Some months ago I posted a blog on the role of geography in history, in which I discussed the Continental Axes argument of Jared Diamond. I found it a highly productive idea – literally so, because it prompted us to collect data to test empirically a particular extension of the argument. But, as I said earlier, there are also things in Guns, Germs, and Steel with which I disagree.
An example of the latter is Diamond’s explanation of the contrast between politically fragmented Europe and perennially centralized China. First, however, it is worth pointing out that the contrast between the “united East” and “divided West” is not quite as black-and-white as it is usually portrayed. Large chunks of Europe (quite comparable in size to Chinese empires) were unified by the Roman Empire, later by the Carolingian empire, and, more fleetingly, by Napoleonic France and the Third Reich. At the other end of Eurasia, China was not always under a centralized state. There were numerous periods of disunity and fragmentation, most notably, the three centuries from the collapse of the Han Dynasty to the Sui/Tang unification. As my friend and co-author Walter Scheidel points out, two thousand years ago two quite similar political organizations, Rome and the Han dominated Western and Eastern Eurasia, respectively. It was after these empires failed, when the “First Great Divergence” came about. In Europe, periods of unification became more fleeting and periods of disunity longer, while in China, on the contrary, periods of fragmentation following recurrent imperial collapse became shorter. So there is a real difference, even though it’s not quite as stark as it is often portrayed.
What were the causes of these divergent dynamics between the Western and Eastern ends of Eurasia? Diamond proposes a geographical explanation. Here’s one version of it (from Jared Diamond’s talk, How to Get Rich, on Edge.org):
“So the real question is, why was China chronically unified, and why was Europe chronically disunified? Why is Europe disunified to this day? The answer is geography. Just picture a map of China and a map of Europe. China has a smooth coastline. Europe has an indented coastline, and each big indentation is a peninsula that became an independent country, independent ethnic group, and independent experiment in building a society: notably, the Greek peninsula, Italy, the Iberian Peninsula, Denmark, and Norway/Sweden. Europe had two big islands that became important independent societies, Britain and Ireland, while China had no island big enough to become an independent society until the modern emergence of Taiwan. Europe is transected by mountain ranges that split up Europe into different principalities: the Alps, the Pyrenees, Carpathians — China does not have mountain ranges that transect China. In Europe big rivers flow radially — the Rhine, the Rhone, the Danube, and the Elbe — and they don’t unify Europe. In China the two big rivers flow parallel to each other, are separated by low-lying land, and were quickly connected by canals. For those geographic reasons, China was unified in 221 B.C. and has stayed unified most of the time since then, whereas for geographic reasons Europe was never unified. Augustus couldn’t do it, Charlemagne couldn’t do it, and Napoleon and Hitler couldn’t unify Europe. To this day, the Europe Union is having difficulties bringing any unity to Europe.”
This argument sounds plausible and, as far as I know, did not generate much disagreement (in fact, Diamond is not the first to put it forth). However, it is wrong in two ways: (1) conceptually, because it does not get right the mechanisms underlying geographic influences on social evolution and historical dynamics and (2) empirically, because it does not reflect correctly the facts on the ground.
First, let us discuss how different features of terrain promote or impede state formation and spread. The basic rule is that anything that promotes ease of travel and communications has a positive effect on state growth and expansion, and vice versa, anything that creates barriers to movement has a negative effect. Topography, plains versus mountains is the most obvious variable, and here I am in complete agreement with Jared Diamond. Mountainous terrain provides natural defenses that make societies living there much harder to conquer. Furthermore, rugged areas interrupt easy communications; it is harder to move goods, people, and information across such terrain. So states encompassing mountainous areas had to work harder to move armies and messengers across these areas, extract resources, and to assimilate highlander population to the imperial culture.
Other geographic features that impede communications and increase defensibility are thick forests (although those can be cut) and extensive marshes (although those can be drained, as Iraq of Saddam Hussein did to suppress the rebellion of the Marsh Arabs).
But bodies of water, such as major rivers, straits, and inland seas, are not dividers, as Diamond proposes, but connectors. They may serve as moats, occasionally, but overall effect is to promote communications rather than impede them. Moving trade goods and armies is an order of magnitude cheaper by water than by land. Historians noted long ago that all major Old World civilizations were associated with major rivers (the Nile, Tigris and Euphrates, Indus, and Yellow River). Some (e.g., Karl Witfogel) thought that this association was explained by the need for irrigation, but that ideas has been rejected by modern historians, because the same pattern holds for areas where no irrigation was practiced, e.g. Eastern Europe. The major river there is the Volga, and true to the pattern, the cores of all major empires in East Europe were located on the Volga: the Khazars, the Bulghar, the Golden Horde, the Khanates of Kazan and Astrakhan, and Muscovy-Russia (the last had a core in the Volga-Oka Mesopotamia). The only exception, Kievan Russia, was based on the Dnieper River, the second major river of Eastern Europe. So this association between a major river and an imperial core has nothing to do with irrigation and everything with the ease of communications.
Inland seas have the same effect. The Mediterranean is the best example, as it was extremely important as the conduit for genes, ideas, armies, and goods. It was the road that the “Sea People” used to wreck the Bronze Age civilizations. Later the Phoenicians and the Greeks spread in the opposite direction. Roman Empire would have been impossible without the Mediterranean. The population of Rome (around a million) could be fed only by bringing grain from North Africa and Egypt by sea. Contrary to what Diamond says, the peninsulas of Europe were well connected by inland seas, and were repeatedly unified within a single state. Roman Empire, of course, unified all of them, but there were plenty of later states that incorporate two or more of these peninsulas. Byzantium briefly reconquered most of the Mediterranean, but even during its weakest periods it had a foot in both Europe (the Balkans) and Asia (Anatolia). The Ottoman Empire later replaced the Byzantium and unified most of the Mediterranean from Algeria to the Balkans. So the Balkan and Anatolian peninsulas were within one state for most of their history. The Spanish also controlled Italy for many centuries, France conquered Algeria, etc. etc. I can multiply examples, but the point is that internal seas are no impediment to imperial expansion. If you look at the map of 20th century Europe, Diamond’s argument makes sense. But in fact the neat division with each peninsula being controlled by a separate country is, historically speaking, unusual.
What about the mountains? I agree on the divisive role of mountain ranges, but is Europe really more divided by mountains than China? Let’s take a look at a map:
A topographic map of Europe (from eo.wikipedia.org)
Yes, Europe is divided by a series of mountain ranges into Northern and Mediterranean parts. But North of these ranges (Pyrenees, Alps, Carpathians) Europe is very flat. The North European Plain runs from Bordeaux in southern France (where it is narrowest) through Germany and Poland to Russia (where it becomes very broad). There are no significant barriers within it to the movement of armies and conquest. So Paris has fallen to the Russians and the Germans (on several occasions) and Moscow to the Poles, the French, and (nearly) Germans. Such conquests did not lead to a lasting unification, but the reason is not geography or, at least, not topography.
China is much more cut up by mountains:
(image from Wikipedia)
One of the most important capital cities, Xian, the capital of the first (Qin) and many other unifying states is cut off from the rest of China by mountains. In fact, the area around Xian is known as the “Land between Passes,” and some Chinese historians have argued that it served as the unifying center precisely because it is a good defensive base from which to expand. Other mountain ranges cut Sichuan and southern China off northern China. So the topography of China is much more complex than that of Europe. The eastern plain of China is indeed flat. But it was not the Yangzi valley or the lower Huang He that served as the nucleus from which China was unified. Instead China was invariably unified from the northwest (the Wei River valley where Xian is) or from the north (Beijing area).
China was also unlucky with the situation of its rivers. Whereas Europe has rivers flowing in all directions, so it is easy to travel both east-west and south-north, China is dominated by rivers flowing from west to east. So it is very difficult to move bulk goods in the North-South direction. The Chinese solved this problem with a truly remarkable piece of engineering – the Grand Canal (length = 1776 km). But it was not the Canal that made unification possible; it was political unification that made building the Canal possible.
So west-east flowing rivers and mountain ranges made southward expansion from the Wei River/Yellow River core quite problematic. An additional difficulty for conquering the South is that Yangzi River valley has a very different terrain and environment than North China. Whereas in the North cavalry reigns supreme, the South can only be conquered by a river navy. A dramatized illustration of this can be seen in one of the best Chinese movies, Red Cliff:
(image from awn.com)
This is why it took the Mongols many decades to conquer the South – they first had to learn how to fight on water.
To conclude, both the geographic facts and our conceptual understanding of how geography affects empire-building contradict the geographic explanation of why Europe and China had such divergent political history, proposed by Jared Diamond. So what is the explanation? Why is Europe not China?
Note that there is nothing particularly unique about Europe in terms of its political disunity. There are lots of world regions that had a similar history. As an example, take Southeast Asia, which was never politically unified. As Victor Lieberman describes in his magisterial two-volume work, Strange Parallels, mainland Southeast Asia typically had at least three concurrent states, roughly where present-day Myanmar, Thailand, and Vietnam are.
What is unique is China – there are no other world regions that were so consistently unified by a megaempire during the last two millennia. And what accounts for the uniqueness of China? An extended answer to this question is worth another blog (especially since this one got rather too long already). But if briefly – the interaction between Inner Asian horse archers and the Chinese. I have already discussed it here, and I plan to return to this issue in the future.
Excellent post! I agree. I was never fully convinced by the reasons J. Diamond gave to explain why China has been a unified region for so long. There must be something else. What about differences in climate? Is China less variable when it comes to rain rates and climate in general?…
I am looking forward to reading your next post. Your super-brief explanation indeed sounds right to me!… If I am not mistaken the great wall of China could be a convincing evidence for your theory.
China is actually more variable. The biomes go from the desert in the Northwest (the Gobi) to tropical wet forest in the South and everything in between. European Northern Plain, on the other hand, is temperate (ranging between warm and cold temperate), with similar rainfalls.
Peter, well, sort of, but China is centripetal: it can’t help focusing on the North China Plain-Yangzi River lowland axis. Europe is centrifugal: it can’t help focusing on the seacoast lowlands of France, Netherlands, north Germany, Italy, etc. No way in the world can Europe ever centralize or unite. No way China could ever stay divided, because anyone who got the center (the Wei River-central Yellow River axis) could roll up anybody else. The divisions were brief and usually into just north and south–not a million little states like Europe. Conversely, the Roman Empire didn’t unite Europe, it united the Mediterranean. Most of Europe wa peripheral and weakly held. So, yeah, you’re right most of the time and Diamond is simplistic as usual, but the geographic difference is real.
Gene, I agree that Chinese history has been centripetal, the question is why. It is not easy at all to unify the Yellow River and Yangzi basins. The topographies and climates are completely different. The North you conquer on horseback, the Yangzi basin has to be conquered with river fleets. The natural geographic center of China is the Yangzi. Take a look at the map. That’s where you have extensive plains unified by the network of the great river with its tributaries. Furthermore, that is where the economic center of gravity is. Highly productive paddy rice agriculture, compared to fairly pathetic millet in the North where you periodically have to deal with droughts. By all usual geographic reasons, the Chinese imperial center should have been somewhere between Wuhan and Shanghai. Except the historical record indicates that the unifier has almost always (13 out 14 times) came from the North. Social evolutionary mechanisms (warfare against the steppe horsemen) trump geographical factors, such as topography, rivers, and productive agriculture.
One thing to keep in mind is that through early Chinese history (really, until the Jin took over northern China and pushed the Song south), the vast bulk of Han Chinese population was in the north. The south may have been more fertile, but permanent immigration in China (probably due to cultural traditions) during non-invasion times has historically been slow. Plus, the soil of the Yellow River valley was more fertile back in the day when it hadn’t been tilled for several thousand years already. So it makes sense that all the early dynasties were from the north. Since the Sui/Tang, you do have the second part right, though it’s not so much warfare vs. steppe horsemen as being taken over by steppe horsemen. The Sui/Tang were (somewhat-)Sinicized Tuoba from the north (somewhat because they still had a rather non-Han tendency towards fratricide and patricide). The Yuan were Mongols. The Qing were Manchus. Of the native Han Chinese dynasties after the Tang collapsed,there was one each from the north (Song) and the south (Ming). Ironically, the Song ended up with their capital in the south and the Ming with their capital in the north.
Diamond’s work has been debunked, a bit, his magisterial visions are rhetorically captivating but more fined grained and detailed scholarship with data, is needed.
Someone like Tim Snyder of Bloodlands challenges the notion of “national narratives.” I believe his idea is that human behavior actually mainly occurs across and outside of neat national stories/boundaries/etc. Nationalism is a very recent invention and seems largely local ideological. Thus, starting with a higher order concept like a country and then, top-down, backwards reading history may not carry much information value.
In all evidence-based knowledge and claims seems best to avoid higher order concepts. Although our brains do love them — because they are cheap and easy to comprehend/talk about.
In contrast, the bottom-up history of WWII, Taste of War, talks about food energy as an organizing principle – national narratives seem trivial.
There’s a recent article available online that’s relevant both to this post and to Peter’s earlier post about farming and the rise of the state:
The authors take up the old Wittfogel idea that “hydraulic societies” heavily dependent on irrigation have a particularly intense autocratic form of government. They are especially interested in whether a history of dependence on irrigation is associated with an absence of democracy in the present.
The authors argue that simply measuring the degree of dependence on irrigation introduces a potential problem of reverse causality: maybe strong states generate larger public works, rather than vice versa. Instead, they construct an index of irrigation potential, based on availability of river water versus rain.
The authors present a map of irrigation potential in the Old World. But there’s something interesting about the map they don’t mention: areas of high irrigation potential tend to adjoin steppe and desert areas unsuitable for farming. This is completely unsurprising of course, given the way they define hydraulic potential. The tradeoff between rain-watered and irrigation agriculture is going to look very different along the Rhine than along the Nile not because the Rhine is so wimpy, but because non-irrigated agriculture is so much easier there. (I seem to remember hearing about a short-lived despotic state that included the Rhineland not so long ago, but that’s another story.)
But this means there may be another side to early state formation than the one involving state management (or at least monitoring) of irrigation agriculture that the authors mention. First, the fact that areas of high irrigation potential are surrounded by steppe/desert may favor early state formation via circumscription, a la Robert Carneiro. Second, steppe and desert are rarely simply empty. Both the earliest and the largest-scale premodern state formation may have involved managing the prickly relation along the metaethnic frontier between “the desert and the sown,” as Peter Turchin has emphasized.
Doug, thanks very much for this lead. In fact, as soon as I come back from my travels, I will get in touch with the authors – the map of irrigation potential is very interesting.
I wonder about scientific redundancy. Here is a similar article correlating rainfall and democracy (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1667332).
It is worth noticing that on more than one occasion China was united by conquerors from the North. Perhaps if the Hun, the Mongols, the Arabs, the Turks and the Russians were more successful, Europe would also be a single empire for much of its history?
[…]While preparing materials for this post I came across Peter Turchin’s note “Why Europe is Not China.” I was surprised to find that he agrees with my assessment and makes many of the same points I planned to make in here using the same historical examples I intended to use[…]
Well, no one in Europe strictly unified the written language (among other things) and promoted the concept of unity by ways of education and upward mobility (by exams) like the Chinese rulers did.
During the Middle Ages, Latin was the language of the ruling elites. The plebes all spoke their own languages, but that was the case in China as well, and in any case, they were all illiterate anyway.
Who would have wondered if Blackburn Big Band may be on our minds a lot considering that We are pretty sure that many individuals don’t even take into consideration Manchester Big Band .
Now and again I think that we should perhaps entrust Blackburn Big Band to the immature era as they can consider Manchester Big Band far easier to fully comprehend.
Whilst most people end up more and more familiarized with Manchester Big Band inside our vocab it’s really extraordinary how frequently Big Band in Bolton appears in the discussion of people someone often won t expect to be preaching about Big Band in Bolton .
I reckon this specific obsession with Blackburn Big Band is without a doubt possibly a symptom of the time I guess its also incredibly amazing just how the second another personality mentions Blackburn Big Band myriads of different individuals will be inclined to run with it and also start off debating Manchester Big Band spontaneously which actually is in fact truly weird when you actually contemplate on it.
it could be superb assuming there are many more people who were also serious about the Managers BIG Band and we could definitely have an exceptional talk in relation to Blackburn Big Band . It’s really unfortunate if people today often don’t adopt their personal passions on the grounds that they presume many other folks would typically not become intrigued.
I accept needless to say that many individuals might possibly get bored when speaking of Manchester Big Band But obviously couldn’t it be the same for whatever topic in addition to the Managers BIG Band .
I admit it is a cultural difference, not a geographical or geo-economic difference, but I believe the alphabet made a critical difference. Both Europe and China are naturally linguistically diverse. However, Europe built on these to develop a profusion of vernacular literatures. Chinese ideographs simply do not permit this. The nation-states the coalesced in Europe coalesced around bureaucracies, which required record keeping. In China, this had to be written Chinese; no alternative existed. In Europe, it was an almost trivial exercise to turn letters used for Latin (in Western Europe) or Greek (in Slavonic areas0 into written vernacular form.
When China divided, it is remarkable that nothing like a We or Cantonese bureaucratized separate state ever emerged. In Europe, there was an explosion of them. A critical example of the impact occurred in at the beginning of the Ming dynasty, when Zhang He’s fleet, the most advanced technologically in the world, came to the Indian Ocean and in short time completely re-ordered it. Had the Chinese stayed, they could easily have taken the next step and gone around to Portugal and Europe.
Instead the successor emperor ordered the fleet dismantled and the accounts destroyed, and further exploration ceased. Had a European monarch tried this, it would have been completely futile: if the French king forbade these (wealth-making) trips, the English or Dutch or Spanish governments would have not followed suit. But the Chinese emperor’s word stucik, and Portuguese arrived a few decades later.
Yes, of course the geography of the two continents varied. But I think the alphabet was a factor.