The Barbarian Reservoir

The peoples of the steppe who harassed and invaded the civilized world for more than two millenia have normally been though of as a blind force, and the nomad invasions are often described with the metaphor of suddenly-released  potential energy: geothermal, electrostatic, or thermodynamic reservoirs producing volcanoes, lightning, or storms. (Secondary metaphors are the womb and the black hole / primal chaos).

“Surrounded by this crust of civilization lie the immense wastes of Central Asia , little known, and unpredictable in their reactions – the magma, the molten core around which most of world history has been built. When it comes to the surface, when it breaks the shell within which the sedentary civilizations endeavor to contain it, man, horror-stricken, speaks of catastrophe” (Sinor, p. 93).

“A succession of waves of Turkish invaders coming from some inexhaustible reservoir of remote Asia” (Boodberg,  p. 13).

“The inner Asian reservoir of tribal invasions” (Lattimore, p. 247).

 “An ever-present reservoir of natural political and military talent”. (Gellner, “Tribe and State in the Middle East” p. 199).

“Bedouins are prior to sedentary people. The desert is the basis and reservoir of civilization and cities.” (Ibn Khaldun, p. 93).

“[Central Asia] appears as a sort of black hole in the middle of the world…. it is darkly huge and hugely dark and sucks the life space of outlying peoples and their civilizations into the black hole in the center.” (Andre Gunder Frank, pp. 1-2)1

“Central areas also find it difficult to subjugate peripheral areas of savannah or mountain, which then harbour cohesive participatory, segmentary communities endowed with great material potential.  Thus they constitute a kind of political womb, a source of new rulers who from time to time displace the old.” (Gellner, “War and Violence” p. 160).

Gellner thus follows Khaldun in taking the steppe potential as a positive, structuring force: “For Ibn Khaldun, urban life is a permanent necessity, and pastoral tribalism is the only source of state-formation, the state being the gift of the tribe to the ever-present city”. (Gellner, in Khazanov, p. xxii).

In this theory, the nomads destroy the old order and replace it with a new order, functioning rather like left-Hegelian revolutionaries: “Destruction is also a creative act” (Bakunin). Evidence for this ordering function can be found in the various nomad dynasties founded in China, Central Asia, Northern India, and the Middle East. (The Germanic and “Norman” states founded following the fall of Rome have something in common with these steppe-based states, but the Germans and Normans were sedentary and not nomadic). 

Besides replacing decadent governments in the sedentary world, nomads and other barbarians also gradually brought order to the “uncontrolled zone” of the northern frontier. By 1000 A.D., most of Western Europe was literate, civilized, and Christian, and the Norse and the “Norman” Rus had created a trade route stretching from Greenland to at least Trebizond, and when by 1280 or so the Mongols had brought China and the interior of Eurasia under unified control for the first time ever, it would have been theoretically possible for a single traveler to travel from Greenland to China and from China to Egypt by sea (and also, possibly from North China even to Alaska. Before the Mongols trade from Song China to the Baghdad Caliphate (the two wealthiest nations in the world) had had formerly to pass through four intervening taxing jurisdictions (Chin, Hsi-hsia, Qaraqitai, Khwarizm, all of which were destroyed by the Mongols), some of the time the Baghdad Mongols and the Xanadu Mongols were neighbors, and the rest of the time there was only one intermediary, the weak Chagatai Monol Khanate, which was theoretically allied to both of its Mongol neighbors).2

The various civilized metaphors for the steppe peoples I have listed (reservoir, black hole, etc.) all make it seem that the steppe is tremendously populous, when in fact we know that it was and is thinly populated – the sedentary peoples outnumbered the nomads by a factor of thirty or more. The illusion of large numbers came from the high level of military mobilization of the nomads and from the nomads’ mobility, which allowed them to gather  troops from a large area and then concentrate them suddenly at the weakest point on the sedentary line of defense.  A second factor leading to this powerful effect (and this may have been the primary reason for Chinggis’ success) was that the Mongol Empire’s armies were highly disciplined under one man’s command, with the result that their effect was controlled and carefully targeted, rather than consisting of a hodgepodge of random impulsive attacks (something which goes against our conception of the “horde” as a chaotic mass). It was the discipline of these armies which allowed them to suddenly release all their enormous potential (like the bursting of a dam, or a lightning strike) at a single point, which usually had been carefully chosen.3

Climatologically Mongolia is a permanent high-pressure zone, and this metaphor can be added to the others: Mongolia as military high-pressure zone. Herodotus and later authors often spoke of chain-reactions, whereby the army defeated in distant Central Asia moved west to defeat its nearest neighbor, who in turn fled West until Europe was finally attacked. Documentation is poor for the earliest period, but during the Chinggisid era we have the complete story. First the Qitai conquered Northern China at the end of the T’ang; then the Jurchen conquered the Qitai, whose refugees established the Qaraqitai state in Central Asia, eventually dominating the Khwarizmian state; then, when Chinggis’ Mongols defeated the Naiman Mongols in the west, they took over the Qaraqitai state; then, after the Mongols had defeated the allied Naiman and Khwarizmian armies, the defeated Khwarizmians fled to the Middle East, where they helped the Muslims conquer Jerusalem.  In the thousand years starting about 300 A.D., in fact, virtually the whole civilized world (except for Byzantium, Japan, and Southeast Asia) was conquered by nomadic or barbarian Arab, German, Hungarian, Turkish, or Mongol invaders.4

For 2000 years the steppe was an uncontrolled zone. In the sedentary world men could be controlled by controlling land, and that’s what the castles were for.  Peasants had no wealth except their land and the crop on the ground, and could not flee. For this reason they could be heavily taxed, and the more impoverished they were made by heavy taxation, the harder it was for them to escape. Nomads also may have been better fed than peasants simply because, however poor they might be, they were harder to tax — on the steppe there was little “surplus”.

On the thinly-populated steppe there was no real property at all, but only movables: herds, flocks, wagons, and yurts. There was no real land ownership — land belonged to whomever had the manpower, the horses, and the military prowess to take it and keep it. Castles were mostly useless on the steppe, since steppe warfare was cavalry warfare, and food stocks were sheep on the hoof, so instead of controlling men by controlling their land, rulers controlled land by controlling menand land was controlled by controlling men. As a result, steppe wars were fought between fluid coalitions, riddled with desertion and fission, which swiftly gathered and could equally swiftly dissolve, and many of the practices of Mongol military organization were intended to keep weakly-committed partners from deserting. Chinggis Qan in the Secret History is not a great hero, but a shrewd, charismatic, persuasive leader who draws men into his service by his fair dealing, and the key points in his rise to power come when members of the corrupt enemy camp switch to his side, bringing crucial strategic information with them.

Since there was little surplus on the steppe, sedentary wars against the steppe were money-losers and had to be financed from taxes on the peasants.  Since the sedentary world was much wealthier than the steppe world, these wars were possible. However, the steppe was extremely inhospitable to armies from the civilized world, since almost all supplies had to be carried in on wagon trains which were vulnerable to attack, and since the steppe cavalry’s speed of concentration in attack was matched by their speed of dispersion in retreat, meaning that the invading armies would not have a single target to conquer, but would have to sweep the entire area one sector at a time. The most common response of the sedentary states was make alliances with one group of nomads so that they would fight the other nomads or to simply to pay steppe mercenaries to protect the sedentary states. This was expensive, but less expensive than trying to invade the steppe, and much less expensive than submission to repeated raids from the steppe.  The danger was in letting the nomads get their foot in the door, and usurpation of power by Mongol, Turkish or Norman military specialists was at least as important as conquest during the barbarian takeover of civilization between ~300 A.D. and ~1300 A.D.

As time went on hybrid civilizations developed which combined the military strengths of the steppe with the logistic advantages of the sedentary world. The Qitan Liao dynasty controlled a rather small area of China proper together with large areas of steppe.5 As a result they were able to wage sustained campaigns which the traditional steppe raiders could not. By the time of Chinggis Qan, hybrid dynasties stretched from the Yellow River to Baghdad. Each of them had a peculiar and unstable dual structure, with the local agricultural areas controlled partly from walled cities which always were capable of declaring their autonomy or switching to a new overlord, and and partly by cavalry armies sweeping the open spaces.6

By and large the hybrid states had an advantage over the pure sedentary states, who often had to buy horses from the steppe peoples (often inferior horses) at exorbitant prices and seldom were able to match the nomads at their own game (though they often did use cavalry to great advantage against their purely sedentary enemies). Ultimately, however, it was the nomadic Mongols  rather than the hybrid states who triumphed. Why?  One reason, I think, is that as the sedentary world adapted its military practices to the imperatives of cavalry warfare, the steppe peoples (notably under Chinggis Qan himself, as described in books 8-10 of the Secret History) were adapting their practices too, especially by mastering siege warfare — in other words, there was a learning curve on both sides of the line. Second, there was often a tendency of the sedentary world to underestimate the steppe and to be overoptimistic, especially during times of peace bought by tribute. For example, the Jurchen Chin of Northern China, a martial non Chinese people ruling a hybrid state, abandoned their Tatar steppe allies and then later maintained an aggressive policy simultaneously against the Mongols and the Sung dynasty of South China far longer than they should have. The “divide and conquer” strategy used for centuries against the steppe peoples by the Chinese, was successfully used by the Mongols against their Chin, the Xixia, and the Sung adversaries.7 


1 Boodberg’s electric comparison is applied to China rather than Central Asia: ” [T]he Western world…. knew China only, so to speak, at her anodic and cathodic electrodes. It was more conscious of a might current of energy emanating from a vast electrolyte at two given points than of the spacious body lying between.” (p. 2).  The two poles are seemingly the Chinese terminus of the overland Silk Road, and the Chinese port-city terminus of the southern sea routes.  These are not polar to each other, but both seem to be cathodes discharging energy toward Europe. Seemingly this metaphor could be adapted to make China and Central Asia together into a single powerful electric battery.

2  I should at least note in a footnote that the pre-Alexandrian Greeks were much more like the Vikings than they were like any of the civilized empires of history. As Lane has noted, forced trade and plunder are just the limit cases of trade, since long-distance trade always requires a degree of violence for protection purposes.

3 In the Chinese military philosopher Sun Tzu, the metaphor of the sudden release of potential energy is also used: “The onset of troops is like the rush of a torrent, which will even roll stones along in its course.” (p. 37, V:12) The commander’s control is also compared to the trigger of a crossbow, which allows a tiny movement of the finger to release the enormous power stored in the bent bow (p. 38, V:15). Still another comparison is made to “a round stone rolled down a mountain thousands of feet in height”. (p. 41, V:23).  The later Mongols did not pay any attention to the military writings of the defeated Chinese, and there is no reason to believe that Chinggis Qan learned anything from Sun Tzu, though it is possible that he was in the Chin Chinese service for a time during his twenties (Ratchevsky, pp. 49-50).

4  The Germans, of course, were not nomads, nor were the Jurchens – though both were closely allied with nomads during critical periods.  The sedentary barbarians of Western Europe and Manchuria were considerably different from the nomadic barbarians of Central Asia, and ultimately these two areas were civilized relatively easily.

5 While steppe land is less productive than sedentary land, agriculture is not impossible on the steppe. In fact, the best pasture land is also the best wheatland, and the Scythians, notably, were grain exporters. Grain production on the steppe, however, was less common than it might have been. Partly this was because of the same military imperatives I’ve been discussing — on the steppe, wheatland is hard to defend. Second, during some periods the steppe peoples were political and military specialists, and simply for reasons of comparative advantage they didn’t grow grain : it was more economical to get it from the sedentary world by some combination of trade and extortion.

6 The Khwarizmian Empire and the Chinese Jurchen Chin dynasty, both hybrid states, appear in Mongol history as pitiful victims, but the Jurchen had often been successful against the Sung, and the Khwarizmian Shah Muhammed had been steadily expanding his empire in the decades before the Mongols destroyed him, ultimately gaining a foothold as far as on the Arabian peninsula. It is notable, however, that many of his conquests involve gaining the nominal adherence of walled cities who retained their own leadership and retained a capacity for independent action. Walled cities and castles make the control of peasants possible, but unless carefully controlled from the center, they tend to be a force toward fragmentation, as in feudal Europe. The Mongols solved this problem by tearing down their walls.

7 A demographic cycle has been seen in China’s dynastic history, whereby the austere dynastic founders are succeeded by increasingly larger numbers of idle and parasitical descendants, whose demands put great pressure on the state treasury. Quite possibly a similar dynamic held on the steppe too, as a founding Qan’s successors would have increasing trouble keeping all claimants for shares in the booty happy. When this factor is combined with intermittent fiscal and agricultural crises and the temptations of complacency and the, it can be seen that the steppe frontier was intrinsically unstable.


Boodberg, Peter A., “Turk, Aryan, and Chinese in Central Asia”, in Selected Works of Peter A. Boodberg, California, 1979: , pp. 1-21.

France, John, Western Warfare in the Age of the Crusades, Cornell, 1999.

Frank, Andre Gunder, VU Press, The Centrality of Central Asia, 1992.

Gellner, Ernest, Anthropology and Politics, Blackwell, 1995.

Ibn Khaldun, tr. Rosenthal, The Muqaddimah, Bollingen / Princeton, 1967.

Khazanov, Anatoly, Nomads and the Outside World, Wisconsin. 1994.

Lane, Frederic, Venice and History: The Collected Papers of Frederic C. Lane, Johns Hopkins. 1966.

Lattimore, Owen, Inner Asian Frontiers of China, 1962, Beacon, “The ‘Reservoir’ and the Marginal Zone” pp, 238-251, .

Lewis, Mark Edward, Sanctioned Violence in Early China, ‎ SUNY, 1989.

Ratchnevsky, Paul, Genghis Khan: His Life and Legacy, Wiley, Blackwell, 1993.

Sinor, Denis, Inner Asia and its Contacts with Medieval Europe, Ashgate/ Variorum, 1977,  I: “Central Eurasia”.

Steensgaard,  Niels, “Violence and the Rise of Capitalism”, Review (of Braudel Center), V:2, Fall 1981, pp. 247-73.

Sun Tzu, tr. Lionel Giles, CMC / Ch’eng Wen reprint, 1978.

Who were the Mongols?

“2000 years of barbarians in 50 minutes”: