Since many readers of this website refer to “genocides,” and all of them were born in the 20th or 21st centuries, I want to put a note here which I think will illustrate why it is important to be careful of the use of particular words and what their connotations are as a function of time. In the modern period, the term “genocide” has a particular valence. The Nazi killing of Jews, the Ottoman genocides of the early 20th century, and the killing of Tutsis in Rwanda. These were, I believe, expressions of the mass politics and mobilization. As such, they are not entirely analogous to ethnic and religious turnover in the premodern era, where death was often secondary or a side-effect.
The Tutsis were a socially and politically dominant minority in Rwanda. A militarized rebel force was fighting the Hutu-dominated government, and that rebel force was dominated by Tutsis. The planned extermination of Tutsis in Rwanda was politics by another means. Similarly, various Christian ethnic minorities in the Ottoman Empire during the first decades of the 20th century were seen as stalking horses of European colonial powers in their conflicts with the Turkish Muslim dominated government.
Note that this was a period when the Ottoman state was attempting to transition toward a more modern understanding of its political system, which depended on the support and mobilization of the populace, less differentiated by religion or ethnicity.
Finally, the attempted extermination of the Jews by the Nazi regime cannot be understood except in the light of the theory of history that Adolf Hitler and many antisemitic nationalists promoted in the early 20th century. It was predicated upon a postreligious materialism whereby the Jew could no longer be assimilated into the non-Jewish majority.
There are two changes that happened with the modern world that made genocide as we understand it more feasible:
- The rise of mass politics, which I allude to above. That is, the transformation of political units from being coalitions of elites (e.g., early modern France) to being expressions of national-folk will (e.g., post-Revolutionary France). Ethnic and religious diversity within the state is not a problem when the state is an expression of the will of an oligarchy, rather than an ethnic and religious group.
- The rise of the state more generally. Pre-modern states were weak and relied on ideology and customary tradition to bind villages together. They were simply not capable of totalitarianism if they wanted to engage in such an activity (in contrast, a single city-state could perhaps engage in totalitarianism, which made the social engineering of the Greek polis more comprehensible). The fiat of the central government had limitations.
The Nazi regime managed to kill 90% of the Jews of Poland. This would not have happened under Frederick the Great of Prussia 150 years earlier. First, though Frederick was a conventional antisemite, he was not a genocidal one. And if the Jews had consented to convert to Christianity (Lutheranism) he would have been forced to tolerate them, despite his contempt for the Christian religion. A biologistic understanding of nationality did not quite exist in a systematic form, though its elements were already present.
And second, though at the time contemporaries were in awe of the size and power of the Prussian state, and in particular its military (e.g, “Prussia is an army that has a state”), it was a far weaker government than what had developed in Germany by the 1940s.
That being said, premodern states could engage in genocide if they wanted to. But it was not in the form of camps and firing squads. They could starve or expel recalcitrant peoples.
In the 18th-century the Dzhungar Confederacy of Oirat Mongols emerged as the last great empire of the steppes. Caught between the two gunpowder polities of the Russians and Manchus, their fate was to close out the 2,000 years reign of militarized pastoralists in Inner Asia. Modern-day Dzhungaria does exist, it is roughly the northern half of Xinjiang. But the native inhabitants are not Dzhungars, but Kazakhs. These were pastoralists that moved into the emptied Dzhungar rangelands.
How did the Manchus accomplish the ethnic cleansing then? Simple: in a premodern world of Malthusian limits you simply drive people away or force starvation by preventing them from extracting calories from their land. Huge numbers of Dzhungars migrated into the Russia Empire, with some of them eventually settling permanently in Kalmykia on the Volga. But the death toll was enormous. Most died during the migration (some eventually migrated back).
The Manchu targeting of the Dzhungars was exceptional. That is because the Dzhungars as a whole were mobilized as a political-military force in Inner Asia. They were a predatory power which extracted rents from vassals of both the Russians and the Manchus and was instrumental in decades of machinations in Tibet. Though a small people, in general, these steppe groups punched far above their weight because all free males were potential soldiers for a campaign, with women overseeing the herds while their menfolk were out on the campaign.
This is very different from agrarian populations. When Genghis Khan conquered northern China some record that this plan was to drive off the settled populace and transform the land into pasturage. Basically, he would have induced famine which would have meant most of the refugee population would probably die. One of his advisors explained to him that the rents produced by farmers would far exceed the wealth generated by animal herds. So the farmers lived.
Such a discussion brings into focus the reason that ethnically targeted physical genocide of whole peoples was usually not used as a tool of politics by agrarian states in the premodern period. People were wealth for elites, and killing people destroying wealth. I here use a very specific term: targeted physical genocide. The mass conversion of pagan Slavs on the Baltic frontier by Germans, and their assimilation into a German Christian identity, was cultural genocide. But the rents that knights could extract were maintained. The people lived. Their identity changed.
In fact, the last pagans in the Baltic were to be found on the estates of German Christians in Latvia, into the early 1400s, because pagan peasants were not subject to the protection of the Church from extreme exploitation. In other words, it was more profitable for German Christian elites to extract wealth from pagan peasants than Christian ones!
The reality is that in the premodern period there were many mass die-offs due to famine. Some of these were due to political and historical events. The province of Sichuan, for example, was repopulated to a great extent from Hunan in the 17th century. Part of the issue here is that famine was induced by massive conflicts between the Manchus and the Ming loyalists. Similarly, the qanat system of irrigation in Iran was massively disrupted during the Mongol invasion of the 13th century. This resulted in such depopulation that the region’s census size did not recover until the modern period. This is not due to the concerted attempt by the Mongols to depopulate Iran, but rather, the vicious instrumentalism of Mongol forms of warfare, which responded to resistance with total organized viciousness against both humans and their capital. Mongol genocide was not an ends, but a means toward showing other people why they should surrend and bend the knee as soon as possible.
All of this brings us to India and the idea that genocide was committed against Hindus by Muslims. To be frank, I don’t pay much attention to these sorts of arguments in detail, but there is not much detail. But let me first say that I now lean toward the position that the great Arab conquests of the 7th century were by a people who we would only vaguely recognize as Muslim today. That is, I think a coherent and recognizable Islam dates to the end of the 7th century, and most definitely by the Abbasids after 750 AD. As I have stated before, I believe that Islam is the product of Empire, it did not conquer an Empire.
Which brings us to the Turkic led predations upon India which began under Mahmud of Ghazni, and continue down until the conquests of the whole subcontinent begun by Muhammad of Ghor.* These conquests need to be understood in the context of steppe pastoralist predations that began with the Xiongnu in the centuries before Christ and continued down to the Dzhunghars in the 18th century.
The steppe is not poor on a per person basis in a relative sense. In the premodern wor,ld the vast majority of the population lived on the Malthusian limit. There were cases, such as in the Roman Empire, where trade and economies of scale allowed for the formation of a “consumer society” after a fashion (e.g., pottery mass-produced in the Mediterranean at particular locations and exported by water transport). But the gap between a Roman peasant and a Sarmartian pastoralist was small in a modern perspective. Rather, wealth is thought of as the aggregate of production of a population given across a region, which was easily understood to be a proxy for wealth for extractive elites.
India and China, or Egypt, were not wealthy because of high per person productivity, but because of high per unit productivity (fertile soil) which translated into large populations. From a purely economic perspective these people were rents that premodern elites could use to translate into wealth and prestige.
Unlike Genghis Khan, more sophisticated agro-pastoralist groups with more exposure to the Chinese world system, such as the Khitan and the Jurchen/Manchus, understood that the high population density of China meant that they could extract wealth out of the Chinese state far beyond what they could extract out of their own subjects. The Khitan operated like the Huns of the Late Roman Period, extracting protection money after threatening invasion. In some cases, the invasion had to be attempted, though ultimately this was a “lose-lose” situation. The Khitan did not necessarily recoup the opportunity costs of invasion through plunder, while the Chinese had to mobilize forces to defend themselves. Ultimately it was often less costly to payoff pastoralists for the Chinese state, and less costly for pastoralists to accept a payoff than work hard to plunder and conquer.
Of course, in some instances invasion did occur. It took the Mongols and Manchus two generations to conquer China. In the short term, this was a high-risk proposition, and the conquest itself resulted in the burning of the capital stock from which the conquerors would eventually extract rent. But, after the small number of Manchus and Mongols established themselves on top of the extraction pyramid that was China, they obtained windfall gains. They were far wealthier than they would have been as steppe warlords.
Which brings us back to the Muslims and India. There are ideological debates about whether we should call them “Muslims” or “Turks” (or Afghans, or whatnot). Ultimately I’m not invested in these debates. It is clear that whether someone was a “good Muslim” or not, since the rise of meta-ethnic religions during the Axial Age, these identities are important in some way for individuals who espouse them. I believe the Arabs constructed the Islamic religion in part as a response to likely assimilative pressures in the Near East.
Sometimes they do result in the sort of genocides that we associate with targeted ethnic cleansing. It is clear that the Frankish Christians who arrived for the first few Crusades killed urban Muslims and Jews in Palestine as a matter of religious commitment. They also encouraged the arrival of whole communities of peasants and artisans, who migrated to Palestine. For a few centuries, these people recreated the social structure of Western Europe in the Near East.
But, after the initial conquests the Christian rulers of Palestine took a far less ideological view because ideological decisions were impractical. Muslim peasants were sources of revenue, and some practices in the Near East were functionally adaptive. New migrants were often shocked at the assimilation, but a synthetic social order sprung up. We can only glean this from historical documents because the eventual expulsion of Christian elites from the Near East resulted in the disappearance of this culture.
So what do I think happened in India with the Turks? To understand this, we need to see what they did in the Balkans, in Iran, Egypt, and Eastern Europe. It is important to remember that the Turkic invasion of India is a piece of a broader dynamic after the year 1000 A.D. when Turkic migrations impacted almost all Eurasian societies outside of Southeast Asia.
Genetically, the Turks of Anatolia are only about ~10% East Asian. Assuming dilution that means Anatolian Turks are probably no more than ~20% descended from the Turkic pastoralists who moved into the region in the 11th century (in contrast, Rumelian Turks, like Kemal Ataturk, are almost certainly descended mostly from converts to Islam from Balkan peoples who Turkicized). Most of their ancestry is from people who spoke Greek, Armenian, and perhaps a form of Kurdish.
Similarly, the Chuvash Turkic people of Russia are genetically more like their Slavic neighbors than the Turks of Anatolia, though like the latter they also have a substantial minority East Asian component.
Unlike most Turkic people, the Chuvash, like the Yakuts of Siberia, are mostly Orthodox Christians. This is due to the fact that on the whole when the Turkic peoples shifted from shamanism to a “world religion” they selected that from the peoples whom they were in contact with, and often engaged in a predatory extractive relationship with. Before the rise of Islam, some Turkic people espoused Persian Christianity and Zoroastrianism in Turan. The Turkic people of western Mongolia during the life of Genghis Khan were nominally Christians of the Church of the East.
As far as Turks and India, it begins with Mahmud of Ghazni. He was a complicated figure. Though Indians are aware of him in large part due to the attack on Somnath, he was a major patron of culture, in particular, al-Beruni and Ferdowsi. Though Turkic slave soldiers came to prominence in the Islamic work under al-Mu’tasim, and Mahmud of Ghazni was from this general class of people, Turkic slaves converted to Islam who nevertheless remained subordinate in many ways to Arab and Iranian culture (e.g., see above the patronage of Ferdowsi, who produced a work valorizing pre-Islamic Iran in the form of the Shahnameh). But it was the period around 1000 AD which saw the emergence of Turkic polities which were fully Islamicized in the form of Kara-Khanids, and, independent of Arab and Iranian polities.
The Turkic “sword” of Islam during this period is clear. Though the Mamluks of Egypt included many Circassians and Georgians amongst their number, their internal lingua franca was a Turkic dialect. Though the Safavids of Iran had Kurdish, Georgian, and Greek, ancestry, their essential presentation was as a Turkic military order.
Some would attempt to dismiss the Muslim character of the Turks of Islam. This is fundamentally wrongheaded. Though the Turks may not have been “good Muslims” (whatever that means), and, their own ethnic-tribal identities may have been very salient, their own self-conception as ghazis and Muslims is very clear. In a similar manner, many of the Western Christian warlords that engaged in warfare in the Baltic and Islamic world on religious grounds may have been barely Christianized, and often concerned more with the material than spiritual conquest, but they clearly saw themselves in some ways as furthering the ends of Christian civilization.
The attempt to secularize Turks makes as much sense as it does to secularize the First Crusade. Material conditions matter, but they are not the only conditions that matter, and if materialism is one’s only concern then much of history would not occur.
How best to understand Muslim engagement with India then?
First, material considerations are not irrelevant. In all places, the Turks went they extracted resources and wealth. Whether it be the indirect Tatar yoke, the direct imposition of Turkic rule on non-Turks, as in Egypt and Iran, without ethnic assimilation, or, Turkic rule and assimilation as in Anatolia.
The reason that extensive targeted and conscious genocide in India is something I am skeptical of is that it is irrational. Turks could have inflected genocidal consequences simply by disrupting local production (e.g., if farmers can’t get to their land and the crops fail and famine ensues). But ultimately the wealth they wished to plunder consisted of the peasants!
But, this does not mean that there couldn’t have been targeted ideologically motivated attacks. The conquest of Italy by the Lombards in the late 6th century resulted in the disappearance of the Roman gentry across much of the peninsula. They were replaced by Germans at the top of the local status hierarchy, above Roman peasants. This was a transition from Catholic to Arian.
In North Africa, the Vandals and Alans replaced the Roman aristocracy in the eastern portion of their territory (modern-day northern Tunisia) but allowed the local structures to remain in place in the west (coastal Algeria). We know this particular detail because the Byzantine armies which conquered North Africa in the 6th century found Catholic Christian elites in the west but not in the east (ergo, elites had to be imported from elsewhere to administer the territory). The Arian Christian Vandal German and Alan population disappeared, as the women were given to East Roman soldiers as wives, and the men were enrolled in Roman armies which were situated on the eastern frontier against Persia.
My assertion is that the existence of Hinduism as a non-Islamic system which remains dominant within the Indian subcontinent is mostly a function of the fact that non-Islamic elites persisted and survived in the subcontinent. In some cases, as with Rajputs and others, Hindu elites were integrated into the Turko-Islamic sociopolitical order as sub-elites. In other cases, such as the Zoroastrian kingdoms of northern Iran during the Abbasid period, or the Nubian Christian kingdoms south of Egypt, or Russian Orthodox principalities under the Tatar yoke, polities which espoused a non-Islamic religious and cultural ideas maintained themselves. In India, like Russia, these non-Islamic states eventually maintained themselves well enough to rollback the Turkic-Islamic tide (in fact, a substantial portion of pre-Communist service elites descended from Christianized “Tatar” nobility, whose martial skills were put in the service of Cossack brigades which eventually conquered Siberia).
In The Rise of Western Christendom Peter Brown asserts that the Islamicization of the Near East occurred much faster once various forms of Syriac declined, to be replaced by Arabic. That is, once the ethnic difference between Christians and Arab Muslims was diminished, the conversions of subordinated Christians to Islam proceeded much faster.
Though Iranians were part of the story of Islam from the beginning, the mainstreaming of explicit Iranian culture into Islam can be dated to the late Abbasid period. The problem is that even in the early Abbasid period, Iranians remained predominantly non-Muslim. The defeat of the last independent Iranian Zoroastrian principalities and the conversion of the rural gentry is probably what resulted in the likely majority position of Islam within Iran around 1000 AD.
Curiously, many scholars have asserted that Islamicization proceeded faster in Turan, that is, north of Iran proper, albeit dominated by Iranic peoples. As per Peter Turchin’s argument, some of this may be a result of the fact that marchlands are generally more open to cultural innovation than cores (in the late Roman Empire, elites from the borderlands became Christian much earlier than those in the Roman core). But, another fact that is relevant is the Turan was more religious balanced it is identity than Iran proper. In Turan Eastern Christianity, Zoroastrianism, and Buddhism all had purchased. Islam may simply have been another option on the table, as opposed to a repudiation of Iranian identity, as may have been the case in Iran proper.
I engaged in this digression because the strong West and Central Asian orientation of the Turkic dominated conquest elites may actually have retarded the growth of Islam in South Asia. Though whole communities converted to Islam, and individual high-status converts were prominent, the differentiation between Hindu Indian and Muslim foreign may have prevented greater diffusion of the new elite religious cult. In Europe during the German “Drive to the East,” many Slavic pagans termed Christianity the “German religion.” The stubborn paganism of groups such as the Wends down into the 12th-century may partly have been due to the idea that conversion entailed alienation from their local identity. From becoming Wends into becoming Germans.
This post is written in response to comments below. On the one hand, is the temptation to argue in terms which leverage modern understandings to comprehend the past. This leads to confusions and misunderstandings. The religious skepticism of al-Ma’arri was tolerated and indulged because of he was a genius from an upper-class background in a society that was highly stratified. This does not mean that Muslims of the period were tolerant of atheism any more than they are today, but in that period mass society did not truly exist, and al-Ma’arri’s eccentricities were not perceived to be corrupting of the masses, as they would be today.
And yet similarly the past is not entirely incommensurable to the present. The religious-nationalist rebellion of Simon bar Kokbha is entirely comprehensible in modern terms. The sectarian conflict between Greeks and Jews in Alexandria that bubbled up from below in the wake of the Jewish rebellion is not hard to understand in terms of its social and psychological roots. Urban Greeks and Jews lived together in Alexandria for centuries, and their mores forced separation, despite the fact that Alexandria’s Jews had taken up Greek as their dominant language.
* Is it me or does “Muhammad of Ghor” sound like a barbarian warlord out of a Tanith Lee DAW novel from the 1970s?