Two servants see a woman on a mountain and go back to report to their king. What they deliver is not a description. It is a schedule of assets, itemised from the thighs upward — nose, arms, anklets, girdle, necklace, hair, earrings, bangles, rings, garments, tilaka — and it closes on a sentence about a collection:
*"A woman as beautiful as her does not exist among asuras, nagas, gandharvas, danavas, or anywhere in the three worlds. It is only fit that someone as worthy as you should enjoy someone like her. O lord! All the women who are jewels have become jewels in your inner quarters."*1
That last clause is the argument. Not she is remarkable, but your collection contains every item of this class, and here is an item of this class.
What follows runs four escalating moves — a report, a proposal, a counter, and threats — and each of the first two is conducted entirely in the vocabulary of acquisition. And the fourth ends with the negotiators dead.
The register of Caṇḍa and Muṇḍa's report is worth reading as a form rather than as content.1
It opens with them incapacitated — "They lost their senses and deluded, fell down on the ground" — and then, having recovered, they enumerate: the attendants and what each is doing (massaging the feet, tending the hair, applying collyrium, holding a mirror, offering betel), then the body part by part, then the ornaments item by item, then the weapons.
⚑ It is the same grammar the volume uses for a donation list or a temple inventory — a sequence of named items, each with an attribute, delivered without a main argument until the last line. And the last line is a completeness claim.
⚑ Note also what is not in it: nothing about who she is, what she wants, or why she is on a mountain with eight armed arms. The report contains no information relevant to the proposal it is recommending. She is described exhaustively and identified not at all.
The messenger Sugrīva delivers Śumbha's actual offer, and it opens with his conquests before it mentions her:
"I have defeated Indra and the others in battles and have taken away their jewels. When gods and others are given shares in sacrifices, I myself enjoy the shares of devas. I think that you are a jewel among women, stationed above all other jewels. Therefore, choose me or my younger brother and enjoy the joy of love."1
⚑ Three claims, in order: I have taken things from gods. I receive what is owed to gods. You are a thing of that class, ranked first.
The proposal's persuasive structure is a portfolio. Her position in it is stated comparatively — above all other jewels — which is the same move as "all the women who are jewels have become jewels in your inner quarters", now applied to her individually. ⚑ And "choose me or my younger brother" is the tell: the identity of the counterparty is a detail. What is being offered is admission to a holding, and either holder will do.
Her reply refuses the frame, and the way she does it is the sharpest thing in the chapter — because she uses an argument that has failed twice elsewhere in this corpus, and it works.
"Mahesha Sadashiva is without a second. He is the supreme brahman… I am his subtle Prakriti. How can I accept anyone else as a husband? Even when she is full of desire, a lioness never chooses a jackal. A female elephant does not choose an ass. Nor does a leopardess choose a hare. O daityas! Your words are futile."1
That is the incommensurability argument — the claim that a union between unlike parties is improper by nature — and this corpus has two prior instances of it, both in women's mouths, both lost:
| Speaker | Where | Figure used | Outcome |
|---|---|---|---|
| Menā | Vol 1, against Śiva as a groom | the disparagement kit — gold/glass, sandalwood/mud, sun/firefly | capitulates |
| Satyavatī | Vol 3 ch 358-6(44), refusing Parāśara | "Like glass and gold, there should not be union between us" | overridden inside the sentence — she is transformed into Yojanagandhā on the spot |
| Devī | Vol 3 ch 361-6(47), refusing Śumbha | the lioness, the elephant, the leopardess | holds |
⚑⚑ The argument is the same. The speaker's position is not. Satyavatī's objection is answered by deleting the ground she stood on — her body is altered mid-refusal. Devī's is not answerable, because there is nothing about her the other party can change.
⛔ [n=3 OBSERVED] across the corpus, and the reading is that the variable is power rather than the argument. That is the sharpest analytical statement of the women's-positions material anywhere in this vault, and it requires all three instances on one page to be visible at all.2
Before the lioness speech, at the first approach, she does something else — and it is easy to misread as a refusal.
*"However, I took a pledge earlier. Hear about it. I will be interested in accepting as a husband whoever can defeat me in battle and whoever can destroy my insolence, not anyone else. This is certain. Go and repeat my words to Shumbha and Nishumbha. Let him then decide whatever is appropriate."*1
⚑ That is not a no. It is a term. She accepts the frame of marriage-by-acquisition and inserts a condition into it — and the condition is one only she can adjudicate, since the test is defeating her.
⚑ And she sends it back through the messenger with the decision explicitly returned to him: "Let him then decide whatever is appropriate." The negotiation is kept open on her side and the next move is his, which is precisely what produces the war.
⛔ Worth stating plainly: the pledge is unattested elsewhere in the volume as far as read. It is introduced here, by her, at the moment it is needed.
The escalation is fast and it is documented in the order it happens.1
Dhūmralocana is sent with instructions that are already coercive — "Go there quickly and bring her here… If she desires to fight, make efforts to fight." He is burned to nothing by a syllable, and the killing becomes her name: "uttering the sound of humkara, Maheshvari burnt him down. Since then, Devi has been known as Dhumavati on earth."
Then Caṇḍa, Muṇḍa and Raktabīja, with the threat stated outright — "Otherwise, we will kill you, your ganas and your mount" — followed immediately by advice on husband-selection, in the same speech.
And then, after her lioness reply, the line that ends the negotiation phase:
"Our minds know that you are a woman. Therefore, we will not kill you. But if your mind still wishes to fight, be ready with your lion."1
⚑ Restraint offered as a courtesy, by three parties who are dead in the next sentence. The clause is doing two jobs at once — it withdraws the threat they made four lines earlier, and it re-states the disqualification that the whole sequence has been running on.
The chapter closes the sequence with a clause that undoes its own moral arithmetic:
*"Though Parameshvari fought with them, she was toying with them and killed Chanda, Munda and Raktabija, the great asura. Their minds were full of enmity towards the gods. Nevertheless, at the end, they obtained a world that only her followers go to."*1
⚑ Enmity stated, and the destination granted anyway — because the death came at her hands. It is the same clause the adjacent chapter applies to Śumbha and Niśumbha: "both of them reached the supreme destination, having come by their deaths through Shivaa's lotus hands."3
⚑ Which places the whole courtship inside the volume's larger structure rather than beside it: the parties who proposition her, threaten her, and are killed for it end where the devotees end. The volume's exemplary lives are overwhelmingly of people its own rules exclude, and this chapter adds four more without appearing to notice.4
You are reading a negotiation and want to know what the initiating party thinks they are negotiating over.
Read the briefing, not the offer. The offer is drafted to be accepted and is therefore polite. The internal report that preceded it is drafted for the principal, and it uses the principal's real vocabulary. ⚑ Here the report is an itemised catalogue closing on a completeness argument about a collection, and everything about the subsequent proposal follows from it.
Check what the briefing omits. This one covers her body, her ornaments, her attendants and her weapons, and contains nothing about her identity, her intentions or her situation. ⚑ An exhaustive description with no identifying content is a strong signal about the category the subject has been placed in — and it predicts, correctly here, that the proposal will treat the counterparty's identity as interchangeable too. "Choose me or my younger brother."
Watch for the term inserted rather than the refusal issued. Her "whoever can defeat me in battle" accepts the frame and installs a condition only she can judge. ⚑ That is a much stronger position than a refusal, and it is the move that keeps the initiative — and it is also what makes the escalation inevitable, because a condition invites an attempt.
And notice when an argument that failed before suddenly works. ⚑ The lioness speech is, structurally, the argument Satyavatī lost with. When the same argument succeeds for one speaker and fails for another, the argument is not what is being tested — and the fastest way to find out what is, is to ask what the losing speaker had that could be altered.
Evidence. Ch 361-6(47) read in full in place, with all four moves quoted directly — the servants' report and its closing sentence, Sugrīva's proposal, the pledge-condition, the Dhūmralocana and Caṇḍa–Muṇḍa–Raktabīja embassies, the lioness speech, the mercy clause and the liberation clause.1 Satyavatī's gold/glass refusal read in place at ch 358-6(44), and the Volume 1 Menā material taken from that build's catalogue.2
Tension 1 — the three-instance comparison spans three volumes and two builds. ⚑ Menā's instances are recorded rather than re-read here, and the figures differ — gold/glass and sun/firefly on one side, lioness and jackal on the other. What is common is the form of the argument, not its images, and the page rests on the form.
Tension 2 — this chapter is in the volume's expansion zone. The Umā Saṃhitā runs forty-six per cent over the shloka budget the text publishes for itself, and the six Devī chapters around this one carry another work's frame and episodes with no note attaching. ⚑ So this courtship may not be this text's composition, which would make the three-instance comparison one that spans a join. [PLAUSIBLE — needs corroboration]; the collation has not been run.5
Tension 3 — "power is the variable" is a reading of three cases. ⛔ It is the best available explanation and it is not the only one: the two losing speakers are also the two who are asserting something about themselves, while Devī is refusing a frame imposed on her. Both readings fit all three cases, and this page states the power reading as the sharper one without excluding the other.
Open question. Devī's "I took a pledge earlier" is introduced at the point of need and is unattested elsewhere in the volume as read. Does the pledge appear anywhere in the corpus — or is it, like several other clauses in this volume, a back-reference pointing at a version never told?
Against Closure Without Refutation: this is the page that supplies that pattern's control case. Its material is women's positions closed without ever being refuted — by renaming, by non-sequitur, by deletion of the stated ground, by purchase, by retroactive reattribution. ⚑ Here the identical argument is deployed and simply holds, and the difference is not in the argument. A pattern with a working control is worth much more than one without, and this is that control.
Against What the Demon Knew: the adjacent chapter, and the same sequence one move later. There Niśumbha reasons correctly, knows he cannot win, and attacks for shame. Here his negotiators make the same choice with less information — and the mercy clause is the moment the calculation is visible, because they withdraw a threat and re-assert a disqualification in one breath.
Against The Corpus of the Disqualified: four more names for that page's file. Caṇḍa, Muṇḍa and Raktabīja are described as full of enmity towards the gods and are given a world only her followers reach, and nothing mediates between the two clauses.
Against Two Kauśikīs: same chapter, and the reason the courtship exists at all. The goddess who conducts it was produced sixteen lines earlier — and the Vāyavīya's account of her production has her commissioned by Brahmā specifically to kill these two. ⚑ On that account the pledge, the negotiation and the escalation are all downstream of a boon the granter could not revoke; on this chapter's own account, none of that context is present.
Value Comparison vs Premium Justification — that page's territory is how a thing's worth is argued: by placing it against comparables, or by asserting a category in which comparison does not apply.
The two halves of this chapter are those two moves, run against each other by opposing parties, and the outcome is unusually legible.
Śumbha's side argues by comparison throughout. The servants' report ranks her — "A woman as beautiful as her does not exist among asuras, nagas, gandharvas, danavas, or anywhere in the three worlds" — and closes with a completeness claim about a set. The proposal ranks her again, "a jewel among women, stationed above all other jewels", inside a portfolio of things taken from gods. Every sentence places her on a scale.
Her reply refuses the scale rather than her position on it. "I am his subtle Prakriti" is a category claim, and the lioness, the elephant and the leopardess are three restatements of a single proposition: the classes do not admit exchange. She does not argue that she is worth more. She argues that the comparison is not defined.
⚑ And the diagnostic that falls out is one the commercial framing makes sharp. A comparison argument can always be answered by improving the comparison — which is exactly what happens to Satyavatī, whose objection is met by altering her body, and to Menā, who is talked round. A category argument can only be answered by changing what the speaker is, and where that is impossible the argument is unanswerable. So the choice between the two is not stylistic; it is a bet about whether the other party can move the terms.
Running back the other way, the chapter supplies something the pricing frame usually cannot show: what happens when the category argument holds. The comparison-side does not concede and does not reprice. It escalates to force — three embassies, a threat, and a war — because a scale it cannot use is indistinguishable, from inside, from a refusal to deal.
Sharpest implication. Two servants see a goddess on a mountain and report back in the grammar of an acquisitions list — attendants, thighs, nose, arms, anklets, ornaments, weapons, itemised — and close on the argument that matters: "It is only fit that someone as worthy as you should enjoy someone like her. O lord! All the women who are jewels have become jewels in your inner quarters." The proposal that follows opens with conquests before it mentions her — "I have defeated Indra and the others in battles and have taken away their jewels… I think that you are a jewel among women, stationed above all other jewels" — and offers a choice of holder: "choose me or my younger brother." Her counter is not a refusal but a term — "whoever can defeat me in battle" — with the decision explicitly handed back. And when the escalation comes, she answers with the incommensurability argument: "Even when she is full of desire, a lioness never chooses a jackal. A female elephant does not choose an ass. Nor does a leopardess choose a hare." ⚑⚑ That is the argument Menā lost with in Volume 1 and Satyavatī lost with three chapters earlier, whose refusal was answered by transforming her body mid-sentence. Here it holds. The argument is identical and the outcome is not, which makes this the corpus's control case: the variable is the speaker's position, not the reasoning. And the negotiators, having said "Our minds know that you are a woman. Therefore, we will not kill you", are killed in the next sentence — and liberated, their enmity stated and the destination granted anyway.
Generative questions.