The Rigveda's Accidental Census: What Its Personal Names Give Away

A hundred head of cattle from a man who should not be in the poem

Near the end of RV 8.46, a poet called Vaśa Aśvya stops praising Vāyu and starts counting. He has had sixty thousand horses, ten thousand cows, two thousand camels. Someone made a hundred camels bleat for him in the grazing herd. It reads like a man reciting a receipt from memory, which is roughly what it is. Then verse 32 names two of the people who paid:

śataṃ dāse balbūthe vipras tarukṣa ā dade

“A hundred the inspired poet took from the dāsa Balbūtha, from Tarukṣa.”

Rigveda 8.46, verse 32. After Griffith (1896), diacritics standardised.

Two things about that line are strange, and Vedic scholarship has been chewing on both for a century. The first is the word dāsa, which everywhere else in the Rigveda marks the enemy, the outsider, the people Indra is repeatedly asked to break. Here it is attached to a man whose gift is being gratefully itemised in a hymn. The second is quieter and, I think, more interesting: neither Balbūtha nor Tarukṣa is a Sanskrit word. You cannot take them apart. They have no roots, no transparent compounds, nothing to translate. They sit in the middle of the most linguistically self-conscious poetry of the ancient world like two pebbles in a bowl of rice.

The hymns of the Rigveda were shaped, reshaped, arranged by family, ordered by metre and length, and transmitted with a fidelity that borders on the unbelievable. What was in them was controlled. But a poet who wants to be paid again must name the man who paid him, correctly, in public. A poet who wants the right to sing at all must name the ancestor he inherited it from. Names are load-bearing. They are the part of the text nobody could tidy.

Which makes the Rigveda something it never intended to be: a census. Not a complete one, not a fair one, and certainly not a deliberate one. But a list of several hundred real human beings from the second millennium BCE, recorded by people who were thinking about something else entirely. This piece is about what falls out when you read that list.

The grammar of being called something

Start with what a Rigvedic name is physically made of, because that turns out to be the whole argument in miniature. These names are not opaque labels drawn from a fixed stock, the way Michael or Priya are today. They are ordinary Sanskrit words and phrases, freshly assembled, and they mean what they say.

Trasadasyu, the Pūru king, is a small sentence: “at whom the Dasyu trembles.” Bharadvāja, the clan name behind Maṇḍala 6, is “bearing the prize.” Viśvāmitra is “friend of all.” Sudās, the Bharata king of the Ten Kings battle, is simply “giving well,” which is exactly what a poet would want his patron to be called. Dīrghatamas is “long darkness.” Romaśā is “hairy.” Ghoṣā is “a shout.”

Notice what is absent. The name-suffixes that later became social barcodes in India, -śarman, -varman, -gupta, -dāsa as a caste-marking tail, are not operating here as a system; Vedic naming works on different principles from the classical period, which is the reason Vedic onomastics needs treating separately at all [1]. Nobody in the Rigveda is named to advertise their varṇa. They are named for what they own, what they did, what they look like, or who bore them.

The -śravas cluster deserves its own moment. Śravas means “fame, what is heard,” and it is the thing Vedic poetry exists to manufacture. RV 1.9.7 asks Indra for śravo bṛhat … akṣitam, “lofty fame, unfailing.” Rüdiger Schmitt showed that this exact phrase, śrávas ákṣitam, matches the Homeric kléos áphthiton, “imperishable glory,” morpheme for morpheme, and belongs to inherited Indo-European poetic diction [2]. Now look at the personal names.

Across the family, people were named after the substance the poets were paid to produce. And the loop closes inside a single hymn: Pṛthuśravas Kānīta, the patron whose name means Broad Fame, is described four verses later as the man who gave a golden chariot and varṣiṣṭham akṛta śravaḥ, “made for himself the loftiest fame” (RV 8.46.24). The name, the payment and the product are the same word, doing three jobs at once.

One of these names is a small delight on its own. RV 1.112.11 thanks the Aśvins for the aid “by which sweetness poured from the vat for Dīrghaśravas, the merchant, son of Uśij.” A trader called Long Fame. The Rigveda is not usually thought of as a text with merchants in it.

Aside. A working caution: we have no birth records. When the text calls a man Dīrghatamas we cannot tell whether that was the name his parents gave him, a nickname that stuck, a poetic persona, or a descriptive epithet that hardened into a name over generations of recitation. Everything below is read with that uncertainty attached, and the honest version of Vedic onomastics never quite escapes it.

Dog’s Tail, Long Darkness, and the Hairy One

A culture that builds names out of live vocabulary will produce some names nobody would choose today.

The clearest case is Śunaḥśepa, who appears in the Rigveda already as a figure in trouble. RV 1.24.13 has him “bound to three pillars” calling on Varuṇa; RV 5.2.7 credits Agni with loosing him “from the stake, bound for a thousand.” The later prose narrative in the Aitareya Brāhmaṇa fills in the story of a boy sold for sacrifice. What the story rarely mentions is what the boy is called. Śunaḥ is the genitive of śvan, “dog.” Śepa is a tail, and in Vedic usage frequently the male member. The hero of one of the oldest sacrificial legends in India is named Dog’s Tail, or worse, and translators have quietly disagreed about which for two centuries. His brothers in the Brāhmaṇa version carry the same joke twice over: Śunaḥpuccha and Śunolāṅgūla, both of which also mean “dog’s tail.”

Then there is Dīrghatamas, “Long Darkness,” the poet credited with the great riddle hymn RV 1.164. Twice the Rigveda refers to him by his matronymic and adds a detail:

ye pāyavo māmateyaṃ te agne paśyanto andhaṃ duritād arakṣan

“Your watchers, Agni, seeing, guarded the blind son of Mamatā from harm.”

Rigveda 1.147, verse 3 (repeated at RV 4.4.13). After Griffith (1896), diacritics standardised.

The word is andha, blind. So the poet whose name means Long Darkness is described in the text itself as a blind man. Whether the name came first or the condition did is unrecoverable, and that is precisely the kind of question Vedic names keep raising and refusing to answer.

Body-names run through the women too, and they run together in a way I have not seen remarked on often. RV 8.91 tells a compact, strange story: a girl named Apālā finds soma by the roadside, presses it in her teeth for Indra, and asks him to make three things grow hair, her father’s bald head, the barren field, and her own body. The verb she uses is romaśā kṛdhi, “make them hairy.” The hymn ends with Indra pulling her through three openings and giving her sūrya-tvac, sun-skin.

khe rathasya khe anasaḥ khe yugasya śatakrato | apālām indra triṣ pūtvy akṛṇoḥ sūryatvacam

“Through the hole of the chariot, of the cart, of the yoke, O Śatakratu: purifying Apālā three times, Indra, you made her sun-skinned.”

Rigveda 8.91, verse 7. After Griffith (1896), diacritics standardised.

Now turn to RV 1.126.7, whose speaker the tradition names Romaśā, “the Hairy One.” She says, in effect, come closer, touch me gently, do not think the hair on my body is thin, I am as woolly as a Gandhāra ewe. Griffith could not bring himself to put that into English and printed the verse in Latin instead, which is the most Victorian sentence in the whole 1896 translation.

Two women, two hymns, one vocabulary item. The name and the joke are the same word. That is what it means for a naming system to sit directly on top of the living language rather than in a separate reservoir of “names.”

Methods note. The attribution of RV 1.126.7 to a speaker called Romaśā comes from the Anukramaṇī, the traditional index of poets and deities, which postdates the hymns by centuries. Traditional attributions are evidence about how the tradition read itself, not independent proof about the Bronze Age. I use them here as data, flagged.

The names that will not parse

Return to Balbūtha. He is not alone.

Scattered through the Rigveda is a set of names that resist Indo-Aryan etymology completely. Some belong to enemies, which is unsurprising. What is surprising is how many belong to people on the inside of the ritual economy: donors, patrons, the men whose generosity the poets are contractually obliged to advertise.

The clearest is Bṛbu. Bharadvāja’s hymn to Indra ends with three verses that have nothing to do with Indra at all:

adhi bṛbuḥ paṇīnāṃ varṣiṣṭhe mūrdhann asthāt | uruḥ kakṣo na gāṅgyaḥ

“Bṛbu has set himself above the Paṇis, on their highest head, like the wide thicket on the Gaṅgā.”

Rigveda 6.45, verse 31. After Griffith (1896), diacritics standardised.

The following verses call his bounty thousandfold and swift as wind, and say that all the singers praise him. Sāyaṇa, commenting two millennia later, glosses Bṛbu as the takṣan, the carpenter or artificer, of the Paṇis. The Paṇis are the Rigveda’s misers, the hoarders who withhold cattle and refuse sacrifice, the antagonists of the Vala myth and the structural opposite of everything the gift economy of poet and patron runs on. And a Ṛgvedic poet is publicly thanking their carpenter, by name, for a gift of a thousand.

The Kīkaṭa verse is worth reading whole, because it shows how the text thinks about people who are outside its system:

kiṃ te kṛṇvanti kīkaṭeṣu gāvo nāśiraṃ duhre na tapanti gharmam | ā no bhara pramagandasya vedo naicāśākhaṃ maghavan randhayā naḥ

“What do your cows do among the Kīkaṭas? They yield no milk-mixture, they heat no gharma pot. Bring us the wealth of Pramaganda; put the low-branch folk in our power, Maghavan.”

Rigveda 3.53, verse 14. After Griffith (1896), diacritics standardised.

The complaint is not that the Kīkaṭas are foreign. It is that they own cattle and do not perform the rituals that cattle exist to fund. Their chief has a name the poet cannot decline properly, and the poet wants his property.

What the linguists actually claim, and where they split

F.B.J. Kuiper assembled the case that a substantial layer of Rigvedic vocabulary is borrowed from languages spoken in South Asia before Indo-Aryan arrived. His figure was 383 specifically Rigvedic words of non-Indo-Aryan origin, roughly four per cent of the vocabulary [3]. Thomas Oberlies, working more conservatively, allows 344 to 358 secure items; and Michael Witzel points out that even after subtracting every local personal name and place name from the list, something like 211 to 250 “foreign” words remain, around two per cent of the total [4]. In a hieratic text composed in inherited Indo-Iranian poetic language, that is a lot of borrowing.

The counter-case is real and was argued hard. Rahul Peter Das published a sustained methodological attack, “The Hunt for Foreign Words in the Ṛgveda,” arguing that the criteria used to declare a word foreign were too loose and that the results were unfalsifiable; Kuiper replied in the same issue of the Indo-Iranian Journal [5][6]. Hans Henrich Hock has surveyed the wider question of pre-Ṛgvedic Indo-Aryan and Dravidian convergence and found the evidence considerably weaker than its advocates assume [7]. And Manfred Mayrhofer, who knew the Rigvedic lexicon as well as anyone alive, devoted a whole monograph in 2003 to Rigvedic personal names under a title that is itself the argument: Sicheres und Zweifelhaftes, the certain and the doubtful [8].

Here is the honest position, and it matters for how you read every name in the tables above. An unetymologisable name is not proof of a foreign language. It may be a lost Indo-Aryan formation, a dialectal variant, a hypocoristic pet-form worn down past recognition, or a transmission error frozen in place. The gap between “I cannot analyse this” and “this is not Indo-Aryan” is exactly where the argument lives.

What survives the deflation is smaller but sturdier, and it is a social fact rather than a linguistic one. Whatever language Balbūtha’s parents spoke, the Rigveda records that a man the poets called a dāsa handed over a hundred head of livestock and got his name sung to Vāyu for it. The category was porous. You can read the polemics against the Dāsas and Dasyus as a closed ethnic wall only by skipping the donor lists.

[!NOTE] The camels in this hymn are not decoration. RV 8.46 counts camels at verses 22 and 31, and camels are not Punjab animals. Together with the unusual donor names, they place Vaśa Aśvya’s patrons on the north-western fringe, in contact with Iranian-speaking populations. Maṇḍala 8 generally leans west.

Who gets a name of their own

Sort the human beings in the Rigveda by how they are named and a hierarchy appears that nobody wrote down.

graph TD
    A[A person in the Rigvedic world] --> B[Named as patron in a danastuti]
    A --> C[Named as poet or ancestor]
    A --> D[Named as enemy or outsider]
    A --> E[Not named at all]
    B --> F[Name fixed in the recited text]
    C --> F
    D --> F
    F --> G[Etymology transparent]
    F --> H[Etymology resists analysis]
    H --> I[Lost Indo-Aryan formation?]
    H --> J[Pre-Indo-Aryan substrate?]

Women pass through three different doors, and the door matters more than the name.

The middle row is the one to sit with. The Vedic suffix -ānī attaches to a man’s name and produces “the woman of that man.” Indrāṇī is not a name; it is Indra plus a suffix. And in RV 10.102, one of the most kinetic passages in the entire collection, the person who wins the chariot race is called Mudgalānī:

ut sma vāto vahati vāso asyā adhirathaṃ yad ajayat sahasram | rathīr abhūn mudgalānī gaviṣṭau

“The wind lifts her garment when she won the car-load worth a thousand. Mudgalānī became the charioteer in the contest for cattle.”

Rigveda 10.102, verse 2. After Griffith (1896), diacritics standardised.

She drives. Her husband Mudgala takes the prize, and the hymn says so at verse 5: “thereby Mudgala won well-pastured cattle in hundreds and thousands.” We know his name. Hers is his name with a feminine ending, the way you might call someone “Mrs Mudgala” in a match report about the race she just won.

The third row is stranger still and, once you see it, hard to unsee. Two of the most important poets in the collection are identified by their mothers. Dīrghatamas is Māmateya, son of Mamatā. Kakṣīvant is Auśija, son of Uśij, and the epithet is used constantly, at RV 1.18.1, RV 1.119.9, RV 4.21.6 and elsewhere. In a corpus this patrilineal, functioning matronymics are an anomaly that has never been fully explained. The women themselves are gone. All that survives of Mamatā and Uśij is the fact that men were identified by having come from them.

And then there is Ghoṣā, who is named twice in the text itself and not merely in the index:

ghoṣāyai cit pitṛṣade durone patiṃ jūryantyā aśvināv adattam

“To Ghoṣā, sitting in her father’s house, growing old, you Aśvins gave a husband.”

Rigveda 1.117, verse 7. After Griffith (1896), diacritics standardised.

At RV 10.40.5 she speaks for herself, a king’s daughter asking the Aśvins directly to be near her by day and by night. Her name is the ordinary Vedic word for a shout, a din, the noise an army makes going into battle; the corpus uses ghoṣa that way dozens of times. The one woman in the Rigveda who is recorded asking out loud for what she wants is called Noise. For the fuller picture of how these voices were preserved and then flattened, see the hidden women of the Rigveda.

Names that are deliberately withheld

All of this rests on an assumption the Rigvedic poets did not share: that a name is a label attached to a thing. In Vedic, nāman (Sanskrit: नामन्) covers more ground than English “name.” It reaches into essence and mode of appearance, the form under which something presents itself. Which is why the text can do something that makes no sense for a label. It can hide one.

bṛhaspatir amata hi tyad āsāṃ nāma svarīṇāṃ sadane guhā yat

“Bṛhaspati grasped that name of the lowing cows, the one that was hidden in their place.”

Rigveda 10.68, verse 7. After Griffith (1896), diacritics standardised.

The cows in the cave have a secret name, and finding it is what wins them. At RV 8.41.5 Varuṇa is praised as the one who “knows the hidden secret names of the morning beams.” At RV 9.75.2 the son establishes “the third secret name of mother and father in the bright realm of heaven.” Frank Köhler, who has worked through these passages recently, puts his finger on the paradox: the secret names are mentioned repeatedly and never once quoted [9]. The text tells you they exist and withholds the content, permanently, which is a fairly sophisticated thing for a Bronze Age poem to do on purpose.

Set that theology beside the donor lists and the donor lists stop looking like accounting. If a name is an essence, then putting Balbūtha’s name into a hymn addressed to Vāyu is not a footnote about a transaction. It installs him inside the ritual, in a text that will be recited unchanged for three thousand years. The poets were not being sloppy when they let outsiders into the record. They were paying in the only currency they minted.

The control experiment: names outliving their language

There is one place where we can check whether onomastics really carries history, because the language itself did not survive.

In fourteenth-century BCE northern Syria, the kingdom of Mitanni was ruled by a dynasty with names like Tušratta, Artatama, Šuttarna and Šattiwaza. The population spoke Hurrian. The state archives are in Akkadian. Nobody there was writing Indo-Aryan. But the royal names keep resolving into Indo-Aryan elements, the treaty between Mitanni and the Hittites invokes Mitra, Varuṇa, Indra and the Nāsatyas in that order, and a Hurrian horse-training manual by a man named Kikkuli counts laps with aika-vartanna, “one turn,” where aika is the Indo-Aryan word for one rather than the Iranian [10].

Paul Thieme’s 1960 study of the Mitanni treaty gods remains the reference point for how carefully this evidence has to be handled, and how much it will bear when handled carefully [10]. The lesson for reading the Rigveda is direct. Names are the most portable and the most conservative thing a population carries. They survive language shift, they survive script change, they survive the loss of everything else. When Vaśa Aśvya wrote down Balbūtha, he was doing, without meaning to, the same thing the Amarna scribes did for Tušratta. For the theological side of that Mitanni evidence, see Mitra, the contract god.

What the list is worth

Three things, in ascending order of how much they change the picture.

First, a correction of scale. The Rigveda is usually read as a book about gods. Run your eye down the names instead and it turns into a book about a few hundred specific people: kings, poets, carpenters, merchants, a woman who won a chariot race, a boy tied to a post, a blind bard, a man with camels on the Iranian frontier. The theology is what they were doing. The names are who was there.

Second, a check on the ethnic reading. The Rigveda contains fierce hostility to outsiders and it also contains their names in its thank-you lists. Both are in the text. Any account that keeps one and discards the other is editing.

Third, and this is the part that keeps me returning to it: the Rigveda preserved this material by accident, in a system built for the opposite purpose. Everything we know about Vedic transmission points to obsessive control, syllable by syllable, generation after generation, through the padapāṭha and the memorised schools. And that control is exactly why the accident holds. A society that could not bear to lose a single accent could not lose Balbūtha either. Fidelity does not discriminate. It preserved the theology the poets cared about and the sociology they did not, in the same breath, at the same resolution.

If you want to see it yourself, do not start with a famous hymn. Open RV 8.46 at verse 21 and read the dānastuti to the end, slowly, the way you would read a ledger, noticing every proper noun. Then read RV 6.45.31 to 33 and ask yourself who Bṛbu was, and why a Bharadvāja poet was willing to say his name out loud in the middle of a hymn to Indra. The Rigveda will not tell you. It only tells you that he was there, and that he paid.