Two Bays, a Donkey and a Team of Goats: How the Rigvedic Gods Got Around

When Is the Yoking of Your Donkey?

In the middle of a hymn to the Aśvins, a poet the tradition names Hiraṇyastūpa of the Aṅgirasa line stops praising and starts asking about equipment. He wants to know where the three wheels of their three-part chariot are, where the three seat-rails are that nest together like eggs, and then this:

kva trī cakrā trivṛto rathasya kva trayo vandhuro ye sanīḷāḥ | kadā yogo vājino rāsabhasya yena yajñaṃ nāsatyopayāthaḥ

“Where are the three wheels of your threefold chariot, where the three seat-rails that share one nest? When is the yoking of your prize-winning donkey, with which, Nāsatyas, you come to the sacrifice?”

RV 1.34, verse 9. My rendering from the Sanskrit; compare Griffith (1896). [2]

A donkey. Not a symbolic donkey, not a donkey standing in for something nobler: a rāsabha, the animal that carries grain sacks, harnessed to the chariot of the twin gods of rescue, and described elsewhere as a vājin, a prize-winner, because it once won a thousand in a race (RV 1.116.2).

Anyone who has seen a Hindu temple knows what a god’s animal is supposed to look like. Śiva has his bull, Viṣṇu his eagle, Durgā her lion, Gaṇeśa his mouse, Sarasvatī her goose, Agni his ram, Indra his elephant. The animal sits underneath, the god sits on top, and the pairing is so standard that it has a technical name: vāhana (वाहन), “vehicle”. Go looking for that arrangement in the Rigveda and it is not there. Not in an early form, not in embryo, not anywhere in a thousand hymns. What is there instead is a transport system described with the attention of someone who has repaired one: poles and yokes, reins that must be well made, a whip, a goad, an outrunner horse harnessed beside the pair, teams that range from two animals to a thousand.

Nobody in it sits on an animal. The gods drive.

That single observation turns out to open onto three others: a way of dating the text by its tack, a theology in which the animal is something a god is rather than something a god rides, and the oddest fact of all, which is that the word vāhana does occur in the Rigveda, eighteen times, and in every one of them it means the one who carries rather than the one carried on.

The Divine Motor Pool

Start with the inventory, because it is more specific than you would expect from liturgy. Each major god has a team, and the team is not interchangeable.

Indra’s standard team is a pair. The word is harī, “the two tawny ones”, a grammatical dual appearing in roughly 135 verses, and he is the god of the two-horse racing chariot, which in the late second millennium BCE was the most expensive object a man could own. The plural harayaḥ turns up in another fifty-odd verses, and when a poet of the Gṛtsamada family wants to flatter Indra he does it by arithmetic, inviting him to come with two bays, then four, six, eight, ten, then twenty, thirty, forty, and on up to a hundred (RV 2.18.4 to 6). That is a joke about hospitality rather than a stable record; the dual is the default, and the hyperbole only works against it. For what the Rigveda does with numbers of that kind, see the counting hymns.

Sūrya’s team is seven, and they are mares: haritaḥ, the tawny ones, feminine, a word that appears in about thirty-two verses and mostly means his. The number is fixed early and never negotiated, which is why the seven horses of the sun are still carved on temple walls three thousand years later.

ayukta sapta śundhyuvaḥ sūro rathasya naptyaḥ | tābhir yāti svayuktibhiḥ

“The sun has yoked the seven bright ones, the daughters of his chariot; with them, yoked by themselves, he goes.”

RV 1.50, verse 9. My rendering from the Sanskrit; compare Griffith (1896). [2]

Vāyu, the wind, has the largest fleet and the strangest word for it. His teams are niyut, a term that turns up in about 57 verses spread over 42 hymns and that belongs to him in the way the bays belong to Indra. He is invited with a thousand of them, with hundreds of them, with a thousandfold chariots (RV 1.135.1; RV 2.41.1). The god who is everywhere at once needs a motor pool, not a pair.

The Maruts drive pṛṣatī, dappled mares, in sixteen verses, and one line gives away how closely the poets were watching real chariots: “you have yoked the dappled mares to your chariots; a red one draws as the outrunner” (RV 1.39.6). The word is praṣṭi, usually understood as the third horse harnessed alongside a yoked pair rather than under the yoke itself. [8] That is a detail from the yard, not from the imagination. On the Maruts as a troop, see the storm band.

Then the teams get eccentric. Agni’s horses are rohita and aruṣī, red and ruddy, which is a way of saying that his horses are his flames (RV 1.14.12). Uṣas yokes ruddy cows (RV 1.92.2). Pūṣan, in five verses, is ajāśva, a compound that means “having goats for horses”, and his goats are explicitly asked to draw his car: “let the goats bear god Pūṣan in his chariot” (RV 6.55.6). He also carries the one goad in the pantheon, a slack one that he keeps lifting (RV 6.58.2), which is exactly the implement a goatherd uses and not what you would give a sky god. His work on the roads is gathered in the lord of paths.

The Aśvins are the loose end. Their chariot is drawn by horses, by the donkey, by buffaloes, and by birds, and Macdonell noted the variety more than a century ago. [3] The birds are not a flourish: they are harnessed.

ā vāṃ śyenāso aśvinā vahantu rathe yuktāsa āśavaḥ pataṃgāḥ

“Let eagles bring you, Aśvins, swift winged ones yoked to your chariot.”

RV 1.118, verse 4. My rendering from the Sanskrit. [2]

Wings, in this text, get a yoke put on them. That is the instinct worth noticing, and it is the thing that later Indian art abandons.

Methods note. The verse counts here come from my own scan of the romanised Saṃhitā text of all ten maṇḍalas held in this project’s corpus, which contains 10,524 verses; the standard printed counts differ slightly depending on how the Vālakhilya hymns are handled. I matched on word-initial stems token by token, which undercounts rather than overcounts, because Vedic sandhi fuses words together and buries a stem inside a longer token. Treat every figure as “roughly this many verses”. Grassmann’s 1873 lexicon remains the authority for exact occurrence lists. [7]

A Driver’s Vocabulary

The Rigveda’s horse words are not vague. A text that mentions horses in around 300 verses has a working vocabulary for the parts of the vehicle and the gear that controls the animal, and the vocabulary is consistently that of a driver.

The verse in that first row is one of the hardest in the collection. RV 1.162.21 packs Indra’s bays, the Maruts’ dappled mares and the Aśvins’ donkey into a single line, puts a prize-steed at the donkey’s pole, and leaves translators from Geldner to Jamison and Brereton disagreeing about who is standing where. [1][6] The whole divine motor pool, jammed into one verse, and nobody can quite park it.

Two other things follow from the table. The first is that the gods mostly drive themselves. Sārathi, the charioteer who holds the reins so that the warrior beside him can shoot, appears in six verses, and the Maruts are told, in a line I find hard to read without smiling, to get on with it themselves: “self-yoked, the immortals came down from heaven at will; with the whip, drive yourselves” (RV 1.168.4).

The second is the shape of the hole. There is no Rigvedic word for a rider. The later Sanskrit terms, sādin and aśvārohin, are simply absent, and so is any word for a saddle or a girth. The verbs of mounting, ā ruh and adhi sthā, take the chariot as their object almost everywhere: you climb onto the car, not onto the horse. The text is not short of interest in horses. It is short of one posture. Almost.

The Handful of Lines Where Somebody Rides

There is an exception, and it is one of the most physical passages in the collection. In the fifth maṇḍala, a poet the tradition calls Śyāvāśva of the Atri family asks the Maruts how they travelled, and then answers his own question:

kva vo ‘śvāḥ kvābhīśavaḥ kathaṃ śeka kathā yaya | pṛṣṭhe sado nasor yamaḥ

jaghane coda eṣāṃ vi sakthāni naro yamuḥ | putrakṛthe na janayaḥ

“Where are your horses, where the reins? How were you able, how did you travel? The seat was on the back, the control at the nostrils. The goad was at their haunch; the men spread their thighs apart, like women in childbirth.”

RV 5.61, verses 2 to 3. My rendering from the Sanskrit; compare Griffith (1896) and Müller’s version in Sacred Books of the East 32. [2][5]

That is a description of men astride horses, written by someone who had watched it happen and reached for the only comparison available for a posture nobody in his audience associated with dignity. The control is at the nostrils, not through a jointed bit in the mouth. The seat is on the back, and the word is sadas, “seat”, not a technical term for a saddle, because there is no technical term.

One more passage may belong here. The hymn that follows the horse-sacrifice liturgy praises the one “who first mounted the courser” (RV 1.163.9), with adhi sthā taking the animal as its object rather than the chariot, which is the grammar of riding. A third candidate, the line in the sacrificial hymn about a horse hurt “with the heel or the whip” (RV 1.162.17), is often read that way too, though the verse is obscure and I would not lean on it. On the ritual those two hymns belong to, see the horse sacrifice.

So riding is known. It is known in roughly three passages out of some ten and a half thousand verses, and in none of them is it how a god travels, how an army fights, or how anything gets carried. Macdonell and Keith made the narrower version of this point in the Vedic Index over a century ago: the Rigveda does not show men riding horses into battle. [4]

What makes that silence informative rather than merely odd is that it matches the hardware. Robert Drews, working from bits, bones and art rather than from texts, argues that effective mounted warfare does not appear before roughly the eighth century BCE, and that the enabling technology was the bronze jointed snaffle bit, which shows up in the Caucasus around 1000 BCE and spreads north from there. [10] David Anthony’s account of the spoked chariot puts its invention in the Sintashta culture of the southern Urals around 2000 BCE. [9] The Rigveda sits in the gap: after the chariot, before the cavalry. A text composed in that window would talk about poles and teams constantly and about riding almost never, which is exactly what this one does. The chariot as a dating instrument is treated at length in the spoked wheel piece.

Aside. Not everyone accepts that the Rigvedic chariot should be read as literally as this. Michel Danino has argued that the text’s horses and chariots carry so much symbolic weight, standing for speed, light and energy, that scholars have over-read a military and technological reality into what is largely ritual imagery. [20] The objection has some force against crude uses of the evidence, and none, I think, against the vocabulary itself. Nobody invents a word for the third horse harnessed beside a yoked pair in order to symbolise energy.

Fast Car, Slow Cart

The Rigveda distinguishes two kinds of vehicle, and the asymmetry between them is almost comic. A form of ratha, the light horse-drawn chariot, appears in roughly 610 verses, about one verse in seventeen, or $610/10{,}524 \approx 5.8\%$ of the text. The word for the ox-drawn cart, anas, appears in three.

Those three are worth their weight. One is in the wedding hymn, where the bride travels not by chariot but by cart:

“Her spirit was the bridal car; the covering thereof was heaven: bright were both steers that drew it when Sūryā approached her husband’s home.”

RV 10.85, verse 10, trans. Griffith (1896). [2]

The two “steers” are anaḍvāhau, draught oxen, a word that occurs twice in the whole collection. The sun’s daughter gets a wagon, and the wagon is built out of her mind and roofed with the sky, which is the poet making something splendid out of a slow vehicle. More on that liturgy in the wedding hymn.

The second occurrence is a wreck. Indra, in a story the poets tell with no apparent discomfort, smashes the cart of the dawn goddess: “there lies her cart, well shattered, in the Vipāś; she has fled far away” (RV 4.30.11). The dawn drives the slow vehicle, and the thunder god runs her off the road.

The third kind of load-bearing is the strangest. In the hymn over the weapons, there is a rathavāhana, a “chariot-carrier”: the cart on which the chariot itself was transported when it was not being raced. [8] A society that moves its racing machine around on a wagon is a society with a clear sense of which vehicle is for show. The rest of that hymn, bowstring and armour and all, is read closely in the weapons piece.

The Animal Is a Predicate, Not a Seat

Here is where the Rigveda does something that the later tradition undoes. The animals that eventually end up under the gods are, in these hymns, animals the gods are.

Indra is a bull. Vṛṣabha and its relatives turn up in about 175 verses, and a great many of them are describing him. He is also, in the opening line of a hymn the tradition gives to Savya, a ram: “that ram, much invoked” (RV 1.51.1), the animal that later belongs under Agni. Rudra, whose later form stands beside a bull, is in the Rigveda “the ruddy boar of heaven, the one with braided hair” (RV 1.114.5), and the boar is later Viṣṇu’s avatar rather than anyone’s mount. The pattern is apposition: the god and the animal occupy the same grammatical slot. Nobody is on top of anybody.

Agni makes the point most sharply, because he is both the animal and the vehicle:

vṛṣo agniḥ samidhyate aśvo na devavāhanaḥ | taṃ haviṣmanta īḷate

“Agni is kindled, a bull, like a horse that conveys the gods. Him the offerers invoke.”

RV 3.27, verse 15. My rendering from the Sanskrit. [2]

Devavāhana, “god-conveying”. The epithet that later names the beast beneath the deity is used here of the deity who does the hauling. Agni is also vahni, “the conveyer”, in about 44 verses, because his job is to carry the offering upward. The later system would take that word, keep the sense of carrying, and hand it to the animal.

Notice that Sūrya is the exception that proves the rule. He is the one god whose Rigvedic arrangement survives intact into the temple, with seven horses and a chariot, because his vehicle was never the problem that needed solving.

Yoked by Speech

If the Rigvedic gods do not ride, and if their teams are often made of flame or wind or birds, what actually gets them moving? The text answers that question in a set of compounds I find the most interesting words in the whole inventory.

Indra’s horses are vacoyuj, “yoked by speech”, in five verses. They are brahmayuj, “yoked by the sacred formula”, in five more. They are manoyuj, “yoked by thought”, in seven, and Vāyu’s are too. The Ṛbhus, the artisan gods, are said to have fashioned Indra’s bays with their minds (RV 1.20.2); on those craftsmen, see the artisan gods. And in one line a poet of the Viśvāmitra family says plainly what he thinks he is doing when he recites:

brahmaṇā te brahmayujā yunajmi harī sakhāyā sadhamādā āśū | sthiraṃ rathaṃ sukham indra adhitiṣṭhan prajānan vidvān upa yāhi somam

“With my formula I yoke your two formula-yoked bays, your companions, your drinking-mates, the swift pair. Mounting the firm, easy chariot, Indra, knowing the way, come to the soma.”

RV 3.35, verse 4. My rendering from the Sanskrit; compare Griffith (1896). [2]

Brahmaṇā … yunajmi: with my verse, I yoke. The poet is not describing the harnessing from outside. He is claiming to be the harness. Indra, correspondingly, is brahmavāhas, “conveyed by the formula”, in eight verses, and the singers call themselves stomavāhas, “praise-conveyers”, in five more.

Which brings us back to the eighteen verses that contain vāhana. Every one is a compound of the same shape: havyavāhana, “oblation-conveying”, which is Agni about a dozen times over; kravyavāhana, “corpse-offering-conveying”, the fire of the dead (RV 10.16.11); vasuvāhana, “goods-conveying”, of the Aśvins’ chariot; madhuvāhana, “honey-conveying”, of the same vehicle at dawn; and the rathavāhana wagon. Not one of them is an animal under a god. The Rigvedic vāhana is whatever carries the sacrifice: the fire, the chariot, and in the end the hymn itself.

Aside. This is a semantic shift, not an etymological accident. Both the Rigvedic compounds and the later noun come from vah, “to convey, to draw”, the root behind Latin vehere and English vehicle. [14] What changes is which end of the relationship the word names. In the Rigveda the conveyer is divine. In classical Hinduism the conveyer is a beast, and the god is cargo.

When the Mount Arrives

So when do the animals get under the gods? Not in the Vedic corpus, and the datable evidence for the change is firmer than you might expect.

The sequence that the material evidence suggests runs in stages, and the stages are not the same thing as a mount.

graph LR
    A["Animal as predicate (Indra is a bull)"] --> B["Animal as team (yoked to the chariot)"]
    B --> C["Animal as standard (Garuda on a pillar)"]
    C --> D["Animal as companion (Kushan coins)"]
    D --> E["Animal as mount (the vahana)"]

Three things are worth separating out in that table. The Besnagar pillar gives us Garuḍa in the right company a century before the Common Era, but as a flag, which is how Viṣṇu’s bird first functions. [17][19] The Kushan coins give us an animal standing with a god rather than beneath him. Only the cave reliefs give us the arrangement a modern worshipper would recognise, and by the sixth century it is everywhere.

timeline
    title Transport technology and the gods who used it
    c. 2000 BCE : Spoked chariot in Sintashta burials
    c. 1500 to 1100 BCE : Rigvedic hymns, teams and poles
    c. 1000 to 800 BCE : Jointed snaffle bit, cavalry becomes effective
    c. 113 BCE : Garuda standard at Besnagar
    c. 2nd century CE : Oesho and bull on Kushan coins
    c. 6th century CE : Indra on the elephant at Badami

Why the change happened is a harder question, and here I am inferring rather than reporting. Three pressures seem plausible, and they are not exclusive.

The first is technological. Divine transport tracks elite human transport with a lag. When the chariot was the prestige vehicle, the gods drove; by the time the Mauryan armies were fielding cavalry and war elephants, a god in a two-horse racing car looked archaic. The elephant is nearly absent from the Rigveda, where even hastin, “the one with a hand”, is a simile rather than a species (RV 1.64.7); the missing animals of the Rigvedic fauna are surveyed in the beast with a hand. Indra’s elephant is not a forgotten Vedic detail. It is a later recruit.

The second is iconographic, and it is the one I find most persuasive. The Rigvedic religion has no images and no temples, a point treated in the missing architecture. The moment you have to carve a god, you need an attribute that fits in stone beside a standing figure. A galloping pair with a pole, a yoke, reins and an outrunner is a scene. One animal under one god is a sign. The vāhana is a solution to a compositional problem that the poets never had, because their medium was sound.

The third is liturgical. Temple festivals move the deity through the streets, and a processional image needs something to be mounted on. Nanditha Krishna’s survey of sacred animals in India notes how thoroughly the mounts became objects of veneration in their own right, which is what happens to a convention once it is doing real ritual work. [18]

Aside. What cannot be concluded from any of this is that the Rigvedic poets would have disapproved. They were perfectly willing to put a god on a goat-cart. The point is narrower and more interesting: the arrangement that now looks like the most natural thing in Indian religion is a historical answer to a historical problem, and it is young enough that we can watch it arrive.

What the Harness Knows

Read the Rigveda for its tack and it stops being a timeless book. The vocabulary of poles and reins and outrunners belongs to a specific few centuries, and the vocabulary that is missing belongs to the centuries after. You can date a text by what it has no word for.

The deeper find, though, is the relationship the harness implies. A god on a mount is above the animal and above you. A god in a chariot is in the posture of a driver, reins in hand, and the Rigvedic poets put themselves at the same end of the equipment: they claim, in those brahmayuj and vacoyuj compounds, that their verses are what yokes the horses. The relation between the human and the divine in these hymns is horizontal and contractual rather than devotional, which is why the text is so full of bargaining, friendship and complaint. For the friendship, see sakhya; for the yoke as the root of a much later discipline, see yoga before yoga.

Two reading recommendations. Open RV 5.61 and read the first four verses slowly, noticing that the poet has to invent a comparison for the posture of a man on a horse because his language does not have a word for it. Then go and look at a stone Nandi in a temple courtyard, sitting patiently in front of a sanctum, waiting to be ridden. Between those two images lies the better part of two thousand years, a change in military technology, and the invention of the religious image.

The gods did not always sit on animals. For a long time they drove.