Laṅkā has fallen. Hanuman comes to Sītā in the grove where she has been held, and makes an offer.
He proposes to punish the rākṣasī women who guarded and tormented her — to trample them, drag them, tear them apart.1
It is a jarring moment. The volume spends twenty essays establishing a figure whose defining act is restraint, who spares Rāvaṇa, who stuns rather than kills, whose speed-epithets are each welded to a governor. And here he volunteers atrocity against a group of women whose crime was following orders in a fallen kingdom.
Most contributors pass over it. Svāmī Akhaṇḍānand Sarasvatī does not, and what he does with it is the most interesting interpretive move in the Character & Virtue group.
Akhaṇḍānand names the difficulty rather than absorbing it.
"It's a performance oddly out of character for him, the essay notes, and reads it as deliberate."1
The offer was staged. Its purpose was not to obtain permission — Hanuman had no intention of doing it — but to create an opening in which Sītā's own answer could be heard.
And the essay is specific about whose account this improves: the staging was to make Sītā's greatness visible, not his own.1
⚑ That is a claim about a character deliberately playing worse than he is, so that somebody else can be seen playing better. He supplies a foil, at his own expense, and the cost is that the offer stands on the record as something he said.
The reading has a real textual warrant. The volume elsewhere insists this figure does not act without weighing, and an unweighed proposal to tear women apart, arriving from him at the moment of victory, is anomalous in a way the essay is entitled to notice. Akhaṇḍānand is not smoothing an awkward passage. He is treating an anomaly as evidence — which is the correct instinct, whether or not the conclusion is right.
The reply the staging was for is worth setting out, because it is doing three separate things and the essay only credits it with one.
She invokes universal fault. Nobody is without wrongdoing; the guards' conduct is placed inside a general condition rather than treated as an exception requiring punishment.
She names her own. She cites her own wrongdoing against Lakṣmaṇa — the moment she drove him from her protection with an accusation, which is what left her unguarded.1 ⚑ She does not argue for mercy from a position of innocence. She argues it from having done the same kind of thing, which is the only version of the argument that cannot be answered with you have never been tested.
And she grounds the conclusion in station rather than sentiment: nobility requires compassion. Not that compassion is admirable, but that it is entailed by what one is — which makes the refusal an obligation rather than a generosity.
Three moves, and the second is the one that makes it an argument rather than a sentiment. The essay reports it and does not remark on it, which is a pattern this build has met repeatedly in the volume: the contributor's summary is consistently less precise than the material he is summarising.
Akhaṇḍānand then extracts a decision rule, and it is the single most usable sentence in the group.
"Does this action increase Rāma's own glory, or that of his devotees? If a decision fails that test, one's own devotion must be judged incomplete."1
Three properties, and each is a repair of a failure this build has documented in the modern extension corpus.
It examines the action. Not the actor's orientation, not his motives, not whether he feels devoted — does this action increase. An act-test can be applied to a proposal before it is executed, and by somebody other than the person proposing it.
It names a beneficiary, and the beneficiary is plural. Rāma's own glory, or that of his devotees. The second clause is doing real work: it extends the test past the master to the community of people affected, which is precisely the third-party term this build found missing from every instrument in the modern corpus.
And it states what failure means. One's own devotion must be judged incomplete. Not that the act is wrong, not that one should feel bad — a specific inference about the actor's own standing, drawn from a specific failure. A test with no stated consequence is a suggestion.
⚑ Run it on the staged offer and it works. Would trampling the rākṣasīs increase Rāma's glory, or his devotees'? Plainly not — a victory concluded with reprisals against captive women diminishes the victor and everyone associated with him. The test returns no, decisively, at the first pass, which is what a working instrument does with an easy case.
The modern strategy corpus this build has been reading alongside the volume has a filter it calls the Rāma Test:
Does this serve my audience's dharma or just my ego/profit?2
Set the two side by side and the 1975 version is better on every axis, including the one the modern version is named for.
The modern test is an aim-test. Serve and ego describe the operator's orientation, and nobody sorts their own work into the second bin — the or-clause decides the answer before the question is asked.
The modern test has no failure consequence. It returns a verdict and says nothing about what a bad verdict implies.
And my audience's dharma asks the operator to rule on obligations that are not his to determine. The volume's version asks about glory — a public, observable quantity, assessable by anybody watching — rather than about the interior moral condition of thousands of strangers.
⚑ The corpus named its filter after Rāma and did not consult the volume in which a Rāma-referenced filter already exists, better constructed, doing the same job. That is the build's standing finding in its sharpest single instance: the extension corpus's intuitions are sound and its execution is downstream of not having looked.
The repair costs nothing. Does this action increase the standing of what I serve, or of the people who serve it with me — and if not, what does that say about me? is the same sentence, portable to any domain, with the aim-test removed and a consequence attached.
The essay pairs the punishment scene with a second refusal, and the pairing is what makes the section more than an anecdote.
Hanuman offers to carry Sītā to Rāma on his back. She refuses.1
Her ground is that she has never knowingly touched another man. Rāvaṇa's abduction was forced and therefore does not count against her; willing touch — "even from someone she calls her own son" — would. And the essay reads the refusal as protecting something structural: the rescue must come by Rāma's own strength, not by her secret escape. Akhaṇḍānand calls it "the entire purpose of the Rāmāyaṇa distilled into a single scene."
⚑ What she is protecting is the form of the rescue, and she pays for it in more captivity. The offer would have ended her imprisonment immediately. She declines, and remains where she is, so that the deliverance happens the way it must rather than the way it could.
That is the same shape as the volume's other great non-action — Hanuman declining to rescue Sugrīva from Kumbhakarṇa, because being rescued would diminish him.3 A worse immediate state accepted deliberately, to protect something about how the outcome arrives.
Two instances, two different characters, two contributors who do not cite each other. [n=2] — not yet a pattern, and worth watching for a third, because if the volume has a general doctrine that the manner of a deliverance is part of the deliverance, it is nowhere stated as one and it would organise a great deal of this material.
The staged reading should not be accepted just because it is elegant, and the alternative deserves stating.
The offer may simply have been genuine.
The volume is not squeamish about this figure's capacity for rage. It prints seven registers of one devotion including wrath, an unsoftened rampage, and a battle-cry whose recorded consequences include miscarriages indoors.4 A tradition willing to print those has no need to explain away a violent proposal.
⚠ And the staged reading has a suspicious property: it protects Hanuman from having had the impulse at all. This build applies exactly that scepticism elsewhere — to the corpus that reads its own inconsistencies as design — and consistency requires applying it here.
There is also a cost to the reading that Akhaṇḍānand does not count. If the offer was theatre, then a datum disappears: the volume loses its clearest instance of this figure proposing something the tradition would not sanction, and with it the evidence that his judgement operates under pressure rather than automatically. The staged reading buys Sītā a stage and sells Hanuman's fallibility to pay for it.
🔴 Both readings are recorded here and neither is adopted. What is not in doubt is the extracted test, which stands regardless of why the offer was made — and which is the part worth carrying out of the essay in either case.
Rewrite your decision filters as act-tests with a named beneficiary and a stated consequence. Does this action increase the standing of what I serve, or of those who serve it with me? survives translation into any domain. The three components are separable and each is doing work: the action rather than the aim, so somebody else can apply it; the plural beneficiary, so the community is in scope; the consequence, so a failed test costs something.
When a source explains away an anomaly, ask what the explanation costs the record. A reading that resolves an awkward passage by making it deliberate removes evidence, and the evidence removed is usually the subject's capacity to be wrong. Keep both readings and note which one the source had an interest in.
And notice when someone stages a foil for you. The move Akhaṇḍānand describes — deliberately proposing the worse option so that the better one can be stated by somebody else — is a real technique with an obvious modern use in meetings and negotiations. It is generous when the credit goes to the other party and manipulative when it does not, and the two are indistinguishable in the moment. The test is who is quoted afterwards.
One Aṅka essay for the episode, the staged reading, the extracted test and Sītā's second refusal,1 and three of this build's pages for the comparisons.234
Bounded, and this bound is exact. The staged reading rests on one sentence in one essay, whose wording — "a performance oddly out of character for him… and reads it as deliberate" — was confirmed by a targeted re-read after two general readings of the same essay disagreed about which episode it described. 🚩 SINGLE SOURCE. No other contributor in the twenty-one essays of this group treats the offer at all.
Bounded again. The claim that the volume's version of the test is better constructed than the modern corpus's is an argument about the two formulations as written, not a claim that the volume applies its test anywhere. It does not. The rule is extracted from one scene and never used again in the essay.
🔴 Two readings preserved. The offer as staged theatre against the offer as a genuine impulse the volume had no reason to conceal. This page adopts neither.
Open question. Does Vālmīki or the Mānas give Hanuman a stated intention at this moment, or only the offer? The staged reading is an inference from character, and the primary texts would show whether there is anything in the narration to support it. A single lookup at the Yuddha Kāṇḍa scene would settle whether Akhaṇḍānand is reading the text or reading the figure.
Akhaṇḍānand's essay is built on a single question — what did the work require — and this scene is where the method reaches something it cannot answer that way.
Everywhere else he explains a quality by a requirement: the vanara form removes the grounds of comparison, celibacy enables the assignments, ugliness is adequacy, the curse is a gift.1 All of it functional, all of it transferable.
The punishment offer has no functional explanation. So he supplies an intentional one instead — the act was performed, for an audience, to produce an effect in somebody else — and the switch is invisible in the essay because the register does not change.
⚑ That is worth noticing as a general habit rather than a fault. A writer with one explanatory instrument will use it past its range, and the place where the instrument bends is usually the place where the material is most interesting. Here the bend is where the volume's best decision-rule comes from, and it arrived because a functional reading failed and something else had to be reached for.
The Rāma Test and the Decision Filters. That page takes apart a modern filter — does this serve my audience's dharma or just my ego/profit — and finds an aim-test whose or-clause nobody fails, with no consequence attached and a claim to know obligations the operator cannot know. This page supplies the same tradition's own version and it is better on every count the other page identified as a defect. What the pairing produces is not a correction but a method for finding corrections: the modern construct is named for a figure whose own volume contains a superior construct doing the identical job, and the superiority is entirely in the grammar — act rather than aim, plural beneficiary rather than singular audience, stated consequence rather than none. The lesson generalises past both corpora: when a framework borrows a name, the source of the name is the first place to look for the thing the framework was trying to build, and it is the last place anybody checks.
Indirect Agency and Deniable Power. That page treats an effect arranged so that it cannot be traced to its author — deniability used to suppress a demonstration rather than to escape a consequence. The staged offer is the same mechanism aimed at producing somebody else's authority rather than concealing one's own. Hanuman proposes the worse course so that Sītā's refusal becomes the recorded statement, and the credit accrues to her. The pairing isolates what the influence literature does not separate: the technique of engineering an outcome you do not appear to have caused is identical whether the beneficiary is yourself or another party, and no examination of the method distinguishes them. Only the destination of the credit does — which means this is a case where the ethical question cannot be answered by describing the act, and has to be answered by asking, afterwards, who ended up quoted.
The reading's real problem is that it makes Sītā's argument into Hanuman's achievement.
If the offer was staged to let her wisdom be heard, then the wisdom is still hers and the hearing is his — he built the occasion, chose the moment, and selected the provocation that would elicit the best version of her. On that account the scene's author, inside the story, is the servant.
And the essay says so, in the phrase that is meant to be generous: the staging was to make Sītā's greatness visible. Visibility is something conferred.
There is a plainer reading available in which she simply refused a cruel proposal, on her own reasoning, without needing anybody to arrange the opportunity — and in which the reason her answer is remembered is that it was good.
The volume has form here. This build has already recorded its Sītā material being absorbed into a category before anybody examined how her speech was constructed, with her most skilled piece of advocacy filed under purity by every contributor who touched it.5 A reading that makes her finest moment into a stage another character built for her is the same move, arriving from the opposite direction and looking like praise.