It seems then that the Nomads, in so far as they remain pure Nomads, are really a “people without a history” after all. Their eruptions out of the Desert into the Sown, like the eruptions of a Vesuvius or an Etna, are the mechanical resolutions of vast but inanimate physical forces.
Toynbee, p. 452.
To explain the great invasions we must look for the cause in the life of these peoples itself, in that history which is supposed not to exist.
Lattimore, p. 248, citing Fox’s Genghis Khan, 1936.
Although most of the above direct migration records have not given descriptions of the causes of the migration except those of the migrations caused by war losses….”
Fang / Liu, p. 163.
Anyone with even a moderate interest in world history will usually know of the various theories holding that nomad invasions are a function of weather and climate – most commonly, that droughts cause desperate nomads to pour into the civilized world. This kind of theory was popularized by Toynbee’s Study Of History (vol. 3, 2nd ed. , 1935) and first developed in Huntington’s 1907 Pulse of Asia. These theories were never empirically well-supported, and they also had serious conceptual deficiencies, but they were taken very seriously for a considerable period and have proven to be surprisingly resilient; they motivate research even today.
On the one hand, I believe that these theories have received much more attention than they have deserved, and are symptomatic of pathologies of method and approach; on the other, it may be that we finally now are on the point of having enough facts to seriously test them (though I doubt that what comes out of the new researches will much resemble the earlier theories).
After discussing the climatic explanation itself, I will speculate about the motivations for these explanations: primarily the modern historian’s desire to provide scientific-seeming deterministic, causal, reductionist explanations; and secondarily, civilization’s need to dehumanize its nomadic enemies.
HUNTINGTON AND TOYNBEE
According to Toynbee, the nomads do not act, but only react. Nomadism itself is “a response to the searching challenge of dessication” (p. 8), and all the recorded eruptions of nomads….
…. are not, as a matter of fact, the spontaneous expressions of the Nomads’ human initiative, but are all produced mechanically by the action upon the Nomads of either one or the other of two alternative external forces: either a pull exerted by one of the sedentary societies in the neighborhood of the Steppes, or else a push exerted by the climate of the Steppes themselves (p. 396).
The “push” is dessication. Toynbee quotes Huntington as follows:
In the deserts, the Nomads increase in number, and their flocks attain great proportions. The crest of a climatic wave is reached. …. [but if there is a reduction of rainfall ] soon he comes into conflict with his fellow Nomads, for all move to the best pasturage and the most permanent waters…. The weaker party is driven out, and begins to wander in search of new pastures and springs. Conflict follows conflict. At length the tribes which have been driven forth grow desperate. Impelled by despair, they pour forth in wild hordes upon the nations round about (p. 435).*
Furthermore, Huntington had claimed that these climate changes were global and occurred in a regular cycle, and Toynbee followed him, claiming that
…. our table of Nomad eruptions is beginning to reveal a regular pattern of movements which are periodical in Time and uniform in Space over the whole extent of the Steppes from the Atlantic coast of the Sahara to the Great Wall of China…. if we think of explaining our six-hundred-year cycle of Nomad eruption as the effect of a climatic push and not of a human pull, we shall find that our hypothesis is reconcilable to the climatic theories of at least one modern school of climatologists. This school, whose most eminent representative is Dr. Ellsworth Huntington, believes that there is a periodic shift of the successive climatic zones that encircle the globe latitudinally, so that the arid subtropical zone…. is oscillating all the time, with a regular periodicity, between two extreme geographic positions” (p. 430; see also p. 437, p. 440).
The outcome is a grandiose theory that sees, between 2025 BC and 2175 AD (sic), neatly repeating 600 year cycles of nomad invasions from from Morocco to Manchuria. In 1928 Owen Lattimore wrote sympathetically of Huntington’s thesis (p. 62), but he was much more skeptical in 1938 (p. 243):
By piling one assumption on another it is possible to tabulate a history of successive migrations in a manner that looks astonishingly accurate and convincing. It is well to remember that the results that look so solid are based largely not only on a chain of speculative causes and assumed effects, buy often on original details which are much too fragmentary to carry so solid a superstructure.
While no one advocates the Huntington-Toynbee thesis in its pure form any more, I am not really beating a dead horse. Toynbee was a giant in the field, and I suspect that the frequent, seemingly obligatory references to the possibility of climatic causation whenever the steppe is discussed, perfunctory though they often are, are monuments to his past influence, and attempts to revive some adapted version of this theory are, in my opinion, unnecessarily common. And as is often the case, regardless of this theory’s status among specialists, it still thrives among generalists and popular writers.
MORE RECENT CLIMATIC THEORIES
I have examined seven recent versions of the climate thesis (Meyer, Lamb, Jenkins, Jacoby, Buell, Atwell, and Fang / Liu). Three of them (Jacoby, Jenkins, and Fang / Liu) actually include data. However, Jacoby’s and Jenkins’ data disagree, and Jacoby’s data seem better:
In conclusion it is, ex hypothesis, my contention that a major climatic downturn did much to encourage an end to the infighting and vendettas among the Mongol clans and make possible their reorganization under Chinggis’ military authority (Jenkins, p. 226).
Jenkins’ hypothesis that unusual cold continued during Genghis Khan’s actual reign (1206-1228) is not supported by modern tree-ring paleoclimate studies for northern Mongolia after the very early 1200s…. Genghis Khan’s consolidation and expansion of empire coincided with a period of increased growth (warmer) (Jacoby, p. 55).
The Fang and Liu paper is more promising: it shows that the climate during the Tang dynasty, when China was united and dominant over the steppe, to have been warm and wet, whereas the preceding and following periods of Chinese disunity and nomad power were colder and drier — i.e., during the pre-Tang era when the Toba and the Turks were powerful, and during the post-Tang period of Uighur, Khitan, Jurchen, Tangut, and Mongol rule). However, these data, taken as such, tell us nothing about the later Uighur-Khitan-Jurchen-Mongol succession during a fairly uniform cool period, when the later nomad groups conquered the earlier nomad states, and in point of fact, the Uighur ascendancy began during the warm period. Nonetheless, it is an interesting result, and should be followed up.
Besides being weak in empirical support, climate-causation theories are also often confused about the mechanism. The original Huntington – Toynbee theory combined demographics and climate change: population increased during a period of good climate, and then a climate change for the worse sent desperate nomads into the sedentary world. (This theory is a strange composite, because demographics during any period could create population pressure even without climate change, though climate change could enhance this effect and trigger outbursts). Of the authors listed, Lamb speculates about this “trigger” theory for the Mongol invasions, with a climate downturn at the end of a prosperous period leading to population pressure, steppe disunity and warfare:
The pastures were in better than usual shape [but there is a] suspicion that some more sudden event connected with the cooling triggered it off (pp. 184-5)….. The Mongol homelands in central Asia seem to have been thriving and overpopulated (p. 317) ….
Jenkins, however, holds that deteriorating climatic conditions had the opposite effect, forcing steppe unity under Genghis Khan and leading to an enormous outburst of steppe aggression:
It can hardly be doubted that the cause of these huge displacements of whole populations must be sought in climatic changes (p. 219, citing Lattimore from 1928).
….a major climatic downturn did much to encourage an end to the infighting and vendettas among the Mongol clans and make possible their reorganization under Chinggis’ military authority (p. 226)
Atwell, on the other hand, holds that declining climate conditions weakened the sedentary world and made it vulnerable to attack:
Of interest here is the fact that virtually everywhere that the Mongols attacked during this brief period recently had experiences, or was just then experiencing, significant economic problems caused, at least in part, by “anomalous” climatic conditions. The extent to which the Mongol homeland itself was being affected by these same anomalies remains to be studied…. (p. 45)
Finally, Buell (2003, pp. 19-21 and 197) proposes a version of Hudson’s note in Toynbee (also endorsed by Lattimore): improving weather conditions led to sedentary encroachment on grazing lands and made the whole border area into a conflict zone.
But after a century we still don’t have what we would hope for, and what would seem essential for a theory of this type: two correlated time series (one of climatic conditions and one of nomad movements) together with a specific theory explaining the correlations. All we are left with is McGovern’s satirical composite theory – “It got cold / hot / wet / dry and they died / flourished / migrated / multiplied / intensified” – and Ladurie’s question: “But what is one to conclude from such contradictory and unprovable speculations?” (both cited in Meyer, p. 289). Of course, the fact that one climatic explanation is the opposite of another can simply mean that one theory is good and the other bad, but one is left with a lingering suspicion that this explanation, however robust it may seem, has broken free from the need for evidence.
A NON-CLIMATIC THEORY OF NOMAD INVASIONS
Nomad invasions / migrations are not mysterious and puzzling events, and climatic explanations of nomad behavior need to be integrated with what is already known. Historians’ generalized discussions of nomad movements frequently confuse several different kinds of event: actual invasions with the goal of conquest, e.g. the Mongol invasions; more or less organized frontier raiding intended to extract plunder and tribute (e.g., the Xiongnu in China); migrations within the steppe, e.g. the Yuehzhi migrations after their defeat by the Xiongnu, or the Torgut migrations from China to the lower Volga and back; and finally, the flights of small groups to the sedentary world, where they are sometimes accepted as subjects and sometimes slaughtered as invaders.
The lines between these categories are not always clear and the category a given movement ends up in can be quite fortuitous, but in many cases the differences are clear enough to make a single-cause explanation doubtful. Military defeat on the steppe is the primary push, and the primary pull is the enormous wealth differential between the steppe world and the sedentary world — the result of the denser sedentary population, the more diversified sedentary economy, the peasant’s relative helplessness in the face of violence, which facilitates the collection of taxes and rents and the accumulation of concentrations of wealth and steep social hierarchies. Neither of these requires a climatic explanation, and the pull of sedentary wealth is is always there regardless of climate.
Given this general explanation of steppe military activities, what is the significance of the climatic factor? First of all, as Lattimore explained, the steppe / sown line is defined to a great extent by climate: areas too dry for crops can still provide excellent pasture. At the same time, climate is not the sole factor drawing that line, since some potential croplands end up reverting to pasture simply because of military pressure from nomads during periods of disorder. But it at least can be said that long- and medium-term climate changes sometimes move the steppe / sown line in one direction or the other, thereby changing the military relationships in various ways — for example, as suggested by Buell.
Second, in the near-subsistence economies of the past, even one or two years of drought or other disaster could cause an economic crisis. This is true both in the pastoral and in the sedentary areas, but it happens more quickly in pastoral areas since pastoralists are not able to store much food except on the hoof. This is the “trigger” explanation which, however, can only be used to explain a specific movement following a sudden dramatic change during the brief period of one or two years, or even just a single spring and summer; the general long-term trends do not trigger anything. And the drought has to be bad enough to motivate movement, but not bad enough to cripple the group in question.
Third, the density of population and military strength both of the steppe world and of the sedentary world are substantially dependent on climate, with warm wet weather making both worlds more prosperous and stronger. The military consequences of this are uncertain, however. They were probably greatest when good conditions in one of the two worlds was accompanied by bad conditions in the other, but most climatic theories so far have assumed general climate change, without specifying variation between different areas.
Fourth, the specific motivational effects of either good or bad conditions on the steppe peoples are uncertain. By and large, nomads went to war when their ponies were fat, and the major, successful nomad military campaigns (the ones which have forced historians to think about this topic) do not at all fit Huntington’s model of desperate tribes pouring forth in wild hordes” The trigger explanation might fit some of the migrations within the steppe or some of the flights of defeated peoples taking refuge in the sedentary world. But even these cases, military defeat is always the proximate cause.
Finally, while unity or disunity on the two sides of the line is the major factor deciding which side gets to play “divide and conquer”, and which side will emerge victorious, neither Huntington’s theory (that declining weather conditions cause disunity) nor Jenkins’ opposing theory (that declining conditions make unity possible) is convincing.
The climatic theory of the nomad invasions was developed during the era of high positivism, when ambitious determinist-reductionist theories, often cyclic, were uncritically welcomed; in the extreme cases the goal of these theories was to describe all human behavior in terms of external causes, leaving no place for any human agency at all. Often these theories were also shaped by grand geopolitical theories of the rise and fall of empires and European dreams of world dominance — stories that had to be simple and vivid in order to be effective. And a final strand was the age-old civilized suspicion that there was something uncanny, unnatural, mysterious, and inhuman about the nomads and their triumphs. At a certain point some version of the climatic theory became a required, obbligato part of any discussion of nomad history, always good for a speculative paragraph and requiring at least a nod even from skeptics.
At the same time, however, this discussion may now be entering a new era, when the theory can actually be tested against data and refined or rejected accordingly. If the data keep coming in, at some point it should be possible to produce a nuanced case-by-case history of the two millennia of nomad incursions within which the role of climate change and weather variations can be plainly seen, in some cases driving desperate refugees from the steppe to become either bandits and raiders or mercenary defenders of civilization depending on their reception, and in other cases allowing the nomads to build up enormous, well-fed armies capable to sweeping all before them. But it seems highly unlikely that these new histories will include grand universal theories of the Huntington-Toynbee type, which seem to me so weakly grounded that they should just be forgotten.
FN The “pull” of the wealthier cultures, of course, is not a mechanical cause like climate or weather, but a motive, identical with one of the main motives of warfare in the civilized world.
The Retreat of the Elephants
In Mark Elvin’s book The Retreat of the Elephants there is what I believe to be an unnecessary pro forma reference to the climatic thesis. As part of his explanation of the disappearance of elephants from China, we read this:
What were the causes of this disaster (from the elephant’s point of view)? In part it was likely to have been the cooling of the climate, as referring back to the last column in Table 1 in the previous chapter will suggest. Elephants do not resist cold well (p. 9)….
Climatic change must have been a factor in their long and eventually fatal southward retreat…. (p. 32)
But Elvin also gives four other causes: the elephant-hunts of the nobility and the king, private commercial ivory-hunting, habitat destruction by agriculturalists, and hunting by farmers protecting their crops. Furthermore, the climate data he gives are not convincing. His Map 2 (”The Retreat of the Elephants”, p. 10) shows the constant retreat of the elephants from 5000 BC onward, and there is little correlation with the climate conditions shown in his Table 1 (“Time-marks in Chinese environmental history”, p. 6). The largest single retreat was between 580 A.D. and 1050 A.D., but according to the Fang / Liu article, cited by Elvin, as well as the Jacoby article I have referenced above, that was was a warm period.
With regard to the disappearance of the very last elephants in SW China around 1400, Elvin (2006, p. 117) cites a paper by Robert Marks (1998, p. 46). “In the case of the forest elephant of Lingnan in South China, its disappearance around 1400 can be explained by either hunting, climatic change, or habitat destruction”. However, while Marks acknowledges the possiblity of a climatic explanation, he does not commit himself to it, and in fact he leans toward the theory that the human factor is most important. In any case, Marks is only writibng about the last three centuries of a 2000 year process.
Thus, between 900 BC and 1100 AD at least, and probably all the way until the elephants’ final disappearance in 1400 AD, it seems to be the case that the various sorts of human pressure were an adequate explanation for the retreat of the elephants, and that in the absence of much more data there is little real reason also to speak of climate at all.
Atwell, William, “Volcanism and Short-term Climate Change in East Asian and World History”, Journal of World History, Vol.12, #1, Spring, 2001, pp. 29-98.
Boodberg, Peter A., Selected Works of Peter A. Boodberg, California, 1979: “Turk, Aryan, and Chinese in Central Asia”, pp. 1-21.
Buell, Paul D., “The Role of the Sino-Mongolian Frontier Zone in the Rise of Chinggis-Qan”, pp. 63-76, Studies on Mongolia, ed. Schwarz, Bellingham, 1979.
Buell, Paul D., Historical Dictionary of the Mongol Empire, Scarecrow Press, 2003.
Curtin, Jeremiah, The Mongols, Combined Books 1996 / Little, Brown 1908: “Foreword”, Theodore Roosevelt, pp. ix-xv,
Douglas, Mary, “Passive Voice Theories in religious sociology,” In an Active Voice, Routledge
Kegan Paul, 1982
Elvin, Mark, The Retreat of the Elephants, Yale, 2004.
Fang Jin-Qi and Liu Guo, “Relationship between Climatic Change and the Nomadic Southward Migrations in Eastern Asian During Historical Times”, Climatic Change, Vol. 22. 1992, pp. 151-169.
Frank, Andre Gunder, VU Press, The Centrality of Central Asia, 1992.
Frank, Andre Gunder, The World System, Routledge, 1993.
Gellner, Ernest, Anthropology and Politics, Blackwell, 1995.
Heather, Peter, Empires and Barbarians, Oxford, 2009.
Ibn Khaldun, tr. Rosenthal, The Muqaddimah, Bollingen / Princeton, 1967.
Jacoby, Gordon, “Tree Rings, Climate History, and Genghis Khan”, in Fitzhugh, Rossabi, Honeychurch, eds., Genghis Khan and the Mongol Empire, Arctic Studies Center, Smithsonian, n.d. (2007 or later), pp. 53-55.
Jenkins, Gareth, “A Note on Climatic Cycles and the Rise of Chingiss Khan.” Central Asiatic Journal,Vol. 18, No. 4(1974): pp. 217-226.
Khazanov, Anatoly, Nomads and the Outside World, Wisconsin. 1994.
Lamb, H.H., Climate, History, and the Modern World, Routledge, 1995.
Lattimore, Owen, Studies in Frontier History, Oxford, 1962 (especially “The Geographical Factor in Mongol History”, pp. 241-258).
Lattimore, Owen, Inner Asian Frontiers of China, 1962, Beacon: pp, 238-251, “The ‘Reservoir’ and the Marginal Zone”.)
Marks, Robert, Tigers, Rice, Silk, and Silt, Cambridge, 1998.
Meyer, William B., “Climate and Migration”, in Bell-Fialkoff, ed., The Role of Migration in the History of the Eurasian Steppe, St. Martin’s, 2000, pp 287-294.
Sinor, Denis, Inner Asia and its Contacts with Medieval Europe, Ashgate/ Variorum, 1977, I: “Central Eurasia”.
Sun Tzu, tr. Lionel Giles, CMC / Ch’eng Wen reprint, 1978.
Steensgaard, Niels, “Violence and the Rise of Capitalism”, Review (of the Braudel Center), V. 2, Fall, 1981, pp. 247-73.
Steensgaard, Niels, Coercion, Capitalism, and European States, Blackwell 1992.
Toynbee, Arnold, A Study of History, vol. III, 2nd ed., 1935.
Wolf, Eric, Europe and the People Without History, California, 1982.