The “clash of civilizations” is a thing, just not the only thing

A few days ago I put up a post, The “Islamic World” Was Not Invented By Europeans. Since then, I have been reading the author’s book, The Idea of the Muslim World: A Global Intellectual History.

It’s an interesting work with a lot of facts. Though so far no facts have been surprising to me, and, many facts were known to me. For example, the author talks about the reality that Muslims were subjects of the Russian Empire in the 18th and 19th centuries. I happen to have read a book about the topic. Specifically, how the ulema in the Russian Empire adjusted to rule under an Orthodox Christian monarch. The author mentions that Protestants fought with the Ottomans at Vienna, and exhibited a cool attitude toward Indian Muslim nationalism against the British. Both of these facts, I knew.

The basic thesis seems to be similar to what I had inferred earlier: that the idea of a unitary Muslim world is a reaction to the European colonial experience, and not deeply rooted. The problem is that a lot of these assertions hinge on semantic interpretations. What does “unitary Muslim world” mean for example? The author, Cemil Ayden, seems to also suggest that both the “West” and the “Muslim world” are modern constructions. And they are. That does not mean these modern constructions don’t build upon and extend pre-modern self-conceptualizations which are very important. We can walk and chew gum at the same time.

Reading Ayden reminds me of encountering Bryan Catlos’ work on Muslim Spain years ago. Catlos’ publicity people at the university press tried to make it as if he was arguing that the line between Muslim and Christian was incredibly fluid and that his work refuted the “clash of civilizations.” But anyone, which includes me, who is aware of the large numbers of conversions of Christians to Islam and then back to Christianity, not to mention the Jews, knows that the categories are a bit more complex than the modern cartoon. Nevertheless, nothing in Catlos’ scholarship refutes the reality that religious identity was a critical, and perhaps the most important, building block of self-conceptualization in medieval Iberia.

One way to avoid the baggage around the word civilization is to rename it a “meta-ethnic” identity, as Peter Turchin does. A meta-ethnic identity allows people from different tribes and ethnicities to unite around something greater. Often, though not always, it is religion. The initial decades of the rise of Islam are complicated by the possibility that the religion wasn’t a meta-ethnic identity, but rather a tribal cult specific to a group of Arabs. This was not sustainable if Muslims were to maintain a multi-ethnic polity. Like the Mongols, they would have been absorbed by those whom they conquered. The rise of the Abbasids around 750 is often characterized as the revenge of the convert peoples, with Iranians in especially prominent in the early years of the dynasty.

Something similar happened with Christianity, which in its early centuries was fundamentally a Roman religion in Western Europe. Eventually, the expansion of the commonwealth of European kingdoms in the early medieval period occurred through the expansion of the Roman religion, and its transformation into something that was post-ethnic (during the medieval period in parts of pagan Eastern Europe Christianity was considered a “German” religion!).

There is certainly something commendable in Ayden’s work in situating current geopolitical tensions and alignments with their early modern precursors. But to the naive these arguments often erase the real deep roots of these configurations and their durability across the millennia. For example, I have stated, justifiably I think, that modern Iran was fundamentally and essentially shaped by the Safavid transformation of the region in the 16th century. That is, unifying the various Iranian and Turkic peoples in present-day Iran under the banner of Twelver Shia religion. But this is not to deny the reality that elements of Persian national self-conception predate the Safavids by thousands of years!

To bring it back to conflict, Christian cities such as Amalfi in southern Italy, often aligned themselves with Muslim pirates and corsairs in the first few centuries Islam. This does not mean that Amalfi was not Christian. Or that the distinction between Christianity and Islam meant nothing. Amalfi came under sharp criticism from Christian polities for its pragmatic alliances with Muslims. Similarly, France’s traditional friendly relations with the Turks due to the common Habsburg enemy came under criticism during the second Ottoman siege of Vienna.

Because of profit or in the exigencies of the moment, strange bedfellows often emerge. The Hungarian Protestants that marched with the Ottomans against the Habsburgs were marching for their cultural survival. The Habsburgs were suppressing and slowly extinguishing the Reformed movement in Hungary and had been doing so for decades. Hungarian Protestantism persisted only under Ottoman protection. This does not mean that Hungarian Protestants are not aligned with Christianity. But before their civilizational commitment could come into play, they had to safeguard their existence, which forced them into making a decision to march with the armies they had and not the ones they would have wished (Viktor Orban is a Hungarian Protestant).

What quantitative scholars like Turchin, and Azar Gat in War in Human Civilization, have shown is that conflicts across meta-ethnic or civilizational boundaries tend to be particularly brutal and characterized by the dehumanization of the enemy. On average. The fact that most Christian states in the pre-modern world bordered on Christian states means that most conflicts would occur between Christian states. But the conflicts at the civilizational boundary would be characterized by more extreme levels of brutality, coercion, and a lack of chivalry.

One might see in most conflicts that they occur within meta-ethnic groups, or that in a large number of cases the alliances cross meta-ethnic identities. For example, Pakistan today is under the grip of Sinophilia, despite China’s objective reality that it is an anti-Islamic state which oppresses Muslims, and Pakistan’s objective Islamic extremism. The fact on the ground currently though is that Pakistan as a nation-state benefits much more from being pro-China in its rivalry with India then rejecting Chinese entreaties on principle due to meta-ethnic solidarity with China’s Muslims. The pragmatic aspect of this alliance does not negate the reality that Pakistanis are sincere Muslims who have strong commitments to a trans-national Islamic identity, as evidenced by the fact that Pakistanis are often represented in trans-national Muslim movements.

Anyone who has read my thoughts knows I reject the idea that religions have fundamental clear and distinct essences. Religions are what people believe they are. What people practice. But people with particular confessions exhibit more solidity in their understanding of group identity than most post-colonial treatments seem to allow. Islam and Islamic identity do not exist only in contrast with Western Christians. In the east Islam interfaces with Indian traditions, Buddhism, and Zoroastrianism. Across all these interactions Muslims have a certain sense of self as distinct and can grade differences between various out-groups (e.g., Christians are not clearly idolaters, Jews are clearly monotheists, and Buddhists are idolaters).

It is simply a fact that post-colonial peoples had a pre-colonial history, and that pre-colonial history is just as important in their self-understanding as the post-colonial one.

Crescent over the North Sea

Pew has a nice new report up, Europe’s Growing Muslim Population. Though it is important to read the whole thing, including the methods.

I laugh when people take projections of the year 2100 seriously. That’s because we don’t have a good sense of what might occur over 70+ years (read social and demographic projections from the 1940s and you’ll understand what I mean). Thirty years though is different. In the year 2050 children born today, such as my youngest son, will be entering the peak of their powers.

First, one has to note that these statistics include a lot of people who are what some would term “Muslimish”. That is, they are not religious believers, but have some identification with Muslim culture. That’s explicitly noted in the methods.

The problem with this is that there is a wide range of religious commitment and identification across Europe’s Muslim communities. On the whole, they are more religiously observant than non-Muslims in their nations of residence,  but, for example, British Muslims are consistently more religious than French Muslims on surveys (or express views constant with greater religious conservatism).

Here are the results of a 2006 survey:

  France Britain Germany
Yes, Westerners are respectful of women 77 49 73
Yes, there is a conflict between being devout Muslim and living in modern society 28 47 36
Yes, sometimes violence against civilian targets in order to defend Islam can be justified 16 15 7
Did Arabs carry out 9/11? (yes) 48 17 35
People in Western countries are selfish (yes) 51 67 57
People in Western countries are arrogant (yes) 45 64 48
People in Western countries are violent (yes) 29 52 34
Do you consider yourself Muslim first? (yes) 46 81 66
In my country Muslims are perceived to adopt customs of nation 78 41 30
     

Numbers such as those above indicate even if France and the United Kingdom both have Muslim minorities on the order of 17% of the population, the nature of those populations differs to such an extent that that similarity in value may mislead.

In God’s Continent Philip Jenkins observes that public statistics of Christians often work to exclude cultural Christians, but those of Muslims include cultural Muslims. What many estimates of “Muslims” in the European context do is give a sense of the proportion of the population which is of Muslim background. This is especially true in a nation like France where religious survey data is not collected by government agencies.

Overall I think this data is important to consider, but there’s nothing really new in a qualitative sense. And, it is important to keep in mind the details. It is highly probable that the idea of a European superstate will have faltered by 2050, and each nation will its own Muslim minority, and engage with them differently depending on local values and context. Though Muslims, broadly construed, will form about the same proportion of the French and British general population, I suspect that in Britain the distinction between Muslim and non-Muslim will be much more obvious and strict than in France.

The question should be “Who Are Salafi Muslims and Why Are Many So Extreme?”

Because of the horrible massacre at a mosque with Sufi tendencies in Egypt, there are a lot of “explainers” out there about sectarian divisions in Islam. The one in The New York Times, Who Are Sufi Muslims and Why Do Some Extremists Hate Them? could be worse. This portion especially gets at the major issue:

For a time, beginning in the 12th century, Sufism was a mainstay of the social order for Islamic civilization, and since that time it has spread throughout the Muslim world, and to China, West Africa and the United States. As Sufism spread, it adapted elements of local culture and belief, making it a popular practice.

Alexander D. Knysh, a professor of Islamic studies at the University of Michigan and expert in modern Sufism, describes it as a “very wide, amorphous movement” practiced within both the Sunni and Shiite traditions.

Specific claims about Sufism beyond the most general fail because vast swaths of Islamic history and Muslim peoples and practice are Sufi. In the modern western media, there is an unfortunate tendency to dichotomize Islam into a harsh and fundamentalist form and a moderate and mystical Sufi variety. Though a small minority of Sufis have drifted toward very heterodox beliefs, the vast majority are orthodox Muslims who also adhere to a school of Islamic law.

And Sufis are not all pacific saints. In the 19th century Libya the Sensussi Sufi movement was critical in the continuation of the trans-Saharan slave trade, and later served as a major focal point for violent resistance against the Italian colonial project. The great anti-philosopher of the medieval period, Al-Ghazali, who is generally agreed to have ushered in the decline of philosophical thinking within orthodox Sunni Islam, was a Sufi.

The question should not be about Sufis. Sufis are not moderate or mystical Muslims, they are simply Muslims. That is, they’re the mainstream. Rather, the crux of the issue is that violent radicals have emerged from the soil of Salafism. Not all Salafis are violent. But violent Salafis are the ones who regularly target other Muslims and their holiest of sites.

Salafism is a modern movement of the past few centuries. Like Protestant Fundamentalism, it is a product of the engagement of traditional religion with the modern world. Self-consciously Salafist Muslims have never known a world where the West was not dominant. Therefore it is no surprise that they look to the accrued tradition of Islamic civilization and see in it failure and decay.

Like some Radical Protestants, the Salafists imagine that they are creating a community of Muslims who are true to the path of the religion in its earliest years before it became tainted with monarchy. Basically, Salafists wish to transform Islam from a religion of history to one of pure axiomatic abstractions.

Why do Salafi radicals attack Sufis? Their tendency to engage in takfir against other Muslims goes back to the proto-Salafi Wahhabists. And Sufi Islam, with a venerable history going back more than 1,000 years, is naturally going to be the target of Salafi rage because it was the Islam that failed to stem the tide of Western ascendancy, the Islam that witnessed the slow and gradual decline from the greatness of the 8th and 9th centuries. The children shall eat their own parents.

More “orthodox” Islam in Indonesia is inevitable


Curfews, Obligatory Prayers, Whippings: Hard-Line Islam Emerges in Indonesia:

In the Indonesian market town of Cianjur, new rules require government workers to clock in with their thumb prints at a downtown mosque to confirm attendance at morning prayers. That’s on the order of district chief Irvan Rivano Muchtar, who also wants a 10 p.m. curfew for the town and is sending police to stop teenage girls and boys hanging out without parental supervision.

One one of the first things I wrote on the internet in 2002 was about Indonesian Islam (on a blog platform which is now gone). The reason for me writing on that topic was that the media representations of Southeast Asian Islam in the wake of 9/11 seemed excessively simple and reductive. For the West Indonesian Islam is often asserted to be moderate, and a counterpoint to the intolerance and exclusion which is the norm in the Middle East. In other words, Indonesia as the world’s most populous Muslim majority name plays a specific role in a broader narrative. A bit part in the grand narrative of moderation and radicalism. In the process, the textured uniqueness of Indonesia itself often gets lost.

First, let’s take a step back and frame the history of maritime Southeast Asia, what eventually became Malaysia, Indonesia, and the Philippines. It is often stated that in Indonesia Islam spread peacefully through trade. This is supposedly in contrast to what occurred in the Middle East or South Asia, where military force was the dominant theme of Islamization. Superficially this is true.

But the reality is that forced conversion was likely a marginal phenomenon at any given time, and especially during the early centuries of Islam. The prominence of men such as Timothy, Patriarch of the Church of the East, under the Abbasids attests to the power of the non-Muslim majority (there was a similar eminence for the Zoroastrian community). But the conversion of Malays around Malacca in the early 15th century after the conversion of the king is not quite as different from what occurred in Persia after the Arabs arrived as one might think. The rapid shift of the Iranian nobility to Islam in provinces under Arab control seems to have triggered a gradual change among the masses (those regions, such as Tabaristan, where local elites maintained the old religion resisted Islamization until they were conquered and converted). It was not a matter of the sword or conversion.

Of course, religious wars were a necessary part of the expansion of Islam in the Middle East and South Asia, even if they did not effect much of the religious change which occurred later. But this is not a qualitative contrast with the Middle East and South Asia in comparison to Indonesia, though it is a quantitative one. Some polities, such as Malacca and Aceh, came to Islam through their integration into the maritime mercantile network of Muslims from Arabia to China. Centuries later the collapse of the Javanese kingdom of Majapahit, whose earlier hegemony served as some of the basis of broader Javanese-Indonesian claims to the whole archipelago, occurred due to attacks from Muslim sultanates who organized in part on the basis of religion. Majapahit fell because of jihad. It was not converted peacefully.

In sum, the history of what became Indonesia and its relationship to Islam is different from that of South Asia, or the Middle East, but that difference is one of degree, not kind. Second, one must also distinguish between Java and the rest of the archipelago. In the article above there is an reference to Aceh, a province where the practice of Islam aligns very strongly with that found in the Middle East. But Aceh is also culturally and historically very different from Java, Islam came to Aceh a the dominant religion at least two centuries before it did in Java, and probably earlier. On the map above Aceh is also geographically rather distant from the central islands, at the far northwest tip of Sumatra.

To a first approximation, orthodox Islam is a much more salient and central aspect of the identities of people from outlying islands than it is to Javanese. The complex indigenous-Hindu-Buddhist-Islamic synthesis which is said to characterize traditional Indonesian Islamic culture is actually a feature most evidently of Javanese culture. Parts of far western Indonesia came to Islam earlier, and integrated into the Muslim cultures of the Indian Ocean more thoroughly, so that earlier Buddhist affinities faded over time (Aceh at one point was aided by the Ottomans in fighting the Portuguese). In contrast, parts of eastern Indonesia, such as Sulawesi, were Islamicized after the fall of Majapahit, but the impact of Indic culture had been relatively superficial (though not trivial, as Indic influence is evidence as far east and north as the islands which became the Philippines).

At nearly 40% of the population of all of Indonesia the Javanese loom large in the identity of the nation-state. Most of the presidents of Indonesia have been Javanese (B. J. Habibie was raised in Sulawesi, where his father was from, but his mother was Javanese; Sukarno’s mother was Balinese, while his daughter’s mother was Sumatran, though both seem to have identified culturally as Javanese). Many of the things people say about Indonesian Islam are really about Javanese Islam, with the model implicitly derived clearly from Clifford Geertz’s tripartite division between santri, abangan and priyayi.

The santri are basically what we define as world normative orthodox Muslims. The priyayi are the Javanese aristocracy, who self-consciously explored mystical concepts and practices with an extra-Islamic origin. But the vast majority of Javanese are the abangan, rural peasants who practice an Islam which emphasizes custom and tradition as much as sharia. Because custom and tradition have deep organic roots within Java they naturally include many elements which are ‘pre-Islamic.’ In Java both the Mahabharata and Ramayana are still part of the living culture, for example.

It strikes me that the attitude of the Javanese may have analogs with that of the Persians in relation to their cultural history. By and large like the Persians the Javanese are Muslims without apology.* But like the Persians the Javanese take pride in a history before Islam, in particular Majapahit, whose writ tentatively spanned most of contemporary Indonesia. And Majapahit can not be separated from a Hindu-Buddhist synthesis which left massive cultural artifacts such as the Borbobudur temple complex (and the modern Balinese also serve as continuous cultural links with the Hinduism of Majapahit).

But the economic and social development of Java will naturally lead to a waxing in the santri tendency. Orthodox Muslims among the Javanese have not been part of the underclass, but rather outward facing portions of the traditional mercantile class or middle class urbanites. Santri Islam is portable, and commensurable with international Islam. Abangan Islam is rooted in the rural landscape of Java, and urbanization will inevitably erode its hold on future generations. Meanwhile, priyayi practices are structurally limited to a narrow class of elites.

Overall then the rise of ‘conservative Islam’ in Indonesia is a complex story with two primary threads. One is regionalism. The regulations introduced in the story above are in Cianjur, in western Java. This area is more Islamic than central or eastern Java, and the native people are not Javanese, but Sundanese.

As local identities were given more freedom of play after the New Order in the late 1990s it was reasonable to expect that more strikingly Islamic practices would become more public, as they were dampened earlier by the dominant Javanese orientation of the Suharto regime. Second, modernization within Javanese culture itself will likely lead to the emergence of a more numerous group of sharia compliant and world Islam oriented group of Muslims, as they can not rely upon community and adat in an urban landscape remote from their backgrounds of origin.

This is not to say that the standard chestnuts about Saudi funding are not important. But it is important to note that portions of Indonesian Islam have long been deeply connected to the Muslims of the Arabian Sea; this is not a function simply of the rise of petro-states, though their wealth has certainly allowed them to put their thumbs on the scale. Maritime Southeast Asia is the eastern segment of what is operationally a Shafi international of Sunni Muslims who ring much of the Indian Ocean. As Indonesia becomes globalized, it will gravitate to other nodes within the international network which it already has long-standing connections. This is probably inevitable in some ways, and the working out of the reality of contemporary Indonesian pluralism has to face the inevitable tensions that modernization will bring. A more universal and non-local Islam will probably also be more exclusive and culturally muscular.

* A minority are Christian or Hindu. A Hindu Javanese kingdom persisted in the east of the island until the 18th century.

Arab Islamic science was not Arab Muslim

Someone stupid who follows me on Twitter said “It seems @razibkhan forgot the Arabs gave us algebra and many other scientific/mathematical advances.” The history of algebra is actually somewhat more antique than the Arabs, as outlined in Unknown Quantity: A Real and Imaginary History of Algebra. But the origin of the word is Arabic. From Wikipedia:

The word algebra comes from the Arabic الجبر (al-jabr lit. “the reunion of broken parts”) from the title of the book Ilm al-jabr wa’l-muḳābala by Persian mathematician and astronomer al-Khwarizmi.

Though Wikipedia says that al-Khwarizmi is Persian, it is more accurate to say he was Iranian-speaking, because as his name attests his origins were in Khwarezm, which is in Central Asia. It is accurate to say that Arab Islamic civilization was intellectually productive in the centuries before 1000 A.D., but it is not accurate to say that Muslim Arab scholars were responsible for this.

A huge number of these scholars were not ethnic Arabs. In the early years a substantial number were not Muslim. Though it is often said that many were Persian, as recounted in Lost Enlightenment many of the “Persians” were not from Persia proper, but from Iranian regions of Central Asia which over the centuries have now become Turkified.

Why is this important? The multicultural nature of early Muslim (and later Ottoman) polities might inform us as to the future of this sort of diverse society. Second, it’s preferable that you don’t seem like an idiot if you want me to listen to anything you say.

Islam in China is not one

Over the past few days I have seen articles in the media which refer to “Chinese Muslims,” and then make such a casual and slight distinction between Muslims in China and the Uyghur ethnic group that I think it’s really misleading to the general public (e.g., Anti-Muslim sentiment is on the rise in China. We found that the Internet fuels — and fights — this).

To review, Muslims in China are multi-ethnic. The two largest groups, the Hui and Uyghurs, comprise nearly 90% of Chinese Muslims. There are marginally more Hui than Uyghurs.

Who are the Hui? The Chinese government defines Hui as an ethnic group, but really they are differentiated by their adherence to Islam. Hui speak a dialect of Chinese specific to their locality. They do not have a “Hui language.” Physically they resemble the Han. Because of their long period of isolation in China after the collapse of the Yuan dynasty, the Hui have gone through several indigenizing phases.In the 18th century in eastern China the Hui intellectual classes synthesized Chinese cultural frameworks with Islam in a fascinating manner. The whole project is recounted in The Dao of Muhammad.

These periods of Sinicization are often followed a reformist globalist revival triggered by missionaries or those who went on pilgrimage to Mecca. Islam with Chinese characterizes recedes for a generation or two, only to come back.

In the 19th century to a great extent the project of social accommodation with the Han by the Hui collapsed in the face of social disorder, anti-Muslim policies by the Manchus, and reformist movements inspired by broader currents in the Islamic world. Though the Hui are a very small minority, unlike the Han a military career was not low status for them, so they “punched above” their weight.

In the 20th and 21st centuries the Hui have been relatively quiescent. Why? There are numerous reasons, but it is important to emphasize that there are many strong contrasts with how the Hui are treated and perceived, and how they perceive China, in relation to  what is meted out to the Uyghurs. The Hui are no less Muslim than Uyghurs, but they are not the political and social problem in China that Uyghurs are.

Though the Chinese state defines Hui as one of the minority “nationalities,” that is really a semantic obfuscation. The Hui are most easily conceptualized as Han Muslims, even though some of their customs separate them very strongly from the Han (e.g., no consumption of pork), and traditionally Han identity has been seen as exclusive from an Islamic identity. That is, a Han who converts to Islam becomes a Hui by definition.

Though in a Chinese context one could never call the Hui “Han Muslims,” from a non-Chinese perspective it is very informative of the relationship and difference of the Hui from the Han, as opposed to the Uyghur from the Han.

Two Uyghur men

Obviously the Uyghur are not Han, they are Turkic. Uyghur nationalists have pan-Turkic associations, and many Uyghurs live in Turkey. As a Turkic people Uyghurs, unlike Hui, do not speak Chinese as their first language. The attempt to educate Uyghur children in Mandarin Chinese to enable them to assimilate and succeed economically has faced resistance because Uyghurs see in this the first steps to assimilation and eventually alienation.

Though Hui are very distinctive in China proper, and live in their own segregated areas in much of the north (in southern China this is less common, and Hui assimilation into Han identity has also been widespread), they are still part of the Chinese landscape. Muslims have lived in China proper since the Tang dynasty, 1,300 years ago. Large numbers of Muslims arrived with the Mongols 800 years ago, and many stayed on. As a minority in a non-Muslim society these people had to navigate how to be both Chinese and Muslim, when much of Chinese identity deviated from world normative Islam in deep ways.

The Uyghurs did not go through any of this because they were not part of China until the 20th century. Though Chinese garrisons and hegemony did exist in Xinjiang during portions of the Han and Tang dynasty, up until the Manchu conquest of these territories in the mid-18th century the Uyghurs had not been part of the same political unit with Han Chinese for over 1,000 years, with the exception of the short Mongol interlude. In fact, the ethnogenesis of modern Uyghurs, as a blending of Turkic migrants from the north and native Indo-Eurpean speaking groups in the Tarim basin, was concurrent with the collapse of Chinese influence in what became the eastern edge of the Turkic world.

Notice I was very specific in saying that they became part of the same political unit with Han Chinese in the middle of the 18th century. This because outside of China proper the Manchu emperors did not necessarily rule as Chinese potentates. Rather, they took on different forms for their different subject peoples, and the conquest of the heart of Eurasia was not a conquest by the state of China, but of the Manchu ruling caste. Any attempts to Sinicize Xinjiang came later, and were halting at best. While Chinese speaking Muslims in Beijing were theorizing how Muhammad actually completed the Confucian vision better than most Chinese, the Uyghurs simply swapped the rule of nearby Tibetan Buddhist Oirat Mongols for a distant Manchu ruler, who was also sympathetic to Tibetan Buddhist religion and claimed a kinship with the Mongols through descent from Genghis Khan’s younger brother.

The problem that the modern Chinese state has is that it rejects the feudal multicultural compromises of the imperial past. Though Communist regimes pay lip service to national self-determination, the reality in Communist regimes has always been that the party has enforced a normative ethnic identity as one that is aspirational for minorities. The Chinese state suppression of the religion of the Uyghurs, the promotion of Mandarin, the encouragement of migration to Xinjiang by Han, and even inducements in some cases for Uyghurs to intermarry with Han, are all part of a general pattern of activity which will result in the assimilation of the Uyghur nation.

It is apparently a fact that while Islamic belief and practice by Uyghurs is sharply frowned upon by Han authorities in Xinjiang, in most of China proper Hui religiosity is relatively tolerated. Hui are even seen as appropriate ambassadors to Muslim nations for purposes of diplomacy and business, because they show how China can accommodate Islam. Unlike the Uyghurs the Hui do not have a geographical region where they are dominant (Muslims are 35% of the population of the small province set aside for Hui). Their national home is China. Additionally, obviously they would not resist Mandarin Chinese instruction, because they are already Chinese speakers. Unlike the Uyghur, who have substantial West Eurasian ancestry, the Hui are also physically no different from Han.

In Central Asia the Hui have a different name. They are called Dungans. And traditionally they have been overrepresented among soldiers and merchants from China. Within China the Hui are exotic and somewhat out of place due to their religion. But in Central Asia the Hui are exotic and somewhat out of place due to their Chineseness. Hui were important in keeping Xinjiang in the Manchu fold after the conquest. Many Uyghurs know this history of cooperation between Han and Hui. In the 2009 Urumqi riots the Uyghurs reportedly chanted “Kill the Han, kill the Hui”.

None of this is to deny that Islam presents challenges as a minority religion within a non-Muslim nation. The Hui rebellions of the 19th century, and periodic flare ups between Hui and Han in the Chinese heartland, attest to this. But differences between Uyghurs and Hui illustrate that excessive focus on Islam misses that Uyghur violence in response to Chinese coercion likely has multiple causes. Islam over the last generation has been the most powerful binding ideology for national resistance among Uyghurs. But it would be far less relevant if the Uyghurs were not a nation in the first place, which they are.

Another way to say it is that Tibet and Xinjiang have many of the same underlying parameters as to why they are hotbeds of ethnic tension and separatism. Religion is part of the story.

Related: Islam in China Revisited.