As Caste Vanishes Only Genes Remain

Raimond Klavins/An ancient Hindu Temple in Suchindram, India.

In February 2022, Seattle became the first U.S. city to ban caste-based discrimination. This follows the California State University student body association banning caste-based discrimination, in addition to the University of California, Davis and Brown University. In 2020, a lawsuit was filed in California against Cisco Systems alleging caste-based discrimination where a lower-caste Dalit employee was passed over for promotion by an upper-caste Brahmin colleague. The case against two individuals accused of discrimination was dismissed, but the suit against Cisco continues to be litigated. What’s going on here?

There are three intersecting factors at play in modern American society that makes caste relevant. First, Indian Americans are now publicly prominent. Long the most educated and wealthiest ethnic group in the U.S., in the past few decades they have become prominent in politics and business. Not only are the CEOs of Google, Microsoft, and Starbucks Indian American, but two Republican candidates for the presidential nomination—Vivek Ramaswamy and Nikki Haley—were Indian American, while the current Vice President, Kamala Harris, is the daughter of an Indian woman. They are now the quintessential visible minority at the commanding heights of American culture. 

Second, the caste system is inextricably linked to Indian, and more precisely Hindu, identity in the minds of most Americans. In fact, most elementary school treatments of Indian history and society include a section on caste. And third, the hierarchical caste system is a perfect foil for the current American elite’s fixation on social panics and policies targeted at grades of status. India’s caste system is the absolute inversion of the egalitarian and flat utopia envisioned by modern social justice acolytes—though ironically both are highly identitarian.

Though the caste system dominates much of Indian life, it does not dominate Indian American life. At slightly over 1% of the U.S. population, only about half of Indian Americans identify as Hindu, the religion from which the broader categories of caste, or varna, emerge. While caste endogamy—marrying within one’s caste—in India remains in the range of 90%, in the U.S. only 65% of American-born Indians even marry other people of subcontinental heritage, and of these, a quick inspection of The New York Times weddings pages shows that inter-caste marriages are the norm. While tensions between the upper-caste minority and middle and lower castes dominate Indian social and political life, 85% of Indian Americans are upper-caste, broadly defined, and only about 1% are truly lower-caste. Ultimately, the minor moral panic over caste discrimination among a small minority of Americans is more a function of our nation’s current neuroses than the reality of caste in the United States. 

But this does not mean that caste is not an important phenomenon to understand. In various ways, caste impacts the lives of the more than 1.4 billion citizens of India18% of humans alive todaywhatever their religion. While the American system of racial slavery is four centuries old at most, India’s caste system was recorded by the Greek diplomat Megasthenes in 300 BC, and is likely far more ancient, perhaps as old as the Indus Valley Civilization more than 4,000 years ago. Indian caste has a deep pedigree as a social technology, and it illustrates the outer boundary of our species’ ability to organize itself into interconnected but discrete subcultures. And unlike many social institutions, caste is imprinted in the very genes of Indians today.

Of Memes and Genes

Beginning about twenty-five years ago, geneticists finally began to look at the variation within the Indian subcontinent, and were shocked by what they found. In small villages in India, Dalits, formerly called “outcastes,” were as genetically distinct from their Brahmin neighbors as Swedes were from Sicilians. In fact, a Brahmin from the far southern state of Tamil Nadu was genetically closer to a Brahmin from the northern state of Punjab then they were to their fellow non-Brahmin Tamils. Dalits from the north were similar to Dalits from the south, while the three upper castes, Brahmins, Kshatriyas and Vaishyas, tended to cluster together against the Sudras.

Some scholars, like Nicholas Dirks, the former chancellor of UC Berkeley, argued for the mobility and dynamism of the caste system in their scholarship. But the genetic evidence seemed to indicate a level of social stratification that echoes through millennia. Across the subcontinent, Dalit castes engaged in menial and unsanitary labor and therefore were considered ritually impure. Meanwhile, Brahmins were the custodians of the Hindu Vedic tradition that ultimately bound the Indic cultures together with other Indo-European traditions, like that of the ancient Iranians or Greeks. Other castes also had their occupations: Kshatriyas were the warriors, while Vaishyas were merchants and other economically productive occupations. 

Brahmins, Kshatriyas, and Vaishyas were traditionally the three “twice-born” castes, allowing them to study the Hindu scriptures after an initiatory ritual. The majority of the population were Sudras (or Shudras), India’s peasant and laboring majority. Shudras could not study the scriptures, and might be excluded from some temples and festivals, but they were integrated into the Hindu fold, and served by Brahmin priests. A traditional ethnohistory posits that elite Brahmin priests and Kshatriya rulers combined with Vaishya commoners formed the core of the early Aryan society in the subcontinent, with Shudras integrated into their tribes as indigenous subalterns. Outcastes were tribes and other assorted latecomers who were assimilated at the very bottom of the social system, performing the most degrading and impure tasks.

Razib Khan/The caste system as a layered varna system with five classes and numerous integrated jati communities.

This was the theory. Reality is always more complex. In India, the caste system combines two different social categories: varna and jati. Varna derives from the tripartite Indo-European system, in India represented by Brahmins, Kshatriyas, and Vaishyas. It literally translates as “color,” white for Brahmin purity, red for Kshatriya power, and yellow for Vaishya fertility. But India also has Shudras, black for labor. In contrast to the simplicity of varna, with its four classes, jati is fractured into thousands of localized communities. If varna is connected to the deep history of Indo-Aryans and is freighted with religious significance, jati is the concrete expression of Indian communitarianism in local places and times. 

Though non-Indians may see in varna the essence of caste, in the subcontinent it is jati through which caste expresses itself in day to day lifein the U.S., jati is almost nonexistent for Indian Americans. Dalits, the former outcastes, may not have varna status, but they have numerous jatis across the nation. Christians, Muslims, and Buddhists who convert from Hindu Dalit jatis still retain their communal identity in the eyes of their new co-religionists. In the Islamic Republic of Pakistan, sanitation positions have sometimes stipulated Christian applicants only, because the majority of that nation’s Christians descend from Dalit Hindu converts who engaged in those occupations.

The deviation of caste from varna theory is illustrated by the reality that in much of South India the only twice-born caste are Brahmins; there are no Kshatriyas or Vaishyas. Rather, the kings and merchant princes of these regions hailed from Shudra jatis. In Gujarat and Punjab in the northwest, Brahmins may still be given religious primacy as the highest caste, but power and social status are the purview of jatis who engaged in trade and landowning of the Vaishya, Kshatriya, or even Shudra varna.

This incredible kaleidoscope of social complexity has played out over thousands of years across a subcontinent that spans the snowy fastness of the Himalayas in the north to the jungles of the deep tropics in the south. For decades, social scientists have speculated about the fluidity and flexibility of the caste system. But over the last generation, geneticists who have looked at the data note that various jati categories seem to be distinct populations which have been endogamous for thousands of years. The traditional theory, held in some suspicion, was that the three twice-born castes were originally Aryan, and genetics actually confirms that these jatis bracketed into the Brahmin, Kshatriya, and Vaishya varna do indeed often have more Eurasian steppe ancestry connecting them to Europeans than regional Shudra jati. And Dalit castes across the subcontinent have the least steppe, ergo Aryan, ancestry.

The best evidence is that Indo-Aryans arrived in the Indian subcontinent between 3,500 and 4,000 years ago. This means it is likely that the varna system originated during this period, and the overrepresentation of steppe ancestry in the subcontinent among elites is nearly 4,000 years old! The economic historian Gregory Clark has stated that India shows the highest persistence of social status among all his data, with heritability in the range of 0.90. But there is a peculiar lacuna in the discourse around caste with its fixation on varna: India is the only Indo-European society with jati. Though, in the abstract, the ritual status of Kashmira Pandits and Tamil Iyer Brahmins is broadly the same, these two communities speak different languages, have different attitudes to vegetarianism and are even genetically very distinct. What matters of affinity, socialization and marriage is the jati of the Iyers and Pandits, not their varna status.

It is possible that jati was an invention of the Indo-Aryans who entered an alien landscape of incredible diversity. But there is evidence that in fact jati may predate Indo-Aryans, and it was varna that was overlaid atop it as a scaffold. Across South India, the non-Brahmin population ranges from regional landlords to Dalits relegated to degrading occupations. Both these groups are notionally non-Aryan, but genetically there are clear differences between Dalits and local Shudra jatis, indicating that Indo-Aryan culture may not be a necessary precondition for the social and genetic segregation typical of the Indian subcontinent.

The Past and Future of Caste

The Indus Valley Civilization spanned much of the northwest of the subcontinent, dominating modern Pakistan, but also extended into Indian Punjab and Gujarat. It was substantially larger than Sumer or Egypt and may therefore have been multiethnic. But it was also characterized by planned cities with ancient public works, like efficient sewage systems. And yet this early Indian civilization did not seem to have thrown up palaces or exhibited much sign of inequality. The evidence from Egypt and Sumer suggests that Bronze Age cultures were hierarchical and dominated by an all-powerful ruler, but the Indus Valley Civilization defies that characterization.

But what if the political anarchism that seems to be suggested by lack of centralized monumental structures at Harappa and Mohenjo-Daro, two of the largest cities of this civilization, mask the reality that these early Indians were organized around tight communal social structures that served to regulate relationships between elements of this early urban world? In other words, what if the Indus Valley Civilization was an assembly of Bronze Age jatis and the jatis persisted as social technologies long after the cities collapsed in the wake of climatic shocks in the century before 2000 BC?

The tendency toward central rule in the civilizations of West and East Asia, with paramount rulers like Khufu in Egypt, Sargon the Great in Akkad, and Qin Shi Huang in China, may reflect the lack of social texture and complexity in these cultures. India’s relatively decentralized character could instead reflect that many of the duties and functions of centralized states and paramount rulers fell upon the jati, those communities that self-organized and self-governed. To this day, caste remains a feature of the Indian republic’s constitution, prohibiting discrimination, but also carving out set-asides for Dalits and lower castes. The Indian state, modeled on British norms, must still grapple with the reality of caste, and many Indian political parties have been implicitly and explicitly reflections of castes.

Indian caste, unlike the Chinese mandarin system and even the European blood nobility, was predicated on an extreme level of fine-grained endogamy. Across much of India the rate of out-marriage seems to be 0.5% per generation within local villages, divided between a handful of customary communities, and often dominated by one jati. The social and cultural forces necessary to induce this level of adherence to norms and values is difficult for Westerners to imagine, but Indian villages may be thought of as early-modern total states, where everyone was in everyone else’s business.

But the current endogamy rates in the subcontinent, closer to 90% than 99%, mean that this thousand-year institution is not going to persist more than a few more centuries. Without the ties of endogamy inducing social cohesion, India will develop a more conventional class system, defined by more fluidity and flexibility. But what will happen when humanity leaves the gravity well of the Earth? 

Will future Martians intermarry with future Terrans? Will colonies in the asteroid belt promiscuously interact with each other? The conventional understanding of the human spacefaring future is that we face the threat of speciation. But speciation implies total isolation and separation. Perhaps we might better think of the human future to be a constellation of endogamous subpopulations united by communication systems and a common ethos into which they inflected their own communal norms. Perhaps the distant human future will again be of different castes, each specially oriented toward their ideal occupations and living in close quarters in artificial habitats. India’s long history shows that humans are capable of such a complex yet stable equilibrium.

Razib Khan is a co-founder GenRAIT, a deep tech company that deploys a platform for the life sciences. He also writes at Razib Khan’s Unsupervised Learning. You can follow him at @razibkhan.