Cry From the South of Our Houses: The Rigveda's Bird Oracles

A Pigeon on the Hearth

Something has gone wrong in the house. A pigeon has come in, probably down the smoke hole, and has settled on the household fire itself, on the wooden fire-drill or the ash beside it. The family does not chase it out with a broom. Someone begins to recite.

“May that which the owl shrieks be in vain, (and may it be in vain) that the pigeon takes his place upon the fire; may this reverence be paid to Yama, (the god of) Death, as whose messenger he is sent.”

Rigveda 10.165, verse 4. Trans. H. H. Wilson (1850-88), with Sāyaṇa’s commentary. [1]

That is RV 10.165, five verses long, and it assumes without argument that a bird arriving uninvited is a communication. Not a nuisance, not a coincidence: a courier, dispatched by Nirṛti, the goddess of dissolution, or by Yama, who rules the dead. The hymn’s response has two moves, and they are worth separating. First it salutes the sender. Then it throws the messenger out.

What makes this hymn strange is not what it says. Half the ancient world would have nodded along. What makes it strange is that it is almost alone. The Rigveda runs to 1,028 hymns and roughly ten and a half thousand verses, and across that enormous corpus there is essentially no divination. No inspection of sacrificial livers. No casting of lots to settle a question. No dream manual, no list of portents, no professional whose job is to read the world for signs. Composed within a few centuries of Bronze Age societies that were obsessed with exactly these techniques, the Rigveda simply declines to practice them.

Except for the birds. Three hymns, six or seven per thousand of the collection, treat a bird’s cry as information. And the Sanskrit word used in them for “bird,” śakuna, will go on to become the ordinary Sanskrit word for “omen,” so completely that a thousand years later an encyclopaedist filing chapters on portents will label the whole section śākuna. The crack in the wall turned out to be the door.

The Techniques the Poets Never Learned

Start with the negative space, because it is unusually clean.

Take the divinatory toolkit available to anyone in western or central Asia in the second millennium BCE. Mesopotamian scholars had extispicy, the reading of sheep livers, supported by clay liver models and omen compendia running to thousands of entries. They had dream books. They had celestial omen series. Anatolia had all of that plus a distinctively local system of lot-drawing. Every one of these techniques is a technology: it needs specialists, training, a transmitted body of precedent, and usually a written archive.

The Rigveda has none of them. Its ritual specialists are graded and titled with great precision, the hotṛ, udgātṛ, adhvaryu, brahmán, and not one of those offices involves reading signs. The standard surveys of Vedic religion, Oldenberg’s and Keith’s, catalogue the priesthoods, the deities and the ritual technology at length, and neither has an institution of divination to describe. [2][3] When the Rigveda wants to know something about the future it does not look for a sign. It asks, out loud, in verse. The mechanism is petition, not inference.

The dream row deserves a footnote of its own, because the Rigveda’s response to a nightmare is not to decode it but to hand it to somebody else, a procedure treated at length in Bad Dreams in the Rigveda. Nor should the dice of the gambler’s hymn be mistaken for an oracle. Those vibhīdaka nuts decide who is ruined; nobody asks them a question. And the Rigveda’s stars, as I have argued elsewhere in Counting Stars, function as a calendar rather than a horoscope. Judicial astrology arrives in India much later, and from outside.

Against that background, three hymns about birds start to look less like a curiosity and more like a clue.

The Bird That Is Asked to Speak

Maṇḍala 2 closes with two short hymns whose subject is a bird called the kapiñjala. The traditional identification, which Sāyaṇa follows and Wilson repeats, is the francolin, the grey partridge whose call rings out across dry scrub at dawn and dusk with an insistence that is hard to ignore. Some later Indian tradition instead identifies it with the cātaka or Jacobin cuckoo, whose arrival announces the monsoon. Macdonell and Keith’s Vedic Index records the uncertainty rather than resolving it. [4] Either way, the poet is listening to a bird that calls at a fixed hour and can be expected to answer.

kanikradaj janúṣaṃ prabruvāṇa iyarti vācam ariteva nā́vam | sumaṅgálaś ca śakune bhavāsi mā́ tvā kā́ cid abhibhā́ víśvyā vidat ||

“Crying repeatedly, and foretelling what will come to pass, (the kapiñjala) gives (due) direction to its voice, as a helmsman (guides) a boat; be ominous, bird, of good fortune, and may no calamity whatever befall you from any quarter.”

RV 2.42, verse 1. Sanskrit after Aufrecht’s edition; translation Wilson. [1][5]

Read that image again, because it is doing real work. Iyarti vā́cam aritéva nā́vam: the bird drives its voice the way a rower drives a boat. Sound is not spilling out of the bird at random; it is being steered. The poet has decided that the call has a direction and a destination, which is precisely the assumption any system of augury needs before it can start.

The word Wilson renders “foretelling what will come to pass” is janúṣam prabruvāṇaḥ, and here the translators part company. Janus- means birth, origin, descent. Griffith, taking the plain sense, has the bird announcing its race or lineage. [6] Wilson, following Sāyaṇa, takes it as announcing what is about to be born, that is, the future. The Sanskrit will bear both. Anyone who wants to weigh the options should set these older versions beside the standard modern translation, Jamison and Brereton’s. [7] The difference is not trivial: on one reading the bird is identifying itself, on the other it is prophesying. That both readings survive tells you the hymn sits exactly on the boundary where description of a bird becomes consultation of a bird.

The second verse turns protective, and it is the most likeable thing in either hymn.

“May no kite, no eagle, kill you; may no archer armed with arrows, reach you; crying repeatedly, in the region of the [Fathers], be ominous of good fortune; proclaimer of good luck, speak to us on this occasion.”

RV 2.42, verse 2. Trans. Wilson. [1]

The poet is protecting his oracle. Hawks, eagles and hunters are asked to leave this particular partridge alone, so that it can keep calling. Note that this is not how a sign works in most omen systems, where the diviner is a passive observer of a message the gods have already sent. Here the relationship runs both ways: the bird is blessed by the man whose future it announces.

Then comes the technical detail, the one line in the Rigveda that reads like a rubric from an augural manual.

áva kranda dakṣiṇató gṛhā́ṇāṃ sumaṅgálo bhadravādī́ śakunte |

“Bird, who are ominous of good fortune, the proclaimer of good luck, cry from the south of our dwellings; may no thief, no evil-doer prevail against us; that blessed with excellent descendants we may worthily praise you at this sacrifice.”

RV 2.42, verse 3. Trans. Wilson. [1]

Position matters. Not merely that the bird calls, but where it calls from: dakṣiṇatáḥ gṛhā́ṇām, to the south of the houses. Verse 2 has already called that quarter pitryā́ pradíś, the direction of the Fathers, and the south is the direction of the dead throughout Vedic and later Hindu practice, the quarter of Yama, the side toward which the funeral rites orient themselves. Sāyaṇa states the rule flatly: a bird’s cry from the south is a good omen. Which is a small puzzle in itself. The auspicious call comes from the inauspicious quarter, perhaps because a message about the future is most credible when it arrives from where the ancestors are.

The companion hymn, RV 2.43, pushes the comparison somewhere unexpected. It hears the bird as a colleague.

“You sing, Bird, like the udgātṛ chanting the sāman; you murmur like the brahmaputra at sacrifices; like a horse (neighing) when approaching a mare do you proclaim (aloud) to us good fortune from every quarter; proclaim aloud prosperity to us from every direction.”

RV 2.43, verse 2. Trans. Wilson. [1]

The bird is an udgātṛ, the priest who sings the melodic portions of the liturgy; it is also the brāhmaṇācchaṃsin, the priest who recites what is spoken rather than sung. The first verse of the same hymn has the bird alternating two kinds of utterance the way a chanter alternates gāyatrī and triṣṭubh metres. In a culture where poetic technique was itself the sacred technology, the highest compliment available was to call the partridge a professional, a point that runs alongside the craft language the poets used of themselves. And the third verse asks that when the bird flies calling, the sound be like a karkarī, a stringed instrument.

Aside. Six verses is not a divination system. There is no rule here for what a call from the north means, no scale of outcomes, no procedure for a second consultation when the first is unclear. What we have is the attitude that makes such a system possible, held at the stage before anyone has written the rules down. It is genuinely rare to catch a practice this early.

The Uninvited Guest

The pigeon hymn is a different animal, and a darker one. In RV 10.165 the bird has not been consulted. It has walked in.

“O gods, let us worship for that, desiring which the pigeon sent as Nirṛti’s messenger, has come to this (ceremony); let us make atonement, may prosperity be given to our bipeds and quadrupeds.”

RV 10.165, verse 1. Trans. Wilson. [1]

Nirṛti is dissolution, the goddess of undoing and decay, and her messenger has crossed the threshold. Verse 3 places the bird in the worst possible spot, on the touchwood at the fire, at the exact point where the household’s relationship with the gods is maintained. Verse 4 pairs it with the owl, whose shriek is to be rendered ineffective, and refers the whole event past Nirṛti to Yama himself. Then the hymn ends by getting rid of it.

“(Praised) by our hymn, O gods, drive out the pigeon, who deserves to be driven out, exhilarated (by our oblation), bring us food and cattle, dissipating all our misfortunes; abandoning our food, may the swift (pigeon) fly away.”

RV 10.165, verse 5. Trans. Wilson. [1]

Two things about this are worth holding onto. The first is the two-step structure: honour the sender, expel the messenger. The hymn never tries to work out what the pigeon means. It has no interest in interpretation at all. It treats the bird as a physical carrier of a bad outcome and responds by physically removing it, which is exactly the logic the Rigveda applies to nightmares, and exactly the logic it does not apply to the francolin.

The second is that this hymn had a long career. It reappears at the head of a run of Atharvavedic charms, AV 6.27 to 6.29, whose opening line repeats the Rigvedic verse almost word for word, and which Bloomfield translates under the plain heading of a charm against pigeons regarded as ominous birds. [8] By the time of the Atharvaveda the pigeon and the owl are a settled pair of ill-omened birds with a standard ritual response, and the ancillary literature will attach procedures to it. The seed in the Rigveda is small; the plant that grows from it is not.

What the Tradition Did With Them

Now look at where these hymns sit, because the transmission itself is evidence.

RV 2.42 and 2.43 are the last two hymns of Maṇḍala 2. Oldenberg’s classic demonstration of how the family books were assembled showed that each runs Agni first, then Indra, then other deities in order of decreasing hymn count, with metre and verse number governing the sequence inside each block. [9] Material that violates the scheme, and material parked at the end, is on his criteria the strongest candidate for late addition. A pair of hymns to a partridge, at the very end of a family book, fails the scheme comprehensively. The dove hymn is in the last stretch of Maṇḍala 10, the layer everyone agrees is latest.

Then there is what the indexes did with them. The Anukramaṇī, the traditional catalogue of poet, deity and metre for every hymn, assigns 2.42 and 2.43 to a deity described as the kapiñjala in the form of Indra. The bird is admitted, then immediately identified with the most orthodox god available. The dove hymn gets a similar treatment: a hymn about a pigeon, an owl, Nirṛti and Yama is catalogued under the Viśvedevas, the All-Gods, the category that absorbs whatever does not fit. [6]

That is what discomfort looks like in a transmission system that is not permitted to delete anything. As the oral machinery of Vedic transmission makes clear, the text itself was preserved with a fidelity that borders on the fanatical. What could be adjusted was the frame around it: where a hymn sat, and what it was said to be about.

graph TD
    A[Bird treated as speaker] --> B[RV 2.42, 2.43 francolin]
    A --> C[RV 10.165 pigeon and owl]
    B --> D[Placed at end of Mandala 2]
    C --> E[Placed in late Mandala 10]
    D --> F[Anukramani: deity is Indra]
    E --> G[Anukramani: deity is Visvedevas]
    C --> H[AV 6.27-6.29 charm]
    H --> I[Grhya and ancillary ritual]
    B --> J[Sakuna omen literature]
    I --> J
    J --> K[Brhat Samhita ch. 95-96]
    J --> L[Banned as low art by Buddhists]

Methods note. Placement arguments cut only so deep. Late in the collection does not automatically mean late in composition; a hymn can be old material that the redactors did not know how to file. What the placement establishes is that the compilers treated these hymns as anomalous, which is a fact about the tradition’s self-understanding rather than about the date of the verses. Both readings are compatible with the evidence, and I would not want to choose between them on the strength of position alone.

Anatolia, Umbria, and the Word for Bird

Step outside India and the Rigvedic hymns stop looking eccentric and start looking like a very thin slice of something enormous.

The Hittites of central Anatolia, whose archives overlap in date with the earlier Rigvedic material, ran the most elaborately documented bird-observation system of the ancient world. Alfonso Archi’s study of the corpus laid out the procedure: augurs watched wild birds within a defined field of vision for a set period, recording species, direction of flight, and behaviour from the observer’s fixed standpoint, then reported the results to the palace in formal protocols. [10] Around thirty bird species are named in these texts. There were dedicated professional titles for the observer and for the ritual practitioner, and the observations were written up as consultation documents on specific royal questions: where the king should winter, whether a campaign should proceed. [11] This is bureaucratic divination, with paperwork.

Italy preserves the other end of the same practice, in ritual rather than report. The Iguvine Tables, seven bronze tablets from Umbrian Iguvium, are the fullest surviving liturgical text of any ancient European religion outside Greece and Rome, and Tablet VI opens with an augural procedure: the officiant marks out the observation field, silence is imposed, and birds are watched by species and by quarter, with woodpecker and crow expected in one direction and woodpecker and magpie in another. [12][13] Rome then made the whole thing constitutional. Roman magistrates could not act without the auspicia, and the augural college maintained a body of law about the taking and interpretation of signs so intricate that its modern reconstruction runs to a monograph. [14] Cicero, himself an augur, wrote a sceptical dialogue about the entire business while continuing to hold the office. [15]

Here the vocabulary becomes interesting. Latin auspicium is transparently avis plus specere, “bird-watching,” and Latin avis is the direct cognate of Vedic ví- (nominative véḥ), “bird,” both from Proto-Indo-European h₂éwis. Greek ornis means “bird” and also “omen.” Sanskrit śakuna means “bird” and also “omen.”

The last column is the important one. What varies across these traditions is not whether people watched birds, which they all did, but what kind of institution happened to write it down and what kind of object happened to survive. Dale Serjeantson’s survey of bird divination in the ancient world, written from the archaeological side, is a useful corrective here: bird-watching leaves essentially no physical trace, so the map of where augury is attested is really a map of where somebody wrote things down. [16]

Aside. The obvious inference, that Indo-European speakers inherited augury from a common ancestor, needs handling with tongs. The three words for “omen” are not cognate with one another; each branch built its omen-word out of its own word for bird, independently. What is shared is a habit of mind, not a reconstructible institution, and habits of mind can spread by contact or arise separately wherever people live alongside noisy, visible, seasonally reliable animals. Brent Vine’s work on Umbrian avieka- and Italic augural phraseology shows how much careful comparative work is needed before even a formula can be traced. [17] The safe statement is that augury was normal across the ancient world and that the Rigveda’s version is unusually small, not that the Rigvedic poets inherited a rite from the steppe.

There is a coda here that I enjoy more than I should. The first critical edition of the Iguvine Tables, with facsimiles, grammar and commentary, was published in 1849 to 1851 by Theodor Aufrecht and Adolf Kirchhoff. [18] Aufrecht then moved to Oxford, worked on the Rigveda, and produced Die Hymnen des Rigveda, the edition of the complete Sanskrit text in Roman transliteration that scholars still cite. [5] The same philologist who published Europe’s oldest bird-augury ritual also published the standard text of the hymns in which a Vedic poet asks a partridge to call from the south of his house. As far as I can tell he never remarked on it.

Why the Rigvedic Version Is So Small

So why do three hymns carry the whole load?

The first answer is about genre, and it is the strongest one. The Hittite bird oracles survive because a royal chancery archived its correspondence and the tablets were baked in the destruction of the city. The Iguvine Tables survive because a priestly college inscribed its operating manual on bronze. The Rigveda is neither an archive nor a manual. It is a collection of liturgical poetry for the great sacrifices, assembled and maintained by priestly families for use by those families, and it systematically excludes the domestic, the small-scale and the practical. We know from the Atharvaveda that Vedic-speaking households had a rich repertoire of charms for fevers, jealousy, insects, hair loss and ominous birds; almost none of that appears in the Rigveda, not because it did not exist but because it was not the Rigveda’s job.

Two cautions follow, and they pull against each other.

The first is that silence in a selective corpus is weak evidence of absence. If bird-watching was practised by ordinary householders at dawn rather than by hotṛs at the soma pressing, the Rigveda would tell us almost nothing about it either way, and the three surviving hymns would be exactly the leakage you would expect.

The second is that we should not overcorrect. The Rigveda is enormous, and it is not silent about domestic life in general: it has hymns about gambling, about frogs, about a wife’s grievances, about herbs and healers, subjects treated elsewhere on this site in the pieces on Rigvedic medicine and the animals the poets did and did not name. If a developed omen science had existed among these people, some technical vocabulary should have surfaced somewhere in ten thousand verses. It does not. The honest conclusion is that bird-listening was a live folk practice with no professional apparatus around it, and that it had not yet been systematised into anything resembling the Hittite protocols.

There is a third possibility worth naming even though it cannot be proved: that the Rigvedic religious programme actively preferred speech. The relationship the poets describe with their gods is contractual and verbal, built on invocation, hospitality and reciprocal obligation, the world of ṛta and of Varuṇa’s watchfulness. In a system where you can address the god directly and expect an answer in kind, reading entrails is a downgrade. You do not interrogate the liver of a sheep when you are on speaking terms with the sky.

How “Bird” Became “Omen”

Whatever the reason for the Rigveda’s restraint, it did not last.

The Atharvaveda already treats the pigeon and owl as a fixed category. The ancillary ritual literature adds procedures. The Ṛgvidhāna, a text prescribing the magical application of Rigvedic verses to practical ends, shows the hymns being repurposed as instruments. [19] And by the sixth century CE the systematisation is total: Varāhamihira’s Bṛhat Saṃhitā, an encyclopaedia of everything a court astrologer needed, devotes a long consecutive run of chapters to omens, with chapter 95 given entirely to the cries of crows and chapter 96 headed śākuna-uttara, the supplement on bird omens. [20] The rules are exactly what the Rigveda lacks: the crow on the right means one thing in the eastern countries and another elsewhere; nesting in a dead tree in Vaiśākha means abundance; the call kava means one outcome and kha-kha means death on the road.

Two thousand years separate áva kranda dakṣiṇatáḥ, “cry from the south,” from Varāhamihira’s regionally indexed tables. The distance is enormous and the continuity is unmistakable, and the word carries it: śakuna, which in RV 2.42 means simply “bird,” is by then the technical term for the entire science.

Meanwhile the practice acquired enemies. The Buddhist canon lists the reading of signs among the debased arts, the tiracchāna-vijjā, by which a wrong-living ascetic makes a dishonest living, and the list in the Brahmajāla Sutta names bird craft and crow craft explicitly, alongside divination by omens and portents. [21] Brahminical law books likewise prohibit a renouncer from supporting himself by explaining omens. By the time the omen manuals reached their final form, India had produced both the most detailed bird-omen literature in the ancient world and a well-developed argument that consulting it was beneath a serious person.

[!NOTE] Neither the Buddhist prohibition nor the legal ones say the omens are false. They say that living off them is a bad way to live. The epistemological objection, that a partridge does not in fact know the future, is not the objection being made.

What the Partridge Was For

The Rigveda’s three bird hymns are worth more than their length because of what surrounds them. Set them against the Hittite tablets and they look like the same practice caught before it hardened, without the paperwork. Set them against the rest of the Rigveda and they look like the one place where the poets are quiet and listening rather than talking.

That is finally what I find striking about RV 2.42. The Vedic religious posture is overwhelmingly a speaking one: the hymn as offering, the metre as vehicle, the god as an addressee who can be praised into a good mood. In these six verses the direction of traffic reverses. The poet asks for information, protects his informant from hawks, promises praise in return, and compares the bird to the priest who chants the melodies, which in that world is not a compliment given lightly. The relationship is oddly egalitarian. A man and a partridge, working the same shift at dawn.

Read RV 2.42 outdoors, early, somewhere in dry country where francolins still call. It takes about a minute. Then notice that you have started, involuntarily, to pay attention to which direction the sound is coming from. That reflex is three thousand years old and the hymn is simply the first person we know of who wrote it down.