A reading method · South Asian religious history

Ask who benefits from the critique, and you can predict what it leaves standing.

Every major internal challenge to Brahmanical authority attacked the part of the system that constrained its own author, and left the rest intact. Run that test across the canon and the pattern is not subtle.

Revolt	Whose	What it broke	What it kept
Upaniṣadic c. 700–600 BCE	Kṣatriya kings — a rival elite	Ritual monopoly. Sacrifice demoted to the returning path	Varna, entire
Bhakti in the Gītā c. 200 BCE–200 CE	A devotional cult seeking mass reach	Ritual gatekeeping of liberation (9.32)	Varna duties by name (18.41–44)
Buddha, Mahāvīra	Outside the twice-born frame	Birth as a spiritual category	—
Basava, Siddhars, Kabir, Ravidās	Low-caste and outcaste	Caste, priest, temple, scripture	—

The first two are internal and survive as scripture. The last two are external and are absorbed or expelled: the Buddha becomes Viṣṇu's ninth avatāra — in some Purāṇas an incarnation taken to delude the wicked — and the Lingayats are persecuted.

Neither the Upaniṣadic revolt nor the bhakti opening is a subaltern revolt. Neither touches the varna order.

Finding 01

“Vedic” is a custody label that reads as an origin label.

The tradition defines Vedic by transmission — attachment to a śākhā, status as śruti — not by where the content came from. That is a defensible definition. The problem is what the word does to a reader who takes it as a statement of provenance.

Arguing “these texts are not really Vedic” invites a definitional counter that goes nowhere. Arguing that the label launders provenance survives contact with a Sanskritist.

The evidence: what has cognates, and what does not

Not even Iranian has the second column — and Iranian shares everything in the first. That is the tighter test, and it localises where the thinking happened.

And the doctrinal complex with no comparanda outside South Asia is precisely the one a Kṣatriya king tells a Brahmin the Brahmins never had (Chāndogya 5.3.7; Bṛhadāraṇyaka 6.2).

Practices are form, not merely meaning

Comparative linguistics works on form because sound correspondence is regular and falsifiable. But ritual sequence, posture, breath technique and initiation structure are also form — transmissible, comparable, and reconstructable in principle. Dumézil attempted exactly that for Indo-European ideology.

Apply the practice-comparison honestly and the result is the same. What is shared: the fire cult, metrical craft, the sacrificial exchange, possibly the youth war-band. What is not: seated meditation, breath control, the yogic body, liberation-seeking asceticism. Widening the test from words to practices does not rescue the inheritance — it makes the gap larger.

The comparative method licenses claims about descent, not about excellence.

Shared words tell you the languages are cousins. They tell you nothing about who thought what. The nineteenth century read a genealogy of grammar as a genealogy of genius — and did so while being drawn to India by exactly the material the inheritance could not account for. They came for the philosophy and claimed kinship through the grammar.

The reason the gap went unnoticed: a story was already in place — arrival, mixture with natives, decline. Anything distinctively Indian could be filed under degeneration rather than counted as evidence that the inheritance was thin. The theory was unfalsifiable by construction.

Finding 02

Patronage is the variable that predicts canonisation.

The Bṛhadāraṇyaka belongs to the Vājasaneyi school — Yājñavalkya's own. The record of Janaka's generosity is written by the lineage that received it. There is no independent account.

The highest honour available is reclassification into the priestly category. The origin myth of the forest peoples is the discarded sin of a king who refused.

Finding 03

Source becomes stain: places turn into categories of defective birth.

Manu's taxonomy of illegitimate birth is populated by the names of real peoples, regions and republics.

Category	What the name is	Where
Vaidehaka	the people of Videha — the region the Upaniṣads credit with rebirth and liberation	Manu 10.11, 10.26
Māgadha	the people of Magadha	Manu 10.11, 10.26
Sūta	the bard-charioteers — narrators of the Mahābhārata	Manu 10.11, 10.26
Licchavi	the ruling clan of Vaiśālī, a republic — Mahāvīra's birthplace, seat of the Second Buddhist Council	Manu 10.22
Draviḍa, Coḍa, Āndhra, Kaliṅga, Puṇḍra, Māhiṣaka	southern and eastern peoples, "degraded through non-observance"	Manu 10.22, 10.43–44, 10.48; MBh 13.33, 13.35

The tradition concedes the direction of derivation and then reverses it. Medhātithi, commenting on Manu 10.44, records that Puṇḍra and the rest originally stood as names of countries, and are used here on the theory that they really denote castes and are only indirectly applied to countries.

And the exit is closed. Manu 10.45: all groups outside the four varnas are dasyu — whether they speak the language of the Mlecchas or that of the Ārya. Speaking Sanskrit does not save you. Nothing is outside; there is only lower.

Degraded through non-observance: absorb by inventing a fall.

The move makes varna original and universal, converts difference into decline, and makes Brahmin presence the condition of rank. It runs on regions (Videha), republics (Licchavi), peoples (Draviḍa), and — in the Vena myth — on the forest itself.

Finding 04 · corpus evidence

Colour narrows: from hosts, to wombs, to skin, to the name of the order.

Computed over the full Rigveda — 164,758 lemmatised tokens, VedaWeb / Cologne, with Arnold's metrical strata per line (Gunkel & Ryan).

Stage	Form	Where
1	kṛṣṇáyoni — "black-wombed", of the Dāsa hosts	RV 2.20.7 · Archaic stratum
2	"the dark fifty thousand"; "the dark ones" driven off with Varcin and Śambara	RV 4.16.13 · 6.47.21 — family books
3	kṛṣṇā́ tvac — black skin	RV 1.130.8 · 9.41.1 — later books only
4	varṇa — "colour" as the name of the social order	first millennium CE

Blackness is attached to descent before it is attached to appearance. Yoni is womb, source, origin — and later, caste. Fifteen of twenty family-book uses of kṛṣṇá are fire, night, cloud and cattle; three are applied to enemies; and the one in the oldest stratum attaches colour to birth.

Named enemies, fortified places

Śúṣṇa 38 · Śámbara 21 (99 forts, RV 7.99.5) · Dhúni 24 · Pípru 11 · Námuci 9 · Varcín 4 · Árbuda 5 — concentrated in the family books, absent or near-absent from Book 10. Indra is puraṃdará, fort-breaker.

The asura chronology, tested

In the family books ásura means lord, applied to Rudra, Varuṇa, Agni, Savitar, Mitra-Varuṇa and Pūṣan. The demotion to "demon" begins in Book 10. India demonised asura late and independently — the familiar story of a mutual Indo-Iranian inversion at the moment of schism does not fit the data.

Open — not settled, and worth the trip to a library