Author: Razib Khan

The religion of Hindus before Hinduism

Razib Khan · September 13, 2020 · 37 comments

India in the Persianate Age: 1000–1765 is a good read and I recommend it. But the author, Richard Eaton, is not a guru or a sheikh, and should not be taken as such.

A comment Eaton makes offhand several times is that the conflict between Turks and Indians should not be understood in confessional terms. This is a commonly asserted, and on some level, it reflects elements of the truth. Hindu Rajputs served under Muslims, and Turkic soldiers served under Hindus. You can’t reduce everything to confession.

But, it is clear that confession and civilizational identity did exist, and it was robust. Going from the specific to the general.

  1. A great deal of text given over to Man Singh’s glorification of his conquests as an Indian warrior, and his patronage of Indian religion, in particular Vaishnavism.
  2. Eaton highlights the rapid Indianization of practices and hegemonic motifs present among the Turks and Afghans who were born and raised in India. And yet despite the syncretistic tendencies which occurred, ultimately these ashraf elites remained identified as Muslims and often were pulled back to world-normative Islam over the generations.
  3. Vijayanagara persisted as a Hindu polity for three centuries. The cross-cultural analysis shows that recalcitrant pagan powers always convert to the religion of their enemies eventually. The leader of the pagan resistance in Saxony became a Christian. Pagan resistance to Christianity in Sweden, Lithuania, and ancient Rome were only temporary, as resistant lineages eventually were assimilated into the new order. Resistance to Buddhism in Japan and Tibet was initially violent, but futile. In iterative games, paganism is the eternal ‘beatable’ strategy.

The only point to posting this is that there is a common assertion that Hinduism as a religion or identity only emerged in the 19th century. I am now convinced that this confuses the name of the phenomenon for the phenomenon. The Indian religion of the Hindus was clearly bundled together in a way that allowed for their elite deployment as a meta-ethnic identity that separated them from the Turks and Afghans who ruled them. Similarly, the Islam of the Turks and Afghans (and variegated Ethiopians, Arabs, and Persians), separated them from the Indians whom they conquered to prevent full assimilation as an Indian elite with popular roots in early modernity.

There is a major issue where our conception of religion qua religion is conditioned on an intellectual revolution rooted in the Second Reformation of the Calvinists. But, I think it is important not to get carried away with this construct, and assert that Calvinist religion is qualitatively different from pre-Calvinist religion. I don’t think it is. Rather, it simply shifts some of the parameter values within the model. Similarly, the identity of a coherent Hindu Rashtra with a post-caste socio-religious identity is an invention of modernity, but its roots are ancient and indigenous, and not postcolonial fictions.

Charles Cameron, R.I.P.

Razib Khan · September 10, 2020 · 8 comments



Charles Cameron, a contributor to this weblog, and an early guest on our podcast has died after an illness. I do not know any details (I saw a Facebook post), but I felt that it was important to mention his passing since his contribution to this forum was appreciated, and from what I knew of him he seemed a man with a great heart and boundless curiosity.

Death is inevitable, and a part of life. We too shall pass in our own time. All we can hope for in this one life is to leave memories that honor what we stood for after we leave.

Islam on the fringes

Razib Khan · September 6, 2020 · 15 comments

Decided to start reading India in the Persianate Age: 1000–1765. It’s a fast and easy read (and, it’s an affordable book for a nonspecialist like me who only spends $100 or more on genetics texts). I jumped ahead in a few passages, which seem to be adaptations from papers I’ve read from the author before.

In relation to discussions on this weblog and comments I’ve made:

1) The author describes islands and pockets of wholly Muslim peasants in eastern Bengal observed by the Mughals in the 16th century. These Mughals indicate that these people are not truly Hindu or Muslim, at least initially. This is line the common thesis that Islamicization is a function of the weak to nonexistent integration of these frontier peasants into Indian culture. Mughal observers also note the physical appearance of these people: small, dark-skinned, and beardless. This seems entirely accurate (I cannot grow a beard!).

The author observes that tax receipts over several decades in Bengal exhibited a pattern that is suggestive of a massive population increase in the east and stability in the west. In fact, there were two-fold increases in the east at the same time that districts in the west declined by 10%.

The thesis by the author, which seems broadly creditable, though not proven, is that the transition to high-intensity rice agriculture in the plains of the eastern delta, and in particular east of the Padma, arrived after the Mughal expansion into the reason. As such, the transition to a “higher religion” occurred under Islam, and therefore these people became Muslim (at least nominally). The analogy that is obvious here would be the Christianization of the Montagnards and other Southeast Asian “hill people” during the colonial period, as they had been detached from Theravada Buddhist civilization.

2) The author deploys a similar model, with modifications, for western Punjab. The model here is that Jat who moved up from Sindh abandoned obligate pastoralism and engaged in agriculture, and fixated upon the tombs and shrines of Islamic eminences. Due to the emphasis on paternal lineage, the author observes that the Islamicization of names occurred quantitatively over 300 years, from the 15th to the 18th century (initiation to completion).

The main qualm I have with this model is this: the Jats/people of Punjab do not look to be from Sindh if modern Sindhis are representative of ancient Sindhis. In the language familiar to readers of this weblog Sindhis are enriched for “Iranian-related ancestry” vis-a-vis Punjabis, who are enriched for “steppe.” The Jats in particular are highly enriched for “steppe” ancestry. Going by genetics alone this model is difficult to accept, though as noted in this space it does seem that caste-like stratification has ancient roots in this region, and so that may serve as some explanation.

Note: The model in Bengal implies that Islamicization and Hinduicization occurred in East Bengal simultaneously. That is, most of the sponsors of agricultural reform and intensification happened to be Muslim, but some were Hindu, and in the case of Hindu sponsored developments the regions became Hindu. Further investigation though would be warranted for Bengali Brahmins in the eastern regions, who would then presumably be migrants from the western zone, because it is hard to credit animistic tribes needing the services of literate priestly elites.

Why do Pakistanis not want to be descended from Hindus?

Razib Khan · September 5, 2020 · 174 comments

First, I want to enter into the record that among Punjabis and Sindhis there is almost no West Asian ancestry in Pakistanis.* I qualify this with “almost” because there is some, particularly in Sindhis. You can tell because of African ancestry, which is distinctive in even small fractions, and which is found in some of the HGDP Sindhis. I haven’t checked the 1000 Genome samples from Lahore (which clearly includes Punjabis but also other ethnicities), but they seem “more Indian” than the HGDP Sindhis.

Most people with half a brain can see the above fact in the data. That being said there is some ideological battle between Pakistanis and Indians about the Hindu origins of Pakistanis. Or, should we say “Hindu”?

On both sides of my family, I have “caste Hindu” forebears within the last few centuries. My paternal grandmother’s father was born a Hindu. So I have no compunction in admitting that my ancestors were Hindu, and my genetics indicate a rather generic Bangladeshi ancestry except for the higher fraction of East Asian (my family is from what was Tippera). It helps I’m not Muslim or Muslim-identified.

From Hindu Nationalists there something of schizophrenia on the topic. On the one hand, they loudly proclaim the Hindu origins of South Asian Muslims (correct). Often, there is also an assertion that these are low caste converts (perhaps correct, but specious to the argument). But then, they flip to the assertion that South Asian Muslims are invaders, oppressors, etc.

It’s not totally coherent. Perhaps more coherent is the position of some Pakistanis: “we were never Hindus.” The argument is straightforward, and about ten years ago I was quite open to it. To be frank, I probably leaned toward the proposition that Hinduism as an identity makes no sense without a reaction to Islam and later the British-Christian experience. Though probably not as extreme as “real Hinduism didn’t exist in the 19th century”, I wouldn’t have laughed that assertion out of the house.

There are several reasons I reject or have evolved from my older views.

Continue reading Why do Pakistanis not want to be descended from Hindus?

Ancient Pakistanis were Hindu

Razib Khan · September 3, 2020 · 20 comments

Over at my other blog, Pakistani British Are Very Much Like Indians Genetically. The title doesn’t refer to genome-wide worldwide affinities. Rather, the preprint looks at British Pakistanis, and finds a pattern that is not going to surprise Indians: endogamy seems to have kicked in for these groups starting 1,500 to 2,000 years ago. This is exactly what you see in the Indian jati data. The similarity is pretty incredible, and to me is a strong rejection of the model that these groups were strongly anti-caste so on the margins of Indic civilization.

There is a second wave of endogamy though, dated from 150-500 years ago, roughly. I think this is likely Islamicization and adherence to cousin-marriage. These Pakistani groups seem to show the tendency of jati endogamy common among Hindus, and, cousin-marriage patterns of the Islamic world.

Finally, the reason I posted over on the other blog is that I think this might speak to the long-term trajectories of Bangladesh and Pakistan: Bangladesh is not in the same mold as Indo-Pak societies. The 1000 Genomes data indicate few runs of homozygosity and not much internal structure. That is, no jati endogamy, and, low levels of cousin-marriage.

If you believe Joe Henrich, this means good things for Bangladesh in the future… (vs. Pakistan)

(the Henrich podcast is already available for Patrons)

Open Thread – 08/29/2020 – Brown Pundits

Razib Khan · August 29, 2020 · 95 comments

The usual. Like I said I’m moderating more, so watch yourself.

On Monday some of the BP crew (and my friend Josiah Neeley of Urbane Cowboys will be talking to Matthew Yglesias of Vox about his new book One Billion Americans: The Case for Thinking Bigger. Open to questions from the peanut gallery here too. I’ll probably also ask non-book questions toward the end of the hour.

Why Turks ruled India for so long

Razib Khan · August 26, 2020 · 117 comments

After finishing Escape from Rome: The Failure of Empire and the Road to Prosperity I continue to believe that geography and economics explain the basic reason for the very long ascendance of Turkic people in the Indian subcontinent, and, their eventual eclipse.

The context for this is the fact that many Indian and Indian American friends have posited cultural rationales for the Turkic hegemony. That is, there was a passivity and disunity in Hindu society which made it vulnerable to the Turks, who were also adherents to a separatist hegenomic ideology in the form of Musim. All this may very well be true, but I have always held that the key factor was that Turks and Muslims more general had ready and easy access to warhorses.

Between 500 A.D. and 1500 A.D. was the heyday of the nomadic pastoralist as a geopolitical force in Eurasia for various reasons. Even in the centuries after 1500 A.D. horses remained critical for mobility. The problem is that in much of agricultural Eurasia there is not sufficient pasturage to raise large numbers of horses.

Pretty much every Eurasian society within reach of the steppe (so basically every society except for those in Western Europe, Japan, and Southeast Asia) had to deal with the menace of armed nomads. Sometimes they paid them off. Sometimes they mobilized enormous armies which incurring crushing costs. And sometime they were conquered.

Between 1700 and 1900 this spectre faded. Improvements in military technology, transport, and mass national mobilization, leveled the gap between the steppe and the settled peoples, to the point where by 1900 the steppe was a marginal factor.

Note: the author of Escape from Rome attributes the rise and fall of Vijayanagara explicitly as a function of its access to horses.

Browncast Episodes 121 & 122: Desi-Rae, heterodox Jamaican American, and Anthony, a black American in China

Razib Khan · August 26, 2020 · 14 comments

Two new BP Podcast episodes are up. You can listen on Libsyn, Apple, Spotify,  and Stitcher (and a variety of other platforms). Probably the easiest way to keep up the podcast since we don’t have a regular schedule is to subscribe to one of the links above!

You can also support the podcast as a patron. The primary benefit now is that you get the podcasts considerably earlier than everyone else. This website isn’t about shaking the cup, but I have noticed that the number of patrons plateaued a long time ago.

First, we talked to Desi-Rae, a Jamaican American woman who has some heterodox views. Her views on race in Jamaica and the United States, relationships, and the fallout from being so vocal. I apologize that my mic was very low on this episode.

Next, I talk to Anthony, a black American who worked in China. We talk about Chinese perceptions of blackness, his daughter, who is half-Chinese, and the prospects for the China-USA relationship.

Brown Pundits