What changed? The genetic record of the transition to endogamy

The Moorjani 2013 paper had a headline finding broader than its dating: India transitioned, sometime between roughly 1,900 and 4,200 years ago, from a society where mixing across populations was common to one where mixing even between closely related groups became rare ^[S1]. Geneticists call this transition the rise of endogamy — the practice of marrying within one's group.

The genetic signal of the transition is unmistakable. In the centuries after about 2,000 years ago, in many populations, the ancestry mix that had been forming over millennia stops changing. New mixing essentially halts. The endogamous group structure that defines so much of South Asian society today — caste, jati, regional and religious identity — starts leaving its signature in the genome.

The genetics is settled on that it happened, and roughly when it happened. The harder question — why — is genuinely open, and how you answer it depends on which lines of evidence you take seriously.

What the genetics actually says

The data is clear and widely confirmed. Multiple studies, looking at different subsets of Indian populations with different methods, find the same broad pattern: ANI/ASI mixing was extensive through at least 4,000 years ago and largely complete by about 2,000 years ago, after which sub-population structures became more rigid [S1, S2].

This isn't subtle. By the time we look at Iron Age and early-historical samples, populations that had been actively mixing across regional and group boundaries are now mostly endogamous. The transition compresses, in many cases, into a few centuries.

What the data does not say is what caused this. Genetics measures the consequence — gene flow stopping — not the cause. The cause has to be inferred from what else was happening in South Asia in the centuries before and during the transition.

Hypotheses for the cause

Several hypotheses are taken seriously by different researchers, often with implicit disagreement about which lines of evidence carry the most weight.

Codification of marriage rules in Brahmanical texts. The literature of Dharmasutra and Dharmashastra — texts on Brahmanical ritual, social, and legal practice — develops formal rules for marriage within varna and jati groups. Many of these texts are conventionally dated to roughly 800 BCE through 300 CE, overlapping with the latter part of the genetic mixing window. The hypothesis: as Brahmanical authority consolidated, formal marriage rules became enforceable in elite groups and gradually shaped behavior across society.

Political consolidation under post-Mauryan and Gupta states. Indian state formation in the post-Mauryan period (roughly 200 BCE onward) consolidated administration over large territories. Large states, in many world regions, have been correlated with formalized social hierarchies — partly as instruments of governance, partly as outcomes of economic specialization. The hypothesis: as Indian states consolidated, group identities calcified along occupational and regional lines.

Demographic stabilization after a long mobile period. From the Indus Valley collapse (~3,500 BP) through the Iron Age, South Asia saw substantial population movement — Steppe arrivals, urban relocations, agricultural intensification. As these movements settled, populations may have stabilized into the regional configurations that produced endogamy as a side effect of geographic and economic settling ^[S2].

Economic specialization. The development of caste-based occupational specialization — jati tied to specific trades — created social pressure for in-group marriage to maintain occupational knowledge, property, and status. The hypothesis: economic structures drove the social ones, not the other way around.

These aren't mutually exclusive. They could all be partially true, and the actual mechanism may have varied by region and group.

Why this is contested

The disagreement isn't primarily about the genetics — it's about how much weight to give different lines of evidence in answering the why.

Researchers focused on Brahmanical texts tend to align the genetic transition with textual chronology and treat the Dharmashastra corpus as central to explaining endogamy's rise ^[S3]. Researchers focused on archaeology and political history emphasize state formation and economic specialization. Researchers focused on comparative anthropology look at parallel transitions in other societies and ask what drives endogamy generally. Each line of evidence has internal disagreements about chronology and interpretation, and the dating of Brahmanical texts is itself a contested area ^[S4].

What's worth flagging clearly: the genetic dating doesn't tell us caste began at any particular date. Caste in the modern sense — fixed jati, occupational lineage, ritual hierarchies — likely emerged through some combination of all of the above, with regional variation and over a long timescale. The genetic signal of endogamy is one of multiple records of that emergence, and it's the one that's easiest to date precisely.

What the next steps look like

The harder questions — when did specific group boundaries form, and what drove them — are starting to be addressable with finer-grained genetic methods. By looking at within-jati versus between-jati genetic distances, and dating those distances precisely, researchers can begin to map when specific endogamous groups formed. Some have formed within the last thousand years; others are older.

For now, the careful position is this: the rise of endogamy in South Asia is a real and dateable demographic transition. Multiple plausible causes have been proposed, and multiple are likely to have contributed. Settling which one was primary — or whether that question even has a single answer — will take continued work across genetics, archaeology, comparative history, and textual studies.

This is one of the live frontiers of South Asian historical research. Swaveda will get into specific aspects of it in future pieces, with the contested-topics framing we've used here.