Wednesday, September 02, 2009

Who's the barbarian now? Empires of the Silk Road   posted by Razib @ 9/02/2009 01:00:00 AM

If there is one word that is applicable to Empires of the Silk Road, Christopher I. Beckwith's magnum opus, it would be dense. The Horse, the Wheel, and Language is a complement of similar density and topical intersection (they do not quite overlap, and The Horse, the Wheel, and Language is missing from the bibliography of Empires of the Silk Road). A quick perusal of Beckwith's ouvre shows that he writes history from the center. What we might term the Ecumene he calls the "periphery." Page after page he defends the "barbarians" of the Heartland against the slander of peripheral scribblers. Though in the introduction the author takes a stand against both the Whiggishness of much of Modernism and the vacuous relativism of Post-Modernism, this work is clearly written as something of a corrective, not an objective treatment where all scales are balanced with care, caution and precision.

One of the themes of Empires of the Silk Road which distinguishes it from most pre-modern history is its mobility, scope and breadth, of space. A history of ancient Egypt or the Roman Empire spans centuries, and even millennia. But their geographic expanse is limited. Roman expeditions to eastern Germany shock and surprise, but the Roman Empire was fundamentally a polity of the western periphery of Eurasia. In contrast Beckwith's subject matter naturally pushes the boundaries to a far greater extent. For example, the flight of the Avars, who dominated what is today Hungary for several centuries, covers from what is today Mongolia to the Danubian plain in less than 10 years! It is a common assert that before the modern era a typical human would be born, live and die with a 10 mile radius. The exact number of miles is irrelevant, the moral is that one wishes to project a sense of incredible spatial stasis which we modern people could not comprehend. Similarly, the fashion in archaeology to assert that a shift in culture as evident through artifact is never through migration, but always through communication, internalizes this assumption. But the reality is that this assumption is only valid for the vast peasantry of peripheral societies (if even there), not for hunter-gatherers, nomads and those who practice mixed-lifestyles. In other words, not for the peoples of the Eurasian Heartland which is the subject of Empires of the Silk Road. The very high likelihood that much of what we today call Xinjiang was inhabited by peoples of Western provenance is one of those "mysteries of the ancient world" which surprises us, but that's because we're the descendants of peasants who are conditioned to assume that pre-moderns were immobile! Conversely, the descriptions of some of the more exotic nomadic groups which arrived on the Central European plain after the collapse of the Roman Empire strongly hint at an East Eurasian provenance for an element of these hordes (the physical characterizations of the Huns and Avars seem qualitativey different from that of Scythians and Samartians).

There are two great pulses, Volkerwanderung, which loom large in Beckwith's narrative, that of the Indo-Europeans and of the Turks. The former is always much more difficult than the latter because in the case of the Turks we have copious records from other groups as to their appearance on the margins of major civilized states. In contrast, when the Indo-Europeans were on the march there was very little widespread literacy, and what there was was often devoted to the kind tedious accounting which dominate the Linear B tablets. The model in Empires of the Silk Road seems to lean strongly on Robert Drews' The Coming of the Greeks, chariot riding martial elites swooping down on the settled peripheral civilizations from the Eurasian Heartland, and rapidly generating synthetic creole cultures in a matter of centuries if not decades. In Beckwith's telling the Indo-Europeans mastered the horse and chariot, and used these technological advances to take over civilized states, though he suggests that instead of full-frontal attack the norm would much more likely be that warbands were invited or their services purchased by the settled elites. Only after this period service would the Indo-European warriors rebel and take over the reigns of power (think the takeover of Britain by Anglo-Saxons after their having been brought over as mercenaries by native warlords). This specific model of servitude is assumed to be an instantiation of a general dynamic which characterizes the ethos of the Heartland, fealty of the comitatus to a leader who they will follow 'till death. We are familiar with this in its classical form among the Mongols or the Japanese samurai, but Beckwith suggests that Islamicized Turks and Christianized Franks continued to express some of the values of this ethos even after repudiation of its more extreme aspect of collective suicide in case of the death of the leader. This sort of non-kin based group cohesion can often be very powerful, and even if it is the exception historically in many societies, these exceptional periods can sometimes serve as hinges which determine the course of nations. The Arabs are a kin-based society, but the presence of non-kin among the Muhajirun, and the obvious non-kin status of the Ansar, is notable as it was they who helped Muhammad rise to power. Though the early Muslim movement was co-opted by the Umayyad clan, it nevertheless relied on an esprit de corps which was not kin-based in its early decades (though it was to a great extent ethnically based, see The Great Arab Conquests).

Though the effect of the chariot in the short term, and Indo-European languages in the long term, are not dismissed, one of the interesting suggestions made in Empires of the Silk Road is that perhaps more critical is the role that Indo-European nomads, and perhaps in particular North Iranians, played as facilitators, communicators and even originators, of ideas and technologies across Eurasia. It is not an exaggeration to note that sometimes Beckwith seems to see Indo-Europeans everywhere. The most controversial and perhaps tendentious claim (which he admits is not the scholarly consensus) is that Indo-Europeans may have played an essential formative role in the emergence of Shang & Zhou China. In particular, it is known that some elements of the culture of the Shang elite is of western origin, in particular the style of chariot warfare. It is also known that the Zhou were a semi-barbarian group from western China, and Beckwith makes a philological argument that the maternal lineage of the Zhou are not Tibetan barbarians, but Indo-European ones. He even goes so far as to offer that perhaps the original Chinese language was Indo-European, now overlain by the indigenous substrate so as to be unrecognizable. I find this last to be rather implausible, but then I know little philology so I can't critique it technically. Nevertheless, the fact remains that we know that Indo-Europeans were a presence in ~2000 BCE in what is today western China, and some aspects of technology were transmitted from them. My skepticism is partly due to the fact that the Han populations of China are genetically very distinct from those of West and Central Eurasia. The Europoid remains from Xinjiang have had enough of their DNA extracted so as that we know that on many genes where Europeans and Chinese are disjoint in frequency, so were they. Additionally, some historical remains of individuals who were claimed to be of western origin have been analyzed, and yes, they were West Eurasian. The thinness of distinctive West Eurasian genes in China proper, genes which are found in abundance among the Uighurs, shows that there were limits to the genetic admixture in this case, to the point where the indigenous substrate totally absorbed and marginalized the exogenous input. This may explain why even if Indo-European warbands had a critical role to play as a cultural stimulus among the elites of the North China plain they seem to have left a rather marginal biological impact.

The contrast with India is illustrative, as there are candidates for genes which the Indo-Europeans brought. Not only that, but the sex asymmetry in terms of genome content is very striking in South Asia, exactly in the direction which Beckwith would argue insofar as they are roving bands of males who engage in societal takeover. India naturally does seem to have a highly creolized culture, most manifest in the Indo-Aryan languages of northern India which show many hallmarks of being the synthetic product of local dialects and an intrusive language. The author barely even manages to conceal his contempt for South Asian nationalists who make arguments to the effect that Aryans were indigenous to South Asia. I think that though the tone could have been a bit more scholarly, anyone who looks as the total range of data from all Indo-European traditions can not escape the conclusion that the probability that Indo-Europeans originated in South Asia is very low, to the point of triviality. He exhibits the same attitude toward those who argue for a deep history of Indo-Europeans in Greece or Anatolia, sentiments with which I sympathize and am in general agreement, but, I would have liked the arguments sketched out in more detail rather than asserted blithely (the footnotes on this topic often don't satisfy since Beckwith cites his future work!).

One interesting facet of the story in Empires of the Silk Road seems to be the winner-take-all and positive-feedback-loop dynamic which reoccurs. The spread of the Indo-European languages is rather amazing, and invites more scholarly interest, but that is not where the story stops. Beckwith notes that for much of antiquity the heart of Asia was dominated by one group of Indo-Europeans, the Northern Iranians. The term "Iranian" here is something of a misnomer, as groups such as Scythians, Samartians and Alans may have had little to do with the contemporary Iran. Rather, the peoples of modern Iran speak a language which is related to the dialects of the Scythians. Up until fall of Rome the Northern Iranian nomads drove all others before them. There is a great deal of circumstantial evidence that the Iranians displaced the Indo-Aryans from the plateau of what later became Persia, and Herodotus seems to describe peoples under Scythian hegemony who would later become Slavs and Balts.

This winner-take-all dynamic repeats with the Turks, who swept the Iranians from the plains of Central Eurasia after marginalizing other groups such as the Avars in their climb to the top. What was so special about the Turks? Perhaps nothing. Perhaps to some extent this is a stochastic process which is characterized by positive-feedback-loops, someone has to win, and winners become even more powerful over the course of time. Eventually these great confederations which have eaten dozens of other smaller confederations expand until they lap up against the margins of peripheral civilizations, at which point the scribes of the settled marvel as to the savage superhuman nature of the nomads at the gates. In Genghis Khan and the Making of the Modern World the author clearly implies that Temujin's personal qualities were essential in allowing for the mobilization of the whole Mongol nation. He notes in particular his rejection of kin-based favoritism and elevation of his comitatus (he does not use that word, but same difference) as critical ingredients in scaling up the Mongol war-machine. But do note that the Mongols had risen before, under Genghis Khan's great-grandfather Khabul Khan, and been crushed by the Tatars. Perhaps Khabul Khan had all the necessary personal qualities, but the fates did not smile upon him, and now he is simply a footnote in history. The domain of the Goturks in the 6th century was nearly was expansive as that of Genghis Khan in his lifetime, but most readers of this weblog are likely unaware of the Goturks. Why? Some of this is simply that the Goturks crested at a lower maximum of territory, but some of it is simply that in the mid-500s much of peripheral civilization was in low ebb and so the Goturks did not come into conflict with the lettered lands (in fact, the Turks were often used sought out as allies against the true barbarians at the gates).

What I'm trying to say about the importance of positive-feedback-loops is not going to shock and surprise anyone. The larger the Roman or Persian Empire got, the more resources were at their disposal. So the more likely victory against other powers became (though at a certain point holding territory becomes more difficult due to diminishing returns on force projection with more numbers). But the parameters are somewhat different in the center. Rome took centuries to reach its "climax" configuration under Augustus (with only changes at the margins under Trajan, etc.). In contrast, most of the area of the Mongol Empire was conquered within 20 years. 30 years after the death of Genghis Khan Baghdad was sacked, and 50 years later all of China was finally conquered. It is understandable why the rate of conquest slowed as the Mongols pushed into the periphery; capture of the territory of nomads by nomads is far different from the taking and holding of cities, or the pacification of mountains. But winner-take-all dynamics on the plains of Eurasia did result in a series of polities which had the numbers and expertise to learn the game of peripheral states, and beat them at it, repeatedly. The Mongols famously learned siege warfare in in the process of conquering China, while the Ottomans hired European gunners, and the Huns made recourse to infantry in forested regions. But the size of their forces due to winner-take-all dynamics, and the cohesion of the comitatus, served them in good stead until the natural decay of soft civilization set in.

About which, Christopher Beckwith has serious objections to the characterizations which posit dichotomies between the soft sedentary and barbarian hardy nomads. He correctly points out that most "nomads" were not pure nomads, but lived a mixed lifestyle (hunting, herding and farming). Additionally, there's plenty of evidence that disgruntled peasants could turn roving brigand when authority was disrupted. So the boundaries between the two are not so clear cut (also, whole populations seem to have switched between settled farming and nomadic animal husbandry several times). The idea of the nomad as the born warrior is also exaggerated, and Beckwith points out the widespread literary evidence which attest to the love of luxury and laziness of the nomad populations (they're humans). And as evidence of the hardiness of peasants, he offers the examples of the citizen soldiers of Republican Rome (before Marius' reforms) and the Greek hoplites who formed the famous phalanx. Good points all, but Beckwith protests too much. He asserts repeatedly that the peoples of Central Eurasia tended to be larger and more well fed than the peasants of the periphery. Why? In a pre-modern society the population would push up against the Malthusian limit, so why were the nomads larger? To some extent this was a function of their higher quality diet in regards to meat and milk, but one might suppose that another issue was that there were greater ecological fluctuations which culled the nomad populations in a manner which was more extreme than among peasants. I bring this hypothesis up because in Empires of the Silk Road this particular problem, starvation due to theft of livestock of their deaths (disease?) is introduced as one issue which made the life of the nomad more uncertain than that of the peasant. But another factor may simply be interpersonal violence on the plains. If the peasant toiled on the margins under the boot of their sedentary nobles, they were often shielded from the sort of violence we read about in the raids common among herders. These facts lead me to contend that that Beckwith overemphasizes the symmetry between nomad and peasant when it comes to their viability as fighters. Finally, I think the body of evidence implies that nomadic populations did mobilize a greater fraction of their able bodied males in organized violence than peasant-based societies (consider Imperial Rome with its professionalized army). This would naturally give the average nomadic male more martial skills simply through experience.

Before I go to the most explosive contentions of Empires of the Silk Road, let me address a peculiar set of chapters near the end of the book. These chapters throw repeated polemical salvos at "Modernism." They attack populism, democracy, and the overall degeneracy of contemporary "high culture." Though they make the factual observation that the old architecture and traditions of Central Eurasia have been totally decimated by Modernist ideology, specifically Communism, the bigger issue here seems to be that Christopher Beckwith objects vociferously to the general slant of the Modernist era in culture. Though I have some sympathy with this, these sections of the book read rather uncomfortably next to the great mass of erudite, if assertive, scholarship which is the majority of the narrative. It is clear that Beckwith believes that Modernism is a base expression of human preferences, and that our better nature has its origins in the genius of the Heartland.

About that genius, the proposition is forwarded by Beckwith that the philosophical production of the Axial Age may have something to due to the facilitative or even original contributions of the peoples of Central Eurasia. As noted above these were peoples who could span great distances in short periods of time, so in some ways it is plausible that they would be the "Pony Express" of their age. Beckwith notes that at least one philosopher who resettled in Greece was Scythian, so the peoples of the Steppe may have been familiar with philosophy. Some of the ideas originated by the Greeks, Indians and Chinese seem eerily similar, while the religious genius of the Jews and ancient Iranians was also operative during this period. Zoroaster is likely a figure of Central Eurasia, not the sedentary world of Fars.

The simultaneous aspect of the Axial Age is rather peculiar. If you are a theist who believes in an active God this is one historical epoch which would seem a likely candidate for divine intervention. Barring that, though I am not convinced by Beckwith's hypothesis, I think a less ambitious model which posits the Central Eurasians not as drivers but as enablers of the spread of ideas is plausible. The civilized areas of the periphery were expanding during this period so they would be in more intimate contact with the Heartland than before, and also naturally more culturally productive because of gains of efficiency due to scale. There are cases later in history where the trade routes and peoples of Central Eurasia were critical in the spread of ideas: Buddhism became a religion of East Asia almost certainly through the cities of Central Eurasia. Islam also spread through the same trade routes, from the Crimean Tatars in the west to the Uighurs in the east. Finally, as Arnold Toynbee observed Far Eastern Christianity spread along these routes, during the time of Genghis Khan most of the tribes of western Mongolia was nominally of Nestorian Orthodox Christian affiliation.

In keeping with Beckwith's thesis that in the period between 2000-1000 BCE a series of cultures arose as syntheses between Indo-Europeans and the sedentary polities of distant antiquity he concludes that Central Eurasians are the spiritual ancestors of modern people, and not the Sumerians, or Egyptians, or the people of the Indus Valley civilization. This is naturally a contentious point. It is for example often noted that of the gods of ancient Greece only Zeus is of undisputed Indo-European origin, even the Mycenaean Greeks may mostly have been culturally indigenized! (or barely Indo-Europeanized if you invert it) Buddhism, arguably India's most impacting cultural export, has been argued by many scholars to be one example of a reemergence of some of South Asia's indigenous religious strands after a period of Vedic Indo-Aryan cultural domination. And so on. I can grant that the ethos of the Heartland was powerful, the expanse of the Indo-European and Turkic languages are manifest evidence of this, but I am not eager to abandon one biased narrative from another. The nature of modern civilization, and its antecedents, are complex. Empires of the Silk Road is admirable at exploring one particular neglected strand, but I'd rather not turn that tree into the forest just yet.

Note: I want to emphasize that did not touch upon many aspects of Christopher Beckwith's argument. When I say that the book is dense, I mean it is dense.