Guns, Germs, and Steel revisited

Jared Diamond’s thesis, in Guns, Germs, and Steel, is that regional differences in civilizational achievement are entirely caused by biogeographical factors, while regional differences in ability have had no effect. It isn’t that he believes that there are no such regional differences: he argues that the populations with the fewest achievements are the most intelligent !

In particular, Diamond argues that people in PNG (Papua New Guinea), are significantly smarter than the average bear. “in mental ability, New Guineans are probably genetically superior [my emphasis] to Westerners”: p21. “Modern ‘Stone Age’ peoples are on the average probably more intelligent, not less intelligent, than industrialized peoples. ” p 21.

This is sufficiently odd that readers of GGS often refuse to admit that Diamond ever said it. They’ll deny that it’s even in the book. They tend to replace this meme with another of their own device: you see, hunter-gatherers are innately better at hunting and gathering – at their own way of life – than developed peoples would be. Of course that doesn’t really work either, since innate superiority at obsolete tasks ( a born buggy-whip maker?) doesn’t necessarily translate to modern superiority, or even adequacy.

I’ve only seen this claim – PNG Über Alles – in one other place, ever. A character in a book by Poul Anderson said “The only true humans on earth, my friends, the main line of evolution, the masters of the future, are the lordly Melanesians. ”
Of course that character was feigning insanity, but still.

In arguing that the last actually are first – that populations that invented calculus and gunpowder and penicillin are duller than those that invented very little – Diamond dismisses the entire field of psychometrics. He mentions no evidence, doesn’t even bother to argue about it. It’s his personal impressions of the locals in PNG versus everybody from Alfred Binet to the College Board. The word “IQ” isn’t even in the book.

It’s a ballsy approach – implying that the whole field is just pointless crap, not even worth discussing. It’s how I would deal with astrology or gender studies. It’s how everybody should have dealt with Freudian analysis.

The problem with Diamond’s non-argument is that aptitude tests actually work. A one-hour paper-and-pencil test gives a reasonable estimate of a student’s general problem-solving ability, which is why everybody uses such tests. The Army find that the top scorers make much more accurate tank gunners – it’s hard to ignore a 120-millimeter DU shell.
Regional scores on IQ tests and other educational tests ( PISA, etc) do track regional differences in S&T achievements. Not perfectly – northeast Asians have the highest scores but have not made the largest contributions to the development of modern technology – but pretty well. Populations that have low average scores on such tests have contributed very little to the development of modern science and technology.

If there was some fatal flaw in our methods of testing academic aptitude, you’d see some people (or whole populations) that scored low ( say 80) but were still whizzes at electrical engineering or molecular biology. That doesn’t happen. To be fair, we do see many people with high scores embrace various forms of madness, everything from Koreshanity (Why, this is Pellucidar, nor are we out of it.) to Fomenko’s New Chronology (Gary Kasparov). But then they’re intelligence tests, not sanity tests.

If Diamond were right (and the tests wrong), there would be tremendous opportunities for arbitrage, just as sabermetrics let baseball managers identify undervalued players. For example, if people from PNG were indeed significantly smarter than the world average, UCLA could develop powerhouse departments, full of likely future Nobelists, at low cost. People would eventually try to look intelligent by putting a bone through their nose. Why hasn’t this happened? Pure stubbornness? Shouldn’t Harvard pre-emptively adopt this policy, in order to stay on top?

If Diamond were right, hunter-gatherers and other backward peoples should be able to catch up with the developed world rather easily, being smarter. In fact, they should be able to rapidly surpass us: even a moderately higher average IQ in a population greatly increases the fraction that scores above a high threshold. PNG should be shot with genius. Yet there’s no sign of it.
Diamond acknowledges as much. “We see in our daily lives that some of the conquered peoples continue to form an underclass, centuries after the conquests or slave imports took place. ” p 25. ” Yes, the transistor, invented at Bell Labs in the eastern United States in 1947, leapt 8,000 miles to launch an electronics industry in Japan – but it did not make the shorter leap to found new industries in Zaire or Paraguay. The nations rising to new power are still ones that were incorporated thousands of years ago into the old centers of dominance based on food production, or that have been repopulated by peoples from those centers.”
“Prospects for world dominance of sub-Saharan Africans, Aboriginal Australians, and Native Americans remain dim. The hand of history’s course at 8000 B.C. lies heavily on us.” p 417.

Why should that be so? If hunter-gatherers are ” probably more intelligent, not less intelligent, than industrialized peoples”, why doesn’t it show? Maybe they aren’t plugged into the old-boy networks, but why don’t they win the math contests and chess tournaments? Where’s their Paul Morphy, their Ramanujan, their George Green? Mathematicians, at least, would cheerfully hire a grizzly bear as long as it ate fewer undergraduates than it solved Hilbert problems.

Where are the practical payoffs? “Many of the living descendants of the Aborigines who survived the era of European colonization are now finding it difficult to succeed economically in white Australian society.” p 19. Again, if they’re so smart, why aren’t they rich? Why do they flunk algebra?

Perhaps we should consider dysgenic effects. Because of low birth rates among highly educated women, IQ is probably declining today in developed countries, at ~1 pt a generation. Probably this hasn’t been going on for very long. . But if it goes on long enough, a day may come when the minds of the men of the industrialized countries fail, leaving the inhabitants of Sentinel Island the smartest people on Earth.
But it is not this day.

Enough about the thesis: it’s a mess. Measurements don’t support it, and none of its implications have gone through the formality of actually happening. Back to the book itself, which is not all bad.

You see the idea that biogeographical circumstances shaped the rise of civilization and technology is not at all crazy. The mistake is assuming that that is the only factor, or that those circumstances never change the peoples exposed to them: never change them above the neck, that is. Diamond is happy enough to admit that selection for disease resistance changed Eurasians and Africans.

Diamond emphasizes the important of domestications of animals and crops, the big step towards civilization. This allowed vast increase in population size and social complexity: you can’t overemphasize its importance.

He discusses various ways in which parts of this big story support his thesis. Often they don’t really, but the discussion can still be interesting.

Most significant domestic animals were domesticated somewhere in Eurasia or North Africa, only a couple in South America (llamas and vicuna), nothing in the rest of the world. Diamond argues that this wasn’t because populations varied in their interest in or aptitude for domestication. Instead, the explanation is that only a few large animals were suitable for domestication.

He’s unconvincing. Sure, there were places where this was true: what were the Maori in New Zealand going to domesticate – weta? And Australia didn’t have a lot of large mammals, at least not after people wiped out its megafauna. But there are plenty of large animals in Sub-Saharan Africa, yet none were domesticated. He argues that zebras were wilder, more untameable than horses – but people have tamed zebras, while the wild ancestors of horses (tarpans, which survived into the 19th century) were usually described as untameable. The wild ancestors of cows (aurochsen, which survived into the 17th century) were big and mean. They enjoyed impaling people on their horns and flinging them for distance. The eland is a large African antelope, and by Diamond’s argument it must be untameable, since the locals never tamed it. But in fact it’s rather easy to tame, and there’s now a domesticated version.

The key here is that one can select for disposition, for tameness, as well as obvious physical features, and an animal can go from totally wild to cuddly in ten generations – remember the selection experiment with Siberian foxes. In the long run disposition is not a big obstacle. Selection fixes it – selection applied to above-neck traits.

Diamond makes a similar argument about domesticating plants as crops: only a few plants were suitable for domestication, and part of the reason that some populations never developed crops was a lack of suitable plant species. I’ll give him Eskimos. but that’s about it.

Here his argument is far weaker: there are a buttload of plants that could be domesticated and might be quite useful, yet have not been. Enthusiastic agronomists keep trying to get funding for domestication of jojoba, or buffalo gourd, or guayule – usually government interest runs out well before success.
The reason that a few crops account for the great preponderance of modern agriculture is that a bird in the hand – an already-domesticated, already- optimized crop – feeds your family/makes money right now, while a potentially useful yet undomesticated crop doesn’t. One successful domestication tends to inhibit others that could flourish in the same niche. Several crops were domesticated in the eastern United States, but with the advent of maize and beans ( from Mesoamerica) most were abandoned. Maybe if those Amerindians had continued to selectively breed sumpweed for a few thousand years, it could have been a contender: but nobody is quite that stubborn.
Teosinte was an unpromising weed: it’s hard to see why anyone bothered to try to domesticate it, and it took a long time to turn it into something like modern maize. If someone had brought wheat to Mexico six thousand years ago, likely the locals would have dropped maize like a hot potato. But maize ultimately had advantages: it’s a C4 plant, while wheat is C3: maize yields can be much higher.

Why didn’t people domesticate foxes, back in the day? Is it because foxes are solitary hunters, don’t have the right pack structure and thus can’t be domesticated, blah blah blah? No: they’re easy to domesticate. But we already had dogs: what was the point? You had to be crazy like a Russian.

One other factor has tended to suppress locally-domesticated plants – what you might call alien advantage. If you grow a crop near its origin, there will be local pests and pathogens that are adapted to it. It you try growing it in a distant land with a compatible climate, it often does very much better than in its own country. So… crops from Central and South America have done very well in Africa, or sometimes in Southeast Asia. Rubber tree plantations work fine in Malaysia and Liberia but fail in Brazil. Maize is the biggest crop in Africa, while manioc and peanuts are important. Most cocoa is grown in Africa: most coffee is grown in South America.

Sometimes, Diamond was wrong, but in a perfectly reasonable way, not in the devoted service of a flawed thesis, but just because the facts weren’t all in yet. We all need to worry about that.

He considered the disastrous impact of Eurasian and African diseases on the inhabitants of the New World, contrasted with a much smaller impact in the opposite direction, and concluded that a major factor had probably been transmission from domesticated animals. Eurasians domesticated quite a few animals, Amerindians not many – perhaps that was the explanation. In Guns, Germs, and Steel (p 207), he mentions measles, tuberculosis, smallpox, influenza, pertussis (whooping cough), and falciparum malaria as likely cases of transmission from domesticated animals.

We know a lot more about this we did twenty years ago, since we’ve been sequencing the genes of everything in sight – and it appears that Diamond was mistaken about the most important members of that list. TB appears to be ancient in humans, smallpox probably came from some East African rodent, while falciparum malaria seems to have derived from a form of malaria carried by gorillas. Measles really does descend from rinderpest, a cattle plague, but then rinderpest (and mumps) probably descend from bat viruses. Domesticated animals do play a role in influenza, along with wild birds. I don’t think we know the origins of pertussis.

So why then was the Old World such a fount of infectious disease? Well, it’s bigger. Civilization was older, had had more time to pick up crowd diseases. Humans have close relatives in the Old World that carried important pathogens (chimps and gorillas), while Sasquatches are germ-free. Important pathogens, especially those with insect vectors like malaria, maybe couldn’t make it to the New World through ice-age Beringia. Transportation and trade were more advanced in the Old World, and spread disease more efficiently.

I don’t think that Diamond was making excuses for Amerindians in this, as he was when talking about domestication: having lots of plagues isn’t usually considered an accomplishment. Origination in livestock seemed like a reasonable idea at the time, considering the state of the art. It seemed so to others as well, like William McNeill. It’s not totally wrong – definitely true for measles – but it’s not a huge part of the explanation.

Sometimes Diamond was right. He says that it’s a lot easier for crops to spread east and west than north and south, and he’s correct. Middle Eastern crops worked in much of Europe, especially southern Europe, and also were important in India and China. On the other hand maize had to adapt to shorter growing seasons as it spread into North America: this took time. Post-Columbian spread of maize in Africa was much faster.
Geographical barriers were major factors in slowing the spread of civilization. Although a few distressed mariners must have occasionally crossed the Pacific in ancient times, nothing significant (in terms of crops or ideas) seems to have made it across before Columbus. Amerindians had to develop everything themselves, while populations in the Old World were sharing seeds and ideas (and plagues). Having to invent everything from scratch is a disadvantage, no question.
The geography of the Americas greatly inhibited contact between Mesoamerica and the Andean civilization: even today the Pan-American highway doesn’t go all the way through. The Sahara was even worse, but most of the budding civilizations of Eurasia did manage some contact.

Conclusion
We could use more serious work on macrohistory and the rise of civilization: it’s an interesting and important subject. In particular I’d like to see a really smart and detailed comparison of the two totally independent births of civilization in the Old and New Worlds. But this book isn’t serious. The thesis is a joke, and most of the supporting arguments are forced ( i.e. wrong). Perhaps the most important thing we can learn from Guns, Germs, and Steel is that most people are suckers, eager to sign on to ridiculous theories as long as they have the right political implications.

Advertisements