Seven Cups, Seven Jobs: The Org Chart Buried in the Rig Veda

Drink from the Neṣṭṛ’s cup

In the thirty-seventh hymn of the second book, a god is asked to drink six times, and each time from a different cup.

Stanza one hands him the Hotṛ’s cup. Stanza two, the Potṛ’s. Stanza three, the Neṣṭṛ’s. Stanza four counts the draughts he has already taken and adds a fourth vessel. The god being served, Draviṇodas, “the wealth-giver,” is a minor figure whom the tradition barely developed; the interesting party in the hymn is not the drinker but the crockery. Nobody in these verses is named. The cups are, and the cups carry job titles.

From the Hotṛ’s cup he has drunk and rejoiced; from the Potṛ’s he was gladdened; the food set out has pleased him from the Neṣṭṛ’s bowl. Let the wealth-giver drink the fourth cup, undiminished, immortal.

(RV 2.37.4, after Griffith 1896, with the priestly titles left untranslated) [1]

Read that as liturgy and it is a sequence of offerings. Read it as an inventory and it is something stranger: a list of stations, each with its own equipment, each staffed by a man with a title rather than a name. The Rig Veda contains no treatise on how a sacrifice was organised. What it contains is this, procedure fossilised inside praise, and it turns out to be the earliest surviving description anywhere of a team whose members hold distinct, specialised, formally named roles.

The catch, and the reason this article exists, is that the description is unreliable in an interesting way. Count how often each title appears across the whole collection and the numbers are wildly lopsided: one office turns up in more than two hundred verses, another exactly once, and that once is a joke about a bird. The skew is not noise. It is the fingerprint of the people who compiled the text, and following it leads to a fair question about how much of the classical Indian sacrificial machine was already standing when these hymns were composed, and how much of it we have quietly read back into them.

One verse that lists the whole payroll

The single densest piece of evidence sits early in Maṇḍala 2, in a hymn to Agni that spends its second stanza telling the fire god that he is, personally, every member of staff.

távāgne hótraṃ táva potráṃ ṛtvíyaṃ táva neṣṭráṃ tvám agníd ṛtāyatáḥ | táva praśāstráṃ tvám adhvarīyási brahmā́ cāsi gṛ́hapatiś ca no dáme

“Yours, Agni, is the Hotṛ’s office, yours the Potṛ’s in its due season, yours the Neṣṭṛ’s; you are the fire-kindler of the man who keeps the rite. Yours is the Praśāstṛ’s office, you act as Adhvaryu, you are the Brahmán and the master of the house in our home.”

RV 2.1.2, my rendering from the Sanskrit; compare Griffith (1896). [1]

Seven titles, plus the householder, in twenty-odd words. Standard reference works treat this stanza as the classic Rigvedic enumeration of the sacrificial staff, and pair several of its titles with their later successors: the praśāstṛ, the director, becomes the Maitrāvaruṇa of the developed ritual, and the brahmán of this list corresponds to what the later manuals call the Brāhmaṇācchaṃsin. [2] The Maitrāvaruṇa is the one office to have received a book-length study of its own, Christopher Minkowski’s, which traces the director’s function from these Rigvedic traces to the elaborate prompting duties of the classical rite. [6]

Two details make the verse more than a list. First, it is not unique. The identical stanza reappears, word for word, near the end of the collection at RV 10.91.10, a fact that belongs to the large class of Rigvedic repetitions Maurice Bloomfield catalogued in 1916, when he reckoned that something like a fifth of the Rig Veda’s verses recur somewhere in it. [3] A formula that useful got reused. Second, and more revealing, the verse exists in order to flatter. The poet is not documenting a workflow; he is telling Agni that the entire division of labour collapses into him. Our earliest organisational chart survives because someone wanted to pay a compliment, and the most efficient compliment available was to name every job in the building and assign all of them to one god.

Once you notice the move it turns up everywhere. At RV 1.94.6 Agni is “Adhvaryu and ancient Hotṛ, Praśāstṛ, Potṛ, purohita by birth, knowing every priestly function.” At RV 7.16.5 he is householder, hotṛ and potṛ in a single breath. At RV 4.9.3 he is led around the altar as hotṛ, “or he sits down as Potṛ.” The joke, if that is the word, is that the theology of Agni as universal priest is the reason we can reconstruct any of the offices at all.

Methods note. These counts are mine, made over the romanised Saṃhitā text of the Rigveda as published on sacred-texts.com and checked form by form. They count verses that contain the title, not occurrences, and I have removed look-alikes by hand: iṣṭāpūrta at RV 10.14.8 is not the Neṣṭṛ, and agnidhāna, “fire-place,” at RV 10.165.3 is not the Agnīdh. Word counts on an oral text with heavy sandhi are approximate by nature; treat the orders of magnitude as solid and the exact integers as within a verse or two. The brahmán figure is the softest, because the word also means the sacred formulation itself, and separating the man from the utterance is a judgement call in about a third of its occurrences. Translations here are Griffith’s or my own; for a modern English rendering of any of these hymns the standard is Jamison and Brereton (2014). [4]

The frequency table is a self-portrait

Look at that last column again. The hotṛ outnumbers the adhvaryu four to one and the agnīdh by something like seventy to one. If the Rig Veda were a neutral record of who did what at the fire, that would be evidence of a wildly top-heavy operation.

It is not a neutral record. The Rig Veda is the hotṛ’s book. The verses collected in it are, overwhelmingly, the material the hotṛ recited, and the men who compiled, ordered and transmitted it were the inheritors of that office. A hotṛ’s anthology mentions hotṛs. The Rigvedic silence about other roles is a silence about other people’s paperwork, not about their existence, and one office demonstrates this beautifully by its near-total absence.

The classical sacrifice cannot happen without the udgātṛ, the chanter who sings the melodies of the Sāmaveda. His title occurs in the Rig Veda exactly once. It occurs in a simile. It occurs, of all places, in a hymn to a partridge:

udgā́teva śakune sā́ma gāyasi brahmaputrá iva sávaneṣu śaṃsasi

“Like a chanter, bird, you sing the Sāman; like a brahmin’s son at the pressings, you recite.”

RV 2.43.2, my rendering; compare Griffith (1896). [1]

A poet trying to describe a francolin’s dawn call reaches for the most familiar sound in his professional world and comes up with the chanter and the reciting student. The title is absent from the hymns not because chanting was unknown, but because the chanters’ repertoire went into a different collection, the Sāmaveda, which draws most of its verses from the Rig Veda and sets them to melody, and which was transmitted by its own schools down its own line (see The Oral Engine). (The hymn’s other, weirder half, the business of birds as omens, is a story of its own; see Bird Omens in the Rigveda.)

The same pattern catches the upavaktṛ, the “speaker-up,” another director-type office. It appears three times, and twice as a comparison: Savitṛ stretches out his golden arms “like an upavaktṛ” at RV 6.71.5, and soma sends out its voice “like the Hotṛ’s upavaktṛ” at RV 9.95.5. You do not build similes out of things your audience has never seen. These offices were common enough to be figures of speech and marginal enough to go otherwise unmentioned, which is exactly what you would expect of jobs held by people who were not compiling the anthology.

Count it up and ten of the sixteen classical titles are either missing from the Rig Veda or present only as a turn of phrase. Six are genuinely there: hotṛ, adhvaryu, potṛ, neṣṭṛ, agnīdh and brahmán. That ratio is the fact any account of Vedic ritual history has to sit with.

A rota disguised as a round of drinks

Return to the cups, because they are the closest thing in the text to an actual roster. Two passages run the same structure. RV 1.15 works through twelve stanzas of invitation in which the recurring instruction is to drink ṛtúnā, “with the season” or “in due course,” each draught pulled from a named station:

Drink from the Potṛ’s cup, Maruts, in due course; purify the rite, for you are the givers of good gifts.

O Neṣṭṛ, with your lady accept our sacrifice; drink in due course, for you are the giver of treasure.

(RV 1.15.2 to 3, after Griffith 1896) [1]

RV 2.36 does the same thing across six stanzas, cycling the Hotṛ’s cup, the Potṛ’s, the Agnīdhra’s, the Brahmán’s and the Praśāstṛ’s, and RV 2.37 picks up the sequence and finishes it. In the classical ritual this cluster becomes the ṛtugraha, the “seasonal draughts,” a set of soma offerings drawn in a fixed order. Whether the Rigvedic hymns already presuppose that developed rite or whether the later rite was built out of these hymns is one of those chicken-and-egg problems in Vedic studies that has no clean answer, and pretending otherwise would be dishonest. What can be said is narrower and still worth saying: by the time these stanzas were composed, a soma pressing had physical stations, the stations had owners, and the owners were identified by function.

“Adhvaryus!”

There is one hymn in the collection that addresses the workers directly, and it does it eleven times in twelve stanzas.

ádhvaryavo bháratendrāya sómam

“Adhvaryus, bring the soma for Indra! Pour the heady drink out with the ladles.”

RV 2.14.1, my rendering from the Sanskrit.

Every subsequent stanza restarts the same way. Adhvaryus, to him who smashed Vṛtra, bring it. Adhvaryus, to him who broke Śambara’s forts, bring it. The hymn is a work chant with a mythological payload, and the vocative is plural throughout: not “priest,” but “you lot.” Whatever the Rigvedic pressing looked like, it involved a crew of men handling equipment while somebody recited over them, and the recitation was partly addressed to the crew. The physical side of that work, the stones, the boards, the strainer, has its own literature in the hymns; see The Soma Workshop.

Notice what the plural does to the picture. Specialisation in the Rigveda is not one man per title in a tidy grid. It is a title that can be filled by several hands at once, in a rite whose scale probably varied with the wealth of the patron. That fits the general texture of Rigvedic society better than any fixed establishment would: the poet’s fee, the dakṣiṇā, is itself variable, negotiated and boasted about, and eventually gets promoted into a goddess (see The Rigveda’s Abstract Deities).

Seven, or seven-ish

Ten verses in the Rig Veda use the phrase “seven hotṛ-” in one grammatical shape or another. Agni is “decked by the seven hotṛs” at RV 3.10.4; Manu offers the first oblation “with seven hotṛs” at RV 10.63.7; the gods are seated “as the seven hotṛs” at RV 10.35.10. Since RV 2.1.2 lists exactly seven titles, the temptation is obvious: the seven-hotṛ formula names the staff, the staff has seven members, and the Rigvedic sacrifice is a seven-person operation.

Three cautions before accepting that.

First, the phrase is not always grammatically about people. At RV 3.4.5 and RV 10.17.11 the text has sapta hótrāḥ and saptá hotrā́ṇi, from hotrā, a feminine noun meaning the priestly office or the invocation itself, not hotṛ, the man. The two words are separated by one syllable and a world of difference, and even the older translators wobble between “seven priests” and “seven oblations” in these lines. Some of the seven, in other words, may be seven utterances.

Second, seven is the Rig Veda’s favourite number and its least literal one. Seven rivers, seven sisters, seven sages, seven flames, seven metres, seven horses of the sun. A collection that reaches for seven this often is not necessarily counting anything when it does so. Hermann Oldenberg, whose 1894 study of Vedic religion remains the point of departure for this whole area, was already careful about how much institutional detail could be extracted from formulaic numbers. [5]

Third, and most important, the later tradition itself does not preserve one list of seven. The Ṛgvedic Brāhmaṇas specify seven hotrakas who recite the litanies, and their list, hotṛ, brāhmaṇācchaṃsin, maitrāvaruṇa, potṛ, neṣṭṛ, agnīdh and acchāvāka, overlaps with RV 2.1.2 but is not identical to it. [2] Two lists of seven, differing in membership, is what evolution looks like, not what a fixed constitution looks like.

What can be concluded is modest and, I think, more interesting than the tidy version. By the late Rigvedic period there were at least seven recognised sacrificial functions with stable names, some of them clearly more central than others, arranged with enough regularity that a poet could rattle them off in a single line and expect his audience to follow. Whether seven men held them, or five men held seven of them, or the number was a poetic convenience that later ritualists took literally and then had to legislate for, the text does not say.

Aside. It is worth resisting the reflex that treats every Rigvedic institution as an early draft of a classical one. The hymns were not written to be the first chapter of the Śrauta Sūtras. They are the working literature of a particular office, composed over several centuries, in a society whose rituals almost certainly varied by family, region and budget. Reading them as a constitution produces a false sharpness that no amount of footnoting can rescue.

How seven became sixteen

Somewhere between the Rig Veda and the classical manuals, the sacrifice was reorganised, and the reorganisation was drastic. The mature soma ritual employs sixteen officiants in four teams of four. Each team is headed by one of the four chief priests, and each chief is tied to one of the four Vedas: the Hotṛ to the Ṛgveda, the Adhvaryu to the Yajurveda, the Udgātṛ to the Sāmaveda, the Brahmán to the Atharvaveda. The fourfold canon and the fourfold priesthood are, on this arrangement, the same object seen from two directions (for the canon itself, see The Map of Hindu Scripture). Below each chief sit three assistants, ranked and paid by rank.

graph TD
    RV["Ṛgveda"] --> H["Hotṛ + 3 assistants"]
    YV["Yajurveda"] --> A["Adhvaryu + 3 assistants"]
    SV["Sāmaveda"] --> U["Udgātṛ + 3 assistants"]
    AV["Atharvaveda"] --> B["Brahmán + 3 assistants"]
    H --> Y["Sixteen officiants, one soma rite"]
    A --> Y
    U --> Y
    B --> Y

The column headings are not decorative. The classical fee scheme divides the dakṣiṇā by rank into four classes, whose names are simply the fractions: the chiefs, then the ardhin or half-sharer, the tṛtīyin or third-sharer, the pādin or quarter-sharer. [2] Kullūka Bhaṭṭa, commenting on Manu’s law code, works the arithmetic out for a fee of a hundred cows: twelve to each chief, six to each half-sharer, four to each third-sharer, three to each quarter-sharer.

$$4 \times (12 + 6 + 4 + 3) = 100$$

The numbers close exactly, which is the point. A hundred-cow fee divides without remainder across sixteen ranked officiants. Somebody designed that.

Two things stand out when you set this grid against the Rigvedic evidence. The Potṛ and the Neṣṭṛ, both attested in the oldest layers, end up in the bottom rows of the classical scheme, demoted to quarter-sharer and third-sharer. And the Udgātṛ, whom the Rig Veda mentions once, in passing, as a simile, is now a chief priest with a Veda of his own and a team of three under him. Ritual reorganisation, like the corporate kind, moves people around.

The historical question is when and why. Michael Witzel’s reconstruction places the decisive systematisation in the Kuru realm of the early first millennium BCE, where the collection of the Vedic texts into distinct branches and the standardisation of the great rites happened together, as a single political and liturgical project; on this account the four-Veda division and the four-team priesthood are two faces of the same reform. [7] J. C. Heesterman argued for a deeper transformation, from an older, riskier, competitive sacrificial world into the classical system’s closed and rule-bound perfection, in which the dangerous parts were engineered out. [8] The two readings are not incompatible. One describes the mechanism, the other the loss.

The priest who says nothing

The strangest office in the finished system is the one with the least to do. The Brahmán, chief of the fourth team, does not recite the invitations, handle the vessels or sing. He watches. He is expected to know all the Vedas, to remain silent through the performance, to notice when anything goes wrong, and to repair the damage afterwards with expiatory formulas. Breaking his silence at the wrong moment is itself understood to injure the rite; the Aitareya Brāhmaṇa is where the repairs are set out in detail. [12]

Put in plainer terms, the classical sacrifice budgeted a quarter of its staff and a chief’s share of its fee for quality assurance. Every step was witnessed by someone whose only job was to catch errors and whose authority came from knowing the whole system rather than any part of it. Frits Staal’s much-argued claim that Vedic ritual is finally about rules rather than meanings, that it is syntax without semantics, looks less provocative once you notice that the ritualists themselves staffed an error-correction post and wrote a procedure for recovering from faults. [10]

The Rig Veda has the word brahmán, and has it often, but the silent supervisor is not yet visible in it. What the hymns show instead is a formulator, a man of powerful speech, sitting among the others: “the Brahmán speaks the knowledge of what exists,” says RV 10.71.11, in a stanza that walks past four differently occupied specialists in as many lines.

One sits nursing the growth of the verses; one sings a chant in the śakvarī metres; the Brahmán declares the knowledge of what exists; and one lays out the measure of the sacrifice.

(RV 10.71.11, after Griffith 1896) [1]

Four men, four skills, one operation. Nobody here is doing everything. The verse belongs to a hymn about speech itself, which is also where the Rigvedic poets come closest to writing about their own craft as a trade with tools (see The Poet as Carpenter).

The positions were never vacated

In April 1975, in the village of Panjal in Kerala, seventeen Nambudiri priests performed the twelve-day Agnicayana, the piling of the fire altar, with a bird-shaped altar of more than a thousand bricks. Frits Staal and Robert Gardner recorded it; Gardner’s film Altar of Fire came out the following year, and Staal’s two-volume Agni (1983) remains the most detailed documentation of a Vedic solemn rite ever assembled. [9] [13] The performance was widely described at the time as possibly the last. It was not: the same rite was performed again at Panjal in 2011, and Nambudiri and Andhra families have continued to hold śrauta rituals in the decades between. David Knipe’s fieldwork among Andhra ritualists, published as Vedic Voices (2015), follows households where the offices discussed in this article are held by living people with telephone numbers. [11]

Aside. Continuity claims about Vedic ritual invite two opposite errors. One is to treat a modern Somayāga as a window onto the Rigvedic period, which it is not: what is performed today is the classical system, the sixteen-priest version that postdates the hymns by centuries and has its own long history of manuals, revivals and adaptations. The other is to dismiss the continuity as staged antiquarianism, which underrates a transmission that has kept an oral corpus and its performance protocols legible across three millennia. The accurate statement is narrow and still remarkable: the job titles in the Rig Veda are, in an unbroken line of instruction, the job titles being filled today.

What the fire actually required

The Rig Veda is usually approached for its gods, its cosmology, its astonishing poetry about dawn and water and the mind. Its organisational content is easy to miss because it never announces itself; it arrives sideways, in a compliment to Agni, in an instruction about whose cup to drink from, in a plural vocative shouted at a work crew.

But that content is there, and it is early, and its shape is unusual. A society that had not yet begun writing had a technical vocabulary for who does what, distinct enough that the words could be used as figures of speech, stable enough to be listed in a single line and reused verbatim eight books later, and durable enough that the same titles were still being filled in a Kerala village in 2011. The Rigvedic poets’ most familiar contribution to Indian thought is ṛta, the idea that the cosmos runs on a fitted order that can be got right or got wrong (see Varuṇa, the Watchful God). It is worth noticing that the same people applied that instinct to their own working day: they divided the labour, named the parts, assigned the equipment, and eventually hired a man to sit in silence and check the whole thing for errors.

The frequency table is where I would send a sceptical reader first. Two hundred and nineteen verses for the hotṛ and one for the udgātṛ is not a fact about ancient religion. It is a fact about who keeps records, and it applies equally well to a Bronze Age fire ritual and to the last project you worked on. Every archive is written by somebody, and the somebody’s own job is always the one that looks indispensable.

Read RV 2.37 straight through, then read RV 2.14. The first tells you where everyone stood. The second tells you what it sounded like when they were told to get on with it.