Jared Diamond. Collapse: How societies choose to fail or succeed. 2005. London: Penguin Press. Paperback. 576 pages.
Kazakhstan is rich in natural resources – and unnatural disasters. The Aral Sea shrank by nine-tenths when Soviet planners redirected its feeder rivers to irrigation of cotton farms. The Semey region, home to hundreds of Soviet nuclear tests, above ground and below, remains economically stunted. So Jared Diamond’s book on environmental disasters should stir up plenty of local readers.
Diamond doesn’t know how to write a dull book. With respect to messy books, however, he’s coming up fast on the inside track.
Collapse is the relatively thin sequel to Guns, germs and steel, the popular classic on why societies flourish. Despite its subtitle, Collapse concerns mainly why societies decline. Diamond posits that the vital causes are environmental and social, and he seeks evidence in 10 case studies ranging from Easter Island to Montana.
The value of Collapse to you depends on what you want. If you seek facts that entertain, then step right up. Diamond’s case studies tell us little that is new, but they convey that little with flair. Of Rapa Nui (the native name for Easter Island), Diamond writes that the nearly 400 stone statues, ranging in height to 70 feet and in weight to 270 tons, had been “thrown down, many of them toppled so as to break them deliberately at the neck” (page 80). If you prefer facts that inform, then you need only to go to Diamond’s sparkling bibliographical essay. But if you want theory -- What is social collapse? Why does it happen? -- then you had better retreat to Tainter and Toynbee. Diamond says he has a conceptual framework, but he must have left it in the garage.
A Cook's tour of case studies
Diamond wrote the book to “[employ] the comparative method to understand societal collapses to which environmental problems contribute” (page 18). We cannot dissect a society in a lab, so we must regard its history as a natural experiment.
Fine. Ideally, the analysis should examine several histories that hold constant all potential determinants of collapse save the one of interest. Here’s a crude example. If we wish to know why empires fall apart, then we may compare Rome’s histories as a republic, an early empire, and a late empire. Some potential causes of social collapse happen to be constant over this period for Rome – its penchant for military imperialism, perhaps. Being constant, they cannot explain changes over time in Rome’s prospects for survival. This leaves us free to focus on a potential cause of collapse that does vary over time. If we are interested in Rome’s incentives to citizens to behave themselves (rather than to stab Gluttonus Maximus in the back), then we can look at how the weakening of such incentives over time might have left Rome more (or less) vulnerable to invaders. Did bread and circuses strengthen the loyalty of residents to Rome? Or did they attract malcontents interested only in their own stomachs?
Such questions are hard enough to answer. We need not add to the mix other societies that fell victim to other causes. The inclusion of Easter Island, which was probably doomed by exhaustion of natural resources, merely confuses matters. Even given evidence of incentives to waste island resources, we would have trouble isolating their contribution to social instability. If natural resources and incentives change over time, then disentangling their effects on Rapa Nui’s survival (without the aid of careful theory) may prove impossible, particularly since they probably were related. To relate these entangled effects to the fall of Rome as well would be an exercise in glibness.
Even with global factors under our thumb, we still must control local factors. An example is bread supply in Rome. The emperors could not always feed hungry rural migrants. Bread riots destabilize, as the recent president of Kyrgyzstan knows full well; but the melees need not arise from the emperor’s reluctance to feed the poor. A bad harvest may be at fault.
My point – that we can analyze only a simple model of the world and not the world itself – is familiar. Were I feeling more Berkeleian this morning, I could argue that the world might be just a mental construct, anyway. For now, I add merely that abstraction does not imply imprecision. Having chosen the factors of interest, we can express them numerically and probably should. To say that Roman bread riots “rarely” occurred is quantitative in nature, anyway. “Rarely” refers to a numerical interval that we shall hide from the reader. Is a bread riot every year rare? What about every month? Rather than tell the reader that the riot is rare, and pray that she trusts us, we might as well provide the rate of incidence and let her judge for herself. The writer can relegate calculations to footnotes if he has the religious conviction that readers hate numbers.
We should simplify and quantify. And that’s enough, I think. Comparative history can only satisfy given constraints.
But Diamond doesn’t satisfice. His methodology is to eschew systematic thinking. He says he will evaluate his 10 studied societies according to five causes of collapse: “Environmental damage, climate change, loss of friendly contacts [with neighboring societies], rise of hostile contacts [with other societies], and the political, economic, social and cultural setting…” (page 21).
Shopping for science
What principle underlies these five causes? Evidently, none. They comprise a shopping list. How did he compile the list? Diamond does not answer this clearly, but he may have drawn upon the very case studies that he would use later to “test” his list (page 420). In that case, his sample is biased from the get-go, and his study has no useful statistical results to impart. (These observations are moot since Diamond doesn’t extract statistics from his case studies, anyway.) Even if he has resisted the temptation to test his hypothesis first and articulate it later, his theory is still scattershot. Why pick these five determinants and not another five?
Shopping lists are not unknown in science, and Diamond might have salvaged some information by categorizing the societies according to the causes of collapse that they seemed to share. But Diamond doesn’t do this, either – although he had co-authored such a study, with Barry Rolett, about the Pacific islands. His failure to state conclusions does not stem from a lack of opportunity: One chapter discusses Haiti and the Dominican Republic, which share an island and thus may enable an analyst to hold geographical factors constant. Diamond fumbles by failing to specify the relative importance of non-geographical factors. (A similar natural experiment might compare East Berlin to West Berlin in the Soviet era. Diamond doesn’t try this one.)
Given that he wants to identify causes of social collapse, Diamond’s research tactics are strange. Presumably he wants to characterize a statistical population of all possible societies, not all of which have come into existence yet nor ever will. This population is indefinitely large. A random sample of 10 societies is probably too small to indicate what the population looks like.
Worse is that Diamond doesn’t pick his sample randomly. He deliberately selects societies so that no two illustrate the same points that he has already decided must be correct. He picks Easter Island for being “as close as we can get to a ‘pure’ ecological collapse.” Pitcairn and Henderson islands illustrate “loss of support from neighboring friendly societies”; Norse Greenland, all five determinants. Of course, “no book on societal collapses would be complete without an account…of the Maya, the most advanced Native American society and the quintessential romantic mystery of cities covered by jungle” (pages 20-21 for all quotes). As a work of science, this book is a terrific travelogue. In any event, the book does not comprise a sample of size 10 of a population – but instead 10 samples of size 1, each of a separate population. As statistical reasoning, the book is worthless.
Why does the author of Guns, germs and steel suddenly turn intellectually timid? The reasons are murky. He criticizes one-cause theories without spelling out what is wrong with them. Why group decision-making sometimes fails is “obviously a complex subject to which there would not be a single answer fitting all situations” (pages 420-421; also see pages 326-328). We can’t have a one-cause theory of group decision-making, because we obviously can’t. Generally, we can’t have a theory because we can’t have a theory. Any fool should see this. Come to think of it, Diamond is not that timid.
Diamond misunderstands the goal of theory. To understand is to categorize the objects of interest, be they societies or baseball teams. To set up a category, we must stress some attributes of the objects and ignore others. For instance, in the economic model of perfect competition, the vital trait of a farm (say) is that it is too small, relative to the corn market, to affect the price of corn. The model ignores the farmer’s educational background as irrelevant to the amount of corn that he can sell, since the latter is wholly determined by the market. Diamond shouldn’t reject environmental determinism as a theory of social collapse just because it’s simple (pages 438-439). Market competition is a simple theory, too; that’s why it has flourished for two centuries. Being simple, it's understandable.
Finally, Diamond hoists himself on his own petard. One cause is insufficient for a satisfactory theory, but five causes will do. Why is five the magic number?
Cause and defect
Diamond may even misunderstand how, for producing some outcome, a cause may be necessary and yet not sufficient. He writes: “To those who would object that there is not a necessary link between Malthusian population pressure and genocide, I would answer, ‘Of course!’ Countries can be over-populated without descending into genocide….Conversely, genocide can arise for ultimate reasons other than overpopulation…” (page 327). I presume that Diamond means to say that overpopulation is not sufficient for producing genocide; overcrowded areas don’t always erupt in murder. But overpopulation may still be necessary to genocide in the sense that under-populated areas never resort to mass murder. At least, Diamond has not provided evidence (or logic) to the contrary. He does cite, as a counter-example, the Seventies genocide in Cambodia “with only one-sixth of Rwanda’s population density” (page 327). But even given that Rwanda’s population was too dense, we cannot conclude that Cambodia’s was OK because it was only a sixth as packed. Such fine-tuning requires a theory of population density. And we already know what Diamond thinks of theories.
Sometimes Diamond criticizes theory sensibly. It is silly indeed to suppose that societies must be like individuals, passing through a cradle-to-grave cycle, merely because they are collections of individuals. But he does not seem to recognize that all good theory is metaphor. Theory simplifies and organizes. Diamond sticks stubbornly to what he thinks he knows to be true: “…Different societies collapsed to different degrees and in somewhat different ways, while many societies didn’t collapse at all” (page 7). Of what value is this vapidity? It’s better to be clearly wrong than vaguely right.
Induction and inundation
Diamond’s hostility toward abstraction creates muddle after muddle. The full title of his book would be “Societal collapse involving an environmental component, and in some cases also contributions of climate change, hostile neighbors, and trade partners, plus questions of societal responses” (page 15). Thank God that we were spared the title, if not the book.
One brief passage in the book speaks volumes. Diamond disparages this newspaper remark: “Rwandas [read: genocides] are endemic, built-in, even, to the world we inhabit.” He observes: “...That fatalistic oversimplified conclusion provokes negative reactions…for three reasons. First, any ‘explanation’ of why a genocide happened can be misconstrued as ‘excusing’ it…” (page 326). So, we should avoid simple explanations? Heavens, we wouldn’t want to make anybody mad at us.
Diamond quickly adds that we must understand tragedies to prevent their recurrence. Then what was the point of criticizing the journalist for simplicity? Well, it is “simplistic” to blame the Rwanda genocide on overpopulation. “Other factors did contribute.” These include past ethnic domination, past mass murders, and stagnant economies that force poor men to join violent groups intent on political power. Doesn’t population pressure play a role in these factors? Diamond’s determination to enumerate all forces that he sees blinds him to obvious connections among them.
Calling Reverend Malthus
Most important, Diamond’s creeping atheorism leads him to misapprehend the theory that might shed the most light on what his case studies mean – Malthus’ principle of population. The first (1798) edition of Malthus’ Essay on population, which Diamond cites, takes only a few hours to read. But Diamond gives what is effectively the Cliff’s Notes version: Exponential growth of the human population will eventually outstrip the slower, linear growth of food production, forcing a crisis that will decimate humanity – until it starts growing again, ripening for another crisis (Collapse, pages 312-313).
That answer will do fine for our Civics midterm, but a book might require a little cogitation. Malthus thought that the growth rate of the population depended mainly on human passion. Since none of us can control our lust, the probability of pregnancy is determined by biological factors that don’t change perceptibly over time. So, the growth rate of the population is a constant. Thus the amount of growth is exponential: As the population increases, the addition to it increases at an accelerating rate.
However, the amount of food produced depends only on the number of acres farmed, Malthus argues. Implicitly, he holds constant the amount of labor and capital available per acre. Assuming that all acres have the same fertility, newly farming a fixed number of acres each year will add a fixed amount to the annual harvest. The food supply thus rises in linear fashion over time.
In short, we add an increasing number of babies each year but have only a fixed amount of new food to feed them. At some point, we won’t have enough new food to fill all the new mouths. People will starve. Those who don’t will fall prey to the concomitants of famine -- disease and war. To fit this Malthusian theory to Diamond’s book, just add a modern covariate – deadly pollution. That’s all that the Club of Rome ever did.
Identifying the holes in Malthus’ theory is a favorite undergraduate exercise. The most obvious weakness is in presuming that the number of acres sown is the only determinant of food supply that can vary over time. In reality, a growing population increases the demand for food, which pushes up farm prices and profits. Yet Malthus’ farmer doesn’t respond by innovating ways to increase his yield per acre -- the use of fertilizer or machines, the conversion of a fallow acre to grazing, or the creation of hybrid crops that grow rapidly and resist insects. The late Norman Borlaug received his Nobel Peace Prize in vain. Malthus’ fatal error was to assume that self-interest would affect sexual decisions but not economic ones.
Oddly, Malthus does recognize that demographics affect technological change. A clergyman, he writes that the long-run tendency of the human population toward misery is God’s way of compelling us to invent. The prospect of starving concentrates your mind wonderfully.
Even had Diamond read Malthus, it is not evident that he would have understood the weaknesses in the clergyman’s argument, because Diamond -- for all his brilliance -- knows less about economics than a good college freshman. “Technology” refers to our knowledge of how to produce, and it is neither pro-environment nor anti-environment. In one of the most confused sections of a confused book, Diamond argues that technology must lead to environmental degradation, because technology is defined (by Diamond) as physical capital that leads to environmental degradation.
“New technologies, whether or not they succeed in solving the problem that they were designed to solve, regularly create unanticipated new problems. Technological solutions to environmental problems are routinely far more expensive than preventive measures to avoid creating the problem in the first place….All of our current problems are unintended negative consequences of our existing technology. The rapid advances in technology during the 20th century have been creating difficult new problems faster than they have been solving old problems: That’s why we’re in the situation in which we now find ourselves. What makes you think that, as of January 1, 2006, for the first time in human history, technology will miraculously stop causing new unanticipated problems while it just solves the problems that it previously produced?” (page 505)
We are privileged to have among us a scholar versed in every day of human history. But there’s more to Jared Diamond than this. “Most of all, advances in technology just increase our ability to do things, which may be either for the better or for the worse” (page 505). A naïve fool like me might have thought that increasing our abilities would expand our opportunities to battle famine and war and leave us free to reject new opportunities for violence -- but no. We are compelled always to abuse our knowledge, since technology always creates problems faster than it solves them. That must be why life expectancies have been rising for two centuries. There are things that man is not meant to know, so let’s just forget everything that we’ve learned. Then we’ll be home free, albeit a trifle damp in the home cave.
Penguin Books has leveled with us about the editorial care that it lavished on Collapse. The best-read page in any book (save those sold in Washington, D.C., where the name index takes priority) is the bionote that precedes the title page. Here, Penguin describes Diamond as the son of a “teacher/musician/linguist mother.” Why not characterize her as a “teacher, musician and linguist”? Did she periodically morph in a Druidic rite?
Although trivial in itself, the failure to correct a clunky phrase suggests that none of Diamond’s transcontinental editors – one in New York, two in London – paid him the courtesy of a critical reading. As a result, Chapter 16 – which presents Diamond’s impulsive responses to “one-line objections” to environmental protection – is a barrel full of catfish. It fails to provide even the reader holding the wrong end of a blunderbuss with an interesting target.
Guns, germs and steel improved the tone of the environmental debate; Collapse dumbs it down. I can’t wait for Diamond’s next book. – Leon Taylor
Revised on May 28, 2010
No comments:
Post a Comment