the Canadian border, actually proves to be at risk from most of the environmental problems that threaten the rest of the world.  Rwanda, the most densely populated country in Africa, suffered possibly the most ferocious convulsion in late 20th-century African history, when in 1994 six million Rwandans killed nearly one million of their fellow Rwandans and drove two million more into exile.  The Caribbean island of Hispaniola is divided between two nations, of which Haiti is the poorest and one of the most overpopulated nations of the New World, while the Dominican Republic is many times more prosperous.  Those contrasting outcomes arose to a smaller degree from environmental differences, and to a greater degree from differences of human history.  China, the world’s most populous nation and its most rapidly growing major economy, is grappling with big problems that inevitably affect the rest of the world, because China releases its gases and wastes into the same atmosphere and oceans that bathe the rest of the world as well, and because China imports essential natural products from the same overseas sources on which other countries also depend.  Australia is instructive as the First World country occupying the most fragile environment (rivaled only by Iceland), and as the First World country now contemplating the most radical solutions to its resulting environmental patterns.

Finally, my book’s last three chapters extract practical lessons for us today.  Chapter 14 asks the obvious question: how could any society fail to recognize that big problems are looming up, and why doesn’t the society take measures to alert disaster?  It was surprise  at this question that caused the archaeologist Joseph Tainter, in his 1988 book The Collapse of Complex Societies, to dismiss out of hand the possibility that complex societies could collapse as a result of depleting environmental resources.  Tainter considered it implausible that complex “societies [would] sit by and watch the encroaching weakness without taking corrective actions.”  But that is precisely what has often happened in the past, and what is happening under our eyes today.  Hence my chapter draws up a roadmap of group decision-making, starting with failure to perceive a problem in its initial stages, and ending with refusal to address the problem because of conflicts of interest and other reasons.  Chapter 15 considers the environmental policies of big businesses, many of which are viewed as, and some of which actually are, among the most environmentally destructive forces in the world today.  But other big businesses are powerful forces for environmental sanity: why do some businesses find it in their interests to protect the environment, while others don’t?  My last chapter lays out the dozen major environmental problems facing the world today, our prospects for solving those problems, and the differences between the dangers facing us and the dangers facing past societies.


Since my book’s  initial publication in 2005, information has continued to accumulate about collapses, and about avoidances of collapses.  I shall mention here three of these recent extensions, to our understanding of Easter Island, the Maya, and Angkor. 

First, I summarize below the striking new evidence that Andreas Mieth and Hans-Rudolf Bork have published concerning the widespread Polynesian felling and burning of trees on Easter Island long before European arrival.

Second, measurements of markers that reflect paleoclimate have provided increasing evidence for severe droughts that contributed to the decline of Classic Maya cities in the Southern Maya lowlands.

Finally, the biggest recent advance (summarized in a new chapter included in my book’s 2011 edition) has been in our understanding of the decline of the Khmer Empire based at Angkor, formerly the world’s most extensive city and capital of Southeast Asia’s largest and most populous empire.  While Angkor’s huge reservoirs have long been famous as tourist destinations and are among the largest structures built by humans in the pre-industrial era, difficulties in understanding how they might have functioned formerly led many scholars to view them as purely ornamental.  A joint Australian, French, and Cambodian research project has now combined aerial radar imaging with ground surveys to map the Khmer water management system in detail.  The giant reservoirs, together with innumerable smaller reservoirs of all sizes and an elaborate web of canals, constituted a complex engineered system to store, release, or dispose of rain water.  Angkor’s eventual decline was due to all five of the factors on my checklist: deforestation, greater climatic variability, hostile neighbors, new opportunities for trade, and the empire’s irreversible commitment to an enormous and hard-to-maintain system of water management that finally became overwhelmed by floods and droughts.

Interesting as the Angkor story is in its own right, it has larger implications.  Angkor exemplified a low-density city of farmland, palaces, temples, and habitations in close proximity to each other, unlike the high-density purely urban cities separate from farmland with which we are familiar today.  Such low-density cities were formerly widespread in seasonally wet tropical environments, other examples having existed in Sri Lanka, Java, Thailand, Vietnam, Myanmar, and of course in the Maya area.  Perhaps the problems that brought down the Khmer Empire also undermined those other low-density cities as well.  The Khmer decline is also useful in reminding us that, while the collapses of small isolated societies such as those of Easter Island and the Greenland Norse are simpler, more easily studied, and more instructive, collapses also afflicted the most advanced pre-industrial societies in both the Old World and the New World.

Unsolved questions. Not even for any of the most-studied cases of collapse, described in my book, do we understand what happened to people in the final phase: were there mass deaths from starvation, or did people abandon the site and resettle elsewhere, or was there just a small excess of deaths over births that produced a slow decline in population?  Visitors from Norse Greenland’s Eastern Settlement came to Western Settlement around AD 1360 and found no one in a settlement that had formerly numbered hundreds of people: did the western Norse all succumb to starvation, or were they massacred by the Inuit?  Did the last pre-H.M.S.-Bounty Polynesian inhabitants of Pitcairn Island build canoes and sail away?  On Henderson Island, which lacks trees big enough to make canoes, did a Polynesian population that probably never exceeded 50 people dwindle until there was just one last person alive, and then no one?  What happened to the millions of people who formerly inhabited the Classic cities of the Southern Maya lowlands? – There are no signs of a mass population influx into the surviving Northern Maya cities.  At the time of the Anasazi decline, pottery finds suggest that a few Anasazi moved to surviving pueblos, but not on a mass scale: what happened to everyone else?  All of those specific questions about the last phases await answers.

There are numerous other collapses, including eight cases summarized briefly on pp. 559 – 561 of my book, for which we can’t yet say which of the five factors are my checklist were most important.  Among those cases are the disappearance, around AD 1300, of Cahokia in the Mississippi Valley, the largest Native North American town before Columbus; the declines of some cities and empires in the Fertile Crescent long before the Christian era; and the end of the Moche in coastal Peru, the Tiwanaku Empire in the Andes, and the Harappan cities of the Indus Valley shortly after 2000 BC, and Great Zimbabwe in southern Africa.