The One-Footed Goat and the Serpent of the Deep: The Rig Veda's Gods Without Stories

Fifteen Verses of Wishing Well

RV 7.35 is a hymn that does only one thing, and does it fifteen times. Every line begins with the same word, śám, which means something like “to our benefit, without harm.” Benign to us be Indra and Agni. Benign the Creator, the Sustainer, broad Earth and Heaven, the pressing stone. Benign the sun rising wide-eyed, the four directions, the steadfast mountains, the waters. Benign the horses, the cattle, the skilful Rbhus. The poet is working through an inventory of everything a household in the Punjab could plausibly want on its side, and the effect after a dozen verses is less like worship than like a man checking that every door in the house is locked.

Then verse thirteen:

Benign to us be the god Aja Ekapād, benign the Serpent of the Deep, benign the sea.

Benign to us be the Child of the Waters, the swelling one; benign be Pṛśni, who has the gods for her herdsmen.

(RV 7.35.13, translated from the Sanskrit)

Aja Ekapād means, on the face of it, “the one-footed goat,” though it may equally mean “the one-footed unborn one.” The Serpent of the Deep is Ahi Budhnya, a snake associated with the bottom of something. No hymn in the Rig Veda is addressed to either of them. No story is told about either of them. No poet describes what they look like, where they live, what they did, what they are for. Between them they occur seventeen times in roughly ten and a half thousand verses, and in every single case they are items in a list.

That is the whole dossier. What makes it interesting is not that we have lost the explanation. It is that the loss appears to be much older than we are. By the time the first Indian philologist sat down to explain the Rig Veda word by word, some centuries before the Common Era, he was already guessing, and he said as much. The gods with no stories are a natural experiment in what a tradition does with a piece of itself it can no longer read.

The Arithmetic of Obscurity

Run the two names through the Saṃhitā text and the distribution is almost comically tidy. Ahi Budhnya appears as a divine name in twelve verses. Aja Ekapād appears in five, and in every one of those five the serpent is standing next to him. They are, in the Rig Veda as we have it, never separated. For a text that pairs its gods constantly, that is a striking exclusivity, and it is a different kind of pairing from the grammatical fusion that produces Mitrāvaruṇā and Dyāvāpṛthivī (see the Rigveda’s dual deities). These two are not fused. They are simply always adjacent, like two names on a list nobody has reordered.

The list is the point. Eleven of the twelve occurrences fall inside hymns the tradition assigns to the Viśve Devāḥ, the All-Gods, a genre whose organising principle is completeness rather than devotion. The twelfth sits in a hymn to Savitṛ, in a verse that is itself a roll call. Not one of the seventeen occurrences is in a hymn to Indra, or Agni, or Soma, or any of the deities who get sustained poetic attention.

Watch how the pair actually enters a verse:

samudráḥ síndhū rájo antárikṣam ajá ékapāt tanayitnúr arṇaváḥ

áhir budhnyàḥ śṛṇávad vácāṁsi me víśve devāsa utá sūráyo máma

“The sea, the rivers, the dusty space, the midair, the One-Footed Goat, the thunder, the flood: let the Serpent of the Deep hear my words, and all the gods, and my patrons.”

Rigveda 10.66, verse 11. Translated from the Sanskrit; diacritics standardised.

That is not a theological statement. It is a manifest. Aja Ekapād (Sanskrit: अज एकपाद्) is wedged between the midair and the thunder with no more ceremony than a comma, and Ahi Budhnya (अहिर् बुध्न्यः) picks up the verb at the start of the second line as though the grammar had simply been waiting for a subject. The poet is assembling a crowd large enough that nothing important can feel excluded, which is also, incidentally, the logic behind the famous Rigvedic habit of counting the gods at three thousand and thirty-nine.

Read down the right-hand column and something surfaces that no summary of Vedic religion will tell you. Of everything a Rigvedic poet can ask a god for, and the repertoire is large, running from cattle and sons to victory in battle and a hundred autumns of life, these two are asked for exactly two things. Hear us. Do not hurt us.

That is the entire theology of the unknown god: attention and restraint. You do not ask a power you cannot describe for anything specific, because you have no idea what it is in a position to give. You ask it to notice you, and you ask it to hold off.

Methods note. The counts above come from searching the Saṃhitā text for the name forms directly, then removing the cases where budhnya is an ordinary adjective rather than part of the divine name: RV 4.55.6 has “the serpent of the depth” in the instrumental as a description, RV 7.6.7 has budhnyā vásūni, “goods of the deep,” and RV 7.56.14 applies the adjective to the Maruts’ glories. Published tallies vary by a verse or two depending on how those borderline cases are treated. Keith’s count of twelve for Ahi Budhnya, made without a computer, matches mine exactly.[1]

Transparent Words, Opaque Referents

Here is the part that makes these two names genuinely unusual rather than merely rare. They are not foreign words. They are not corrupt. Every element in them is common Vedic vocabulary with clean Indo-European ancestry. You can parse the names completely and still have no idea what is being talked about.

Áhi is a snake, and one of the better-attested inheritances in the family: Avestan aži, Greek óphis, Latin anguis. Budhná is the bottom, base or root of a thing, the floor of a valley, the butt of a tree, and it belongs to the same set as Greek puthmḗn, Latin fundus, and, by a long road, English bottom.[2] The derived adjective budhnyà, “belonging to the depth,” is alive and working in the Rig Veda outside the divine name, which is how we know the poets still understood the word. Nobody was reciting a fossil syllable here. They knew they were saying “the snake that has to do with the bottom.” They simply do not tell us the bottom of what: the sea, the atmosphere, the sky, the world, the sacrificial ground.

The goat is worse, because Sanskrit hands you two words spelled identically. Ajá is a he-goat, cognate with Lithuanian ožys and Avestan azī. But a-já is also “unborn,” the negative prefix plus the root jan, “to be born,” and that second reading turns the name into something metaphysical rather than barnyard. The ambiguity is not a modern invention. The Śvetāśvatara Upaniṣad later builds an entire verse on exactly this pun, describing the “one unborn she-goat” of three colours, and it is impossible to read that passage without sensing that the poet is enjoying the double exposure.

Two observations follow. The first is that etymology has limits people rarely acknowledge. Knowing what every morpheme means gets you a literal gloss, not a referent, in the same way that knowing what “morning” and “star” mean does not tell you that the phrase points at Venus. The second is that this failure is unlike the more famous kind of Rigvedic obscurity, the words with no Indo-European ancestry at all that probably entered the language from populations already in the northwest (discussed in the Rigveda’s foreign words). Those are opaque on the outside. These are opaque on the inside.

Yāska Was Already Guessing

Somewhere between the fifth and third centuries BCE, a scholar named Yāska produced the Nirukta, a commentary on an older word list called the Nighaṇṭu. It is the earliest surviving attempt anywhere in the world to explain a scripture’s vocabulary systematically, and it exists because the Rig Veda had already become difficult. The language had moved on. Words that had been ordinary were now technical or unintelligible, and the recitation schools were transmitting sounds with a shrinking grip on sense.

The Nighaṇṭu sorts the gods by region: earth, midspace, heaven. Ahi Budhnya it puts in the middle region, the atmosphere, which is what you would expect from a serpent of the depths of the air. Aja Ekapād it puts among the celestial gods, which fits badly with the Rigvedic verses themselves, where he keeps company with the sea, the rivers and the thunder. Yāska, working through the deity chapters, glosses ajá not as “goat” and not as “unborn” but as “driving,” from a third root, and explains ekapād with the flat literalism of a man who has nothing else: he who has one foot.[3] His commentator Durga, much later, decides the whole thing means the sun.

Look at what has happened there. The earliest expert reader in the tradition, working with access to living recitation, priestly practice and texts now lost, produces for these names a derivation nobody has since accepted and a gloss that explains nothing. He is not being careless. He is being honest about a text that has outlived its own referents.

Yāska’s book preserves something even more startling. He records, in order to argue against it, the position of a predecessor named Kautsa, who held that the mantras of the Veda are anarthaka: meaningless. Kautsa’s argument, as Yāska transmits it, is a serious one built on evidence that the mantras are obscure, contradictory, and ritually effective whether or not anyone understands them. Yāska answers at length, and the exchange has been called a turning point in the history of Vedic interpretation.[4] What matters for our two gods is that the question was live. In the centuries when the Rig Veda was becoming India’s most protected text, a school of thought existed which said out loud that the words might no longer mean anything, and the tradition’s most influential philologist felt obliged to answer it.

Aside. The deity assignments in the table above come from the Anukramaṇī, the traditional index that names a poet, a metre and an addressee for every hymn. It is old and it is useful, but it is a later systematisation, not the poets’ own labelling, and on contested hymns it is frequently a guess. Treat it as evidence about how the tradition organised itself rather than as an authority on what a hymn is doing. The same caution applies to the classification of hymns by stratum, which is what makes the arrangement of the collection its own puzzle: see how the Rigveda was sorted.

A Century of European Guesses

When European Indology took up the problem in the nineteenth century, it brought comparative method, enormous erudition, and a conviction that a nature myth lay behind every Vedic name. The result was a set of proposals that agree on nothing except that the answer must be an atmospheric or celestial phenomenon.

Macdonell’s lightning reading is the most quoted, and it has a certain elegance. A goat leaps between the cloud-mountains; lightning has one leg and it touches the earth once. Set against the actual verses, though, it has nothing to stand on. Not one of the five passages says anything about lightning, or about leaping, or about striking. They say: may he hear us, may he be benign. The reading is an inference from the name alone, which is to say it is a guess of the same species as Yāska’s, made two and a half millennia later with better comparative data and no better evidence.

Keith, who was as sober as Indology got, allowed himself one observation with real force. Twice, in the same words, the serpent is asked not to hand his worshippers over to harm:

mā́ no ‘hír budhnyò riṣé dhād

“Let not the Serpent of the Deep deliver us to injury.”

Rigveda 7.34, verse 17, with the same formula at 5.41.16. Translated from the Sanskrit.

You do not say that to a god you consider safe. The phrasing preserves a hint of menace, and it is the only thing in the Rigvedic material that looks like a character trait.[1] Whether that makes him a domesticated relative of the dragon Vṛtra, as some have argued, or a benign water-power whose danger is merely the ordinary danger of deep water, the text does not say.

What can we actually conclude? Three things, and the third is the useful one.

1. The names are transparent and their referents are not recoverable from the Rig Veda.
2. Every identification on offer, ancient and modern, is derived from the name rather than from the verses, which means the evidence is the same for all of them and cannot decide between them.
3. The pattern of the evidence is itself the finding. A god who appears only in catalogues, only with one companion, and only as the object of “hear us” and “do not harm us” is not a god who has been badly preserved. He is a god who was already preserved in that form.

Griffith, Three Ways

There is a small piece of evidence for how disorienting these names are that anyone can check in an afternoon. Ralph Griffith’s complete English Rig Veda of 1896 was the version through which most of the anglophone world met the text. Griffith had to render áhir budhnyàḥ twelve times. He did it three different ways.

In eight places he translates it: “the Dragon of the Deep.” In four he gives up and transliterates: “Ahibudhnya.” In RV 10.64.4 he splits the difference, producing “Ahi of the Deep,” which is neither. Nothing in the Sanskrit differs across those passages. The phrase is identical. What varies is the translator’s nerve.

May Aja-Ekapād and Ahibudhnya, and Earth and Ocean hear our invocation;

All Gods who strengthen Law, invoked and lauded, and holy texts uttered by sages, help us.

(RV 6.50.14, Griffith 1896)

Compare that with his handling of the same pair fifteen verses later in another book, where the serpent becomes a Dragon and the goat keeps his hyphen, and you can watch a translator negotiating in real time with a name he cannot picture. It is a useful reminder that a translation is a series of decisions under uncertainty, and that the smooth surface of a printed page hides the places where the translator was as lost as you are.

What Survived Was the Name

Now the strange part. Two gods with no myth, no hymn and no agreed identity did not fade out. They propagated.

The Yajurveda tradition, building the ritual calendar, needed a presiding deity for each of the twenty-seven lunar mansions. Two of those mansions are Pūrva Bhādrapadā and Uttara Bhādrapadā, the two halves of the Great Square of Pegasus. The deity assigned to the first is Aja Ekapād. The deity assigned to the second is Ahi Budhnya. The list is given in the Taittirīya Brāhmaṇa, and it is the same list that still runs in Indian almanacs: a couple named in a wedding calculation in 2026 is being matched against a one-footed goat and a serpent of the depths whose meaning was already gone when the Brāhmaṇas were composed. For the wider story of how the mansions were organised, see the Vedic nakshatra calendar.

The Purāṇic tradition went further. When the eleven Rudras had to be enumerated, both names were pressed into service as members of the group. And then, at the far end of the line, a medieval Vaiṣṇava school of the Pāñcarātra produced a substantial theological treatise, some sixty chapters on cosmology, mantra, yoga and the philosophy of divine power, and named it the Ahirbudhnya Saṃhitā, the compendium of the serpent from the depths, on the premise that its teaching was revealed by him.[5] A god about whom the Rig Veda says nothing at all ended up with a book.

graph LR
    A["RV: catalogue verses"] --> B["Nighantu: regional lists"]
    B --> C["Yaska: Nirukta 12.29"]
    A --> D["Taittiriya Brahmana"]
    D --> E["Two lunar mansions"]
    E --> F["Modern almanacs"]
    C --> G["Puranic eleven Rudras"]
    G --> H["Ahirbudhnya Samhita"]
    G --> I["Ekapada Shiva images"]

The sculpture is the last stage. From roughly the seventh century onward, temples in Odisha and elsewhere carry a form of Śiva standing on a single leg, the Ekapāda-mūrti, which iconographic handbooks describe in detail.[6] Whether the sculptors were thinking of the Rigvedic name or had arrived at a one-legged god by another route is genuinely disputed, and the honest answer is that a gap of well over a thousand years sits in the middle of the argument. What is not in dispute is that the image exists and that the name attached to it is the Rigvedic one.

Why a Canon Keeps What It Cannot Use

The reflex, when a religious tradition preserves something it no longer understands, is to call it dead weight. That gets the mechanism backwards.

The Rig Veda was transmitted by a method designed to make the sound of the text invulnerable to the comprehension of the person reciting it. Syllable order, pitch accent and metre were locked down by drilling the verses in permuted sequences, forwards, backwards, and in interlocking pairs, so that an error would break the pattern audibly (this machinery is the subject of the Rigveda’s oral transmission). A system like that has a property worth stating plainly: it preserves meaningless material exactly as faithfully as meaningful material, because it is not tracking meaning at all. Aja Ekapād survived for three thousand years for the same reason a name in a legal document survives a change of ownership. The container did not care.

The all-gods hymn is the second half of the mechanism. A poet composing for Indra needs to know what Indra does, because the hymn is an argument that Indra should do it again. A poet composing a Viśve Devāḥ hymn needs only a list long enough to leave nothing out. That genre is where the pantheon keeps its long tail: the abstractions like Anger and Trust that the poets promoted to divinity (see the Rigveda’s abstract deities), the half-remembered figures like Trita Āptya who clearly once had a story (see Trita and the bad dream), and these two, who may never have had one in this language at all. The insurance policy became a museum by accident.

There is a third thing here, and it is the one I keep returning to. The Rigvedic poets had a working posture toward the unknown, and it is legible in the verb census: hear us, do not harm us. They did not resolve the obscurity by inventing a myth to cover it, which is what a later and more systematising literature would do, and did. They did not drop the name either. They kept it in the list, addressed it in the only terms that made sense for something undescribed, and moved on to the next line. That is a more interesting relationship with ignorance than most traditions manage, and it sits oddly with the popular image of the Veda as a body of total and unbroken knowledge. The text is at its most confident about Indra’s fights and at its most careful exactly where it cannot see.

Open RV 7.35 and read it straight through, aloud if you can, letting the śáṁ naḥ build. Somewhere around verse eleven you stop registering the individual names and start hearing the shape of the thing, which is a man methodically asking the whole world not to turn on him. Then verse thirteen arrives with two powers in it that neither he nor you can identify, and they get the same request as the sunrise and the mountains. That is not a failure of the hymn. It is the hymn working as designed.