Evidence page · material culture, counted
Every material term in the Rigveda, sorted by metrical stratum. Forts are old and villages are new. Chariots outnumber merchants two hundred to one. And the vocabulary of farming arrives at the very end.
| Word | Meaning | Total | Archaic | Popular |
|---|---|---|---|---|
| púr | fort, stronghold | 83 | 30% | 3% |
| víś | clan, settlement | 200 | 27% | 6% |
| kṣáya | dwelling-place | 70 | 27% | 17% |
| dáma | house, home | 51 | 19% | 3% |
| grā́ma | village | 9 | 0% | 33% |
| véśman | dwelling, house | 2 | 0% | 100% |
Pur — the fort — is 30% Archaic and 3% Popular. It declines sharply.
Grāma — the village — is 0% Archaic and 33% Popular. It appears from nothing. And véśman, a dwelling-house, occurs twice and both are in the latest layer.
The corpus begins in a world of strongholds to be broken and ends in a world of villages to live in.
And note what pur keeps company with: ṛjiśvan 71× · Pipru 70× · Śambara 55× · Divodāsa 50× · āyasá, "of metal", 49× · navatí, ninety, 40× · dṛ "to burst" 32× · bhid "to split" 22×.
Named enemies, a metal material, a count of ninety-nine, and two verbs for breaking things open. The oldest settlement word in the corpus is something the composers are attacking.
| Word | Meaning | Total | Archaic | Popular |
|---|---|---|---|---|
| kṛṣí | ploughing, cultivation | ABSENT FROM THE CORPUS | ||
| vrīhí | rice | ABSENT FROM THE CORPUS | ||
| lā́ṅgala | plough | 1 | 0% | 100% |
| sī́ra | plough | 2 | 0% | 100% |
| sī́tā | furrow | 2 | 0% | 100% |
| yáva | barley | 23 | 17% | 8% |
| urvárā | arable field | 8 | 50% | 37% |
| khála | threshing floor | 1 | 100% | 0% |
Every word for the plough and the furrow occurs in the Popular stratum only — the youngest layer. Lāṅgala once. Sīra twice. Sītā twice. All Popular.
And lāṅgala, sīra and sītā are among the classic substrate loanwords — words with no secure Indo-European etymology, in the semantic field where substrate borrowing always concentrates.
The composers of the old layers have barley and a field. They do not have the implement. The vocabulary of ploughing enters the corpus at the same moment as the village — and it enters as foreign words.
| Word | Meaning | Total |
|---|---|---|
| rátha | chariot | 471 |
| rayí | wealth | 535 |
| vā́ja | prize, booty | 376 |
| gó | cattle | 543 |
| hiraṇyáya | golden | 67 |
| vaṇíj | merchant | 2 |
| śulká | price, toll | 2 |
| niṣká | gold ornament used as value | 3 |
| maṇí | jewel | 1 |
| karmāra · tákṣan | smith · carpenter | absent · 1, Popular |
| nau · plava | boat · raft | absent |
Wealth vocabulary is everywhere — 535 mentions of rayi, 543 of cattle, 376 of vāja, which means both "prize" and "booty."
Commercial vocabulary is essentially absent. Two merchants. Two mentions of a price. One jewel. No smith. No boat, of any kind, in 1,028 hymns.
Wealth is counted constantly and never bought. It is won, given, or taken.
This bears directly on the Meluhha question. A civilisation trading carnelian to Ur across a maritime route had merchants, weights, seals and ships. The Rigveda has none of that vocabulary — no boat, no merchant class, no standard of value, no port. Whatever the composers were, they were not the people running the Indus trade network, and the corpus does not describe that world from the inside.
| Word | Meaning | Total | Archaic | Popular |
|---|---|---|---|---|
| vájra | the bolt, mace | 138 | 27% | 8% |
| ratheṣṭhā́ | chariot-warrior | 7 | 71% | 0% |
| vadhá | slaughter, weapon | 43 | 16% | 25% |
| íṣu | arrow | 11 | 0% | 45% |
| páraśu · ási | axe · sword | not found as lemmas | ||
| sená · pṛ́tanā · sáṃgrāma | army · battle-line · battle | not found as lemmas | ||
Ratheṣṭhā — the man who stands in the chariot — is 71% Archaic and 0% Popular. The oldest layer's characteristic fighter is a chariot warrior, and the type fades.
The arrow inverts: 0% Archaic, 45% Popular.
The caveat matters: senā, pṛtanā and dhanus almost certainly do occur in the Rigveda and are missing here through lemma-form mismatches in my search, not through absence from the text. Treat the "not found" rows as a limitation of the query, not a finding. The stratum percentages for the words that did match are sound.
Kṣayádvīra occurs 8 times. Four are with Rudra — three of them in RV 1.114, his principal hymn, plus 10.92.9. The other four (1.106.4, 1.114.10, 1.125.3, 8.19.10) are not attached to him. So the epithet is Rudra-weighted but not exclusive.
Kṣáya — 69 stanzas — means DWELLING. Its collocates are dūraka "far off" 122× · jīv "to live" 29× · pitumant "having food" 57× · jána "people" 3.3×. A place where people live and eat.
Kṣatrá — 44 stanzas — means DOMINION. Its collocates are dhṛ "to hold" 9.7× · rā́jan 4.9× · bṛhánt "lofty" 4.3× · Varuṇa 3.1× and Aryaman 6.1×. Held power, belonging to the Ādityas.
Both come from √kṣi "to dwell, rule." And Sanskrit has a second, homophonous √kṣi meaning "to destroy." Kṣayádvīra sits on that ambiguity: it reads as "ruling over men" and "destroying men" with equal grammar. The word for sovereignty and the word for annihilation are the same sound in this language, and Rudra's epithet uses both.
| Colour | Stanzas | Who or what it is applied to |
|---|---|---|
| hári — tawny, golden | 256 | Indra 120 · Soma 32 · cattle 18 · Agni 7 |
| aruṣá — ruddy, red | 59 | Agni 18 · horses 6 · Uṣas 5 · cattle 5 |
| babhrú — brown | 24 | Rudra 4 · Soma 4 · cattle 3 · Agni 2 · Sūrya 2 · Indra 1 |
| śyāvá — dark brown, dusky | 10 | scattered; no dominant association |
Each major god has a colour, and they are not the same colour. Indra is hari — tawny-gold, 120 co-occurrences, overwhelmingly his. Agni is aruṣa — red, 18. Rudra is babhru — brown, and he is its single strongest divine association.
So the observation stands on its own terms: the god of medicines, of the wild, of the plea to be spared, who has no ritual vocabulary and no clear Iranian cognate, is the brown one, while the god the poets identify with is the golden one.
What the data does not establish is that this encodes human skin colour. Babhru is applied to soma juice and cattle in the same corpus, and the word is inherited Indo-European. The colour-coding of the pantheon is real and patterned. Reading it as racial is an inference the collocations neither support nor refute — and it is a separate question from RV 3.34.9, where varṇa is applied to a people directly.
471 chariots. 543 cattle. 83 forts, mostly in the oldest layer, mostly being broken open. 7 chariot-warriors, 71% Archaic.
2 merchants. 2 prices. 1 jewel. No smith. No boat. No plough until the final stratum, and then in borrowed words. No village until the final stratum.
This is not a description of an urban trading civilisation, and it is not a description of a settled farming one. It is a mobile, chariot-equipped, cattle-counting population that attacks fortified places, and that acquires villages, ploughs and furrows at the very end of its textual record — in vocabulary that is not its own.