Where A = area, c = state's resources translated into geopolitical power, r is the growth rate, h is the spatial scale of power project, a is the geopolitical pressure from the hinterland and S is average polity-wide level of collective solidarity.

"Where A = a reasonably hard, if hardly diamond-crisp, number, and c, r, h, a, and S = subjective fudge factors." Note that this is exactly the number of variables posited, by Professor von Neumann, for the elephant to wiggle his trunk.

A good general principle of computation is that if you plug fudge into an equation, you will get fudge out. Professor Turchin is hardly the sole offender in this category, but his offense is novel. So please excuse my need to give this festering equine one last kick.

The rest of the essay reminds me a bit of the search for "Nostratic" or other vaporous linguistic megafamilies. Or, for that matter, of the many other attempts that have been made to construct a "theory of history in general." I suppose I do prefer differential equations to Spenglerian high mysticism, but not by much.

The exercise is actually quite useful, I think. When you try to reason along with Stark, Turchin, and their ilk, you find yourself operating in a space so large that an almost infinite variety of spurious patterns can be recognized, and made to wiggle the elephant's trunk as it were. Moreover, the patterns you are trying to recognize are hopelessly contaminated by 20th-century presentism - what, for example, is a "religion?" Democracy? Confucianism? Stoicism? Marxism? Nazism? When we conceptualize "paganism" as a "religion," we are looking through a basically Christian frame that was never held by any actual "pagan."

You seem to realize this, and when you write:

you can probably tell that I like both general deductive models, and an attention to contingent detail

I see a great deal of justification for the latter, and very little for the former. Perhaps you could get along just as well with just the contingent detail?

A question like "whatever happened to Babylonian paganism?" is full of exciting detail. A paragraph seems short shrift for it. Or you could explain how the Allies managed to convert Nazi Germany to democracy, or how America abolished respectable segregationism, or how the PRC defeated and discredited the Tiananmen movement, or some other exciting and contemporary equivalent of the demise of Babylonian paganism.

But my guess is that the deeper you go into these contingent details, the fewer neurons you will have left for general deductive models. And perhaps this is a good thing. Surely you'll admit that general deductive models of history don't exactly have the greatest track record...

I see a great deal of justification for the latter, and very little for the former. Perhaps you could get along just as well with just the contingent detail?

theory gives you information for "free," so there's a reason for it.

Surely you'll admit that general deductive models of history don't exactly have the greatest track record...

refuting deduction via induction. clever >:-) in any case, you could say the same of medicine up until the 20th century. one must learn from the past, but not assume that the past totally constrains the future.

I haven't read the book and probably won't, and I tend to agree with Mencius. Factors that don't seem to be included are geographical defensibility, densoty of population, quality of military, and transportation and communication factors.

Applied only to Chinese polities the theory would be more testable. Some of the polities listed (Tufan, which I think means Tibet, W. Liao, and Choresm) are really quite different than the Chinese states. W. Liao and Choresm were rather loose entities composed of semi-independent parts. Choresm completely collapsed under Mongol attack after overextending itself into Afghanistan and as far as the Persian Gulf. The Choresmian empire was just a very fragile balancing act; many of its components switched voluntarily to the Mongols as effortlessly as they had submitted to Choresm in the first place.

It seems to me that most people today have the religion they do because of violence - actual or threatened - against their ancestors in the past.

When Christian missionaries were peacefully converting the Japanese - only to have their work largely rolled back - very different methods were being used to convert Latin America. The evidence suggests that, for the proselytizer, you get the best results by killing people.

K.S. Lal analyzed Indian demography from the time of Mahmud of Ghazni through to the end of the Delhi Sutanate, and concluded that Muslim campaigns in India killed around 80 million Hindus. Timur the Great's memoirs boastfully dwell on the mass slaughter he inflicted upon Hindus. With mere swords he managed to achieve casualty rates comparable to the Russian Front in World War Two. Thus was Islam brought to India.

The Abrahamic religions seem to have more of this than, say Tibetan Buddhism. Does the word "Jihad" have any translatable meaning in Buddhism?

It seems to me that most people today have the religion they do because of violence - actual or threatened - against their ancestors in the past.

the interpretation of *threatened* is key.

When Christian missionaries were peacefully converting the Japanese - only to have their work largely rolled back

this isn't correct. most of the christians were christians because missionaries convinced the daimyos, who compelled their subjects to become christian by fiat. some were converted through the conventional witness, but these were a minority. please be careful with facts, they're your friends. if you aren't sure about something, don't say it. i allude to this fact in the post above so i'm assuming you didn't read the whole thing. or read closer.

I recall that when there were riots in Java in the 1990s against the Chinese many fled to Hindu Bali. Here the proximate dynamic isn't simply reducible to civilizational gaps, after all, both the Balinese and Chinese are outsiders in the mix of the Muslim majority and so a natural empathy might arise (and a substantial number of Chinese Indonesians are Christians, even if only nominally).

I think you overplay the religion issue, in Indonesia as often as not it is race/ethnicity/nationality that matters. I think the Chinese fled to Bali, same reason they fled to Singapore, because it was safe. Bali is a tourist place, they know which side their bread is buttered on, they want to keep it safe. But go back to 1965/66, before the time of mass tourism, and it was a bloodbath there, just like in central and east Java.

Or look at Manado, where the native people are mostly Protestant, I am led to believe there is considerable hostility to the Chinese, whether the latter go to church, temple, or both (which is rather common). I also have some Javanese Christian friends who are wont to utter the most nasty, crude anti-Chinese stuff you'll hear.

I think you overplay the religion issue, in Indonesia as often as not it is race/ethnicity/nationality that matters.

right, and race and ethnicity are what matters in myanmar and thailand, where there have been anti-chinese riots. OTOH, one could make the case that the greater integration of thailand chinese is due to a shallower cultural chasm, mostly because of religion.

I think the differering levels of intermarriage / assimilation of Desi Hindus and Muslims in Myanmar and Malaysia shows the potentials, and the limits, of a common religion overcoming a cultural gap.

Also, the writing was really good in this (extremely long) post. You resorted to boldface only three times, and there were a normal amount of paragraph breaks at appropriate places!

razib,

Let's try a reductio ad absurdum here.

What Turchin is writing about is not really the history of "peoples," but the history of governments. We tend to conflate the two and say "China" or "the Mongols" or "America," but what we are really talking about is the life and death of organizations - the Tang dynasty, the empire of Genghis, the foul beast that squats in Washington, etc, etc.

For most of history, the reason that governments rise and fall has been a matter of military competition. Their life and death, in other words, is the consequence of battles - usually quite a few, for each such lifecycle. Change the results of the battles, and all the pretty curves are different. "Contingent detail."

Therefore, a simpler problem than Turchin's is to understand the outcome of individual battles. What determines the outcome of a battle? Well, we have Na and Nb, the number of soldiers on each side; Qa and Qb, the quality of their equipment; Ma and Mb, their morale; W, the weather for the battle; g, the grade of the slope of the terrain from a to b; and so on.

Our cheerful professor can go through a hundred historic battles, maybe two hundred, maybe five, and assign all of these numbers. He can then produce a set of differential equations which predict the outcome, and calculate R-squared for it. And publish a book. And so on.

And what does this effort contribute to military history, the art of war, or any other subject? I'd like to think that we agree: the result is nothing more than a gigantic mathematical turd. But perhaps I am being too optimistic here.

Or you could use the same methods to predict the lifecycle of corporations: N, the number of employees; Q, the quality of the product line; and so on. This is actually more fun than predicting battles, because there is no futures market for battles. If Professor Turchin applies his differential equations to this arena, he'll never have to apply for a grant again. Move over, George Soros. Take a back seat, Warren Buffett. Sell all, John Paulson. The Turchmeister is on the floor.

And there's always the NBA. Or the ponies. Lucrative possibilities abound...

Mencius, your theoretical modeling of battle outcomes has already been done by the late Col. Trevor Dupuy. The one book I've read by him is the quite enjoyable "Future Wars: The World's Most Dangerous Flashpoints" which has a lot of wrong predictions, but about whether battles would occur. Some people have claimed that his prediction about the effectiveness of a Palestinian intifada has been falsified, but in his scenario it was just a sideshow to a tank invasion by Arab states. Some books by him I haven't read but might be relevant to this topic are "Numbers, Prediction and War" and "Attrition: Forecasting Battle Casualties and Equipment Losses in Modern War". I think Lanchester's N-Squared law might be the most well known of mathematical imperialism in military matters.

Awesome blogpost. An amazing amount of info and analysis.

TGGP, Onan is just unstoppable these days. (I think the Israelis even have a tank named after him.)

Hi Razib

Maybe you'd like to amend this Wikipedia entry:

http://en.wikipedia.org/wiki/Kirishitan

to make clear that the majority of Japanese who converted to Christianity were coerced into doing so.

I was rather surprised by your omission of the western hemisphere from your analysis. Perhaps you could say more about that in a future post.

Exactly why I hate history and books boasting of "truths" that collide with reality. And why
Anthropology is not a science.

The world is full of loudmouthed spinners.