They have run out of their houses in the middle of the night. Milk left half-drawn. A baby put down mid-feed. Husbands left with food in front of them. Four categories of male relative physically tried to stop them and did not manage it.
And the man they ran to opens with this:
"O slender-waisted ones! This night is terrible in form and is frequented by beings that are terrible. Go back to Vraja. You should not remain here."1
Then a worry about their families. Then a compliment about the moonlight. And then he stops being polite and starts being doctrinal:
"The supreme dharma for a woman is to faithfully serve her husband, ensure the welfare of the relatives and take care of the children, without any deceit. A woman who desires the worlds must not abandon her husband, even if he is evil in conduct, unfortunate, aged, foolish, diseased or poor… For women who are from noble families, illicit intercourse is always condemned, even if it happens to be a minor one. This does not lead to heaven. It brings ill fame and hardship."1
That is a complete, orthodox statement of strīdharma, delivered by God, to women who have just abandoned exactly what it tells them not to abandon.
Read the closing instruction carefully, because it goes further than the argument requires:
"You should therefore not have any proximity to me, through hearing about me, seeing me, meditating on me, loving me and chanting about me. Hence, return to your homes."1
Five modes, named individually. Hearing. Seeing. Meditating. Loving. Chanting.
Those are not the components of an affair. They are the standard limbs of devotional practice — the things this entire book exists to recommend. He has just prohibited, to these women, the whole programme the Purāṇa is written to advocate.
⚑ The lecture does not merely tell them to go home. It tells them not to be devotees. And the volume prints it without comment, in the mouth of its God, as the strongest case that can be made against the thing about to happen.
Compare the list to what the corpus recommends elsewhere and the overlap is not approximate. Śravaṇa — hearing. Smaraṇa — remembering, here as meditating. Kīrtana — chanting. Three of the nine classical limbs, named in sequence, with seeing and loving added. These are not activities that happen to be adjacent to devotion. They are the vocabulary the tradition uses to define it, and Volume 2's Prahlāda material sets out the nine-limb list explicitly.
So the sentence is doing something odd even by the standards of a staged objection. A strīdharma argument needs to prohibit going to a man at night. It does not need to prohibit thinking about him, and it certainly does not need to prohibit singing about him — which no husband, relative or moral code in the volume has ever objected to and which the entire book will spend thirty more chapters urging on everyone.
⚑ The prohibition overshoots its own justification by a wide margin, and the overshoot is the evidence. A text writing a merely plausible objection stops at the adultery. This one keeps going until it has forbidden its own programme — which is what you do when you are trying to lose the argument honestly rather than stage a loss.
The text does not hurry past their reaction. It stays on it for four sentences:
"Their faces were downcast and they grieved. They sighed. Their lips were like bimba fruit and those dried up. With the nails on their toes, they scratched on the ground. The tears smudged and washed away the collyrium and the kunkuma on their breasts. They stood there, silently bearing the heavy burden of misery."1
Toenails scratching at dirt. That is a very specific image of humiliation, and the narrator supplies it rather than moving to the reply.
Then: "He was their beloved, though he had not addressed them in the words of a beloved."
Inside the absolute-sounding rule there is a clause that is not absolute.
A woman must not abandon her husband, even if he is evil in conduct, unfortunate, aged, foolish, diseased or poor — "as long as he is not a great sinner."
A limit. Six named tolerances, and then a line past which the rule stops binding.
It is in the text, it is doing real doctrinal work, and neither party uses it. The gopīs do not argue that their husbands are great sinners. Kṛṣṇa does not invoke it either. The exception sits there unused, which is worth recording because it changes what kind of rule was being stated: not an unconditional prohibition, but a threshold rule, presented as though unconditional.
Here is the move, and it is not defiance.
They begin by granting his authority completely:
"You know about dharma and you have said that the natural dharma of women is to serve their husbands, offspring and well-wishers."1
They concede the premise. All of it. And then:
"O dear one! O lord! This instruction should actually be applied to you. You are loved by all those who possess bodies. Indeed, as the atman, you are the relative."
That is the whole rebuttal.
Serve your husband, relatives and children — agreed. But he has told them elsewhere, and the volume states it flatly at 10(14), that he is the ātman in all beings, which makes him the relative that husbands and sons are approximations of. So serving him is not a breach of strīdharma. It is strīdharma performed against its actual object.
⚑ They do not break his rule. They apply it correctly, and the correct application produces the opposite of what he asked for. Then the closing line, which is the only aggressive thing they say: "What do husbands, sons and the others amount to? They lead to nothing but hardships."
You are told the policy. You do not dispute the policy. You restate it back — accurately, generously, better than the person who quoted it at you — and then you show that the policy, applied properly, requires the thing they were using it to forbid. There is no route back for them that does not involve abandoning the policy.
Someone opens by making your case against you, in full, before you have said anything. It feels like an ambush and it is closer to a gift: whatever they say next, they cannot be accused of not having considered it.
You need to refuse someone and you want the refusal to hold. So you do not soften it. You state the strongest version, you watch it land, and you leave a silence — and if they come back with something that survives all of that, you were going to lose the argument anyway and now you know.
The obvious reading is that the lecture is a test — he never meant it, it is there to make the gopīs prove themselves. That reading is available and the volume leaves room for it.
But look at what the text spends on making it not obviously a test.
The doctrine is orthodox and correct. It is not a caricature of the objection; it is the objection at full strength, including the threshold clause that a careless version would drop. The distress is printed at length and is not undercut. The narrator calls his words "disagreeable" and "cruel" in the gopīs' own mouths, and does not correct them. And the prohibition extends to hearing, seeing, meditating, loving and chanting — a needlessly total ban if the point were only to see whether they would stay.
⚑ A staged objection would not have had to be this good, and would not have had to hurt this much. Whether it was sincere is not decidable from the chapter. What is decidable is that the volume built the case against itself to full strength, printed the damage, and then let the other side answer — and the answer wins on the merits rather than on devotion.
Worth separating, because they are often run together.
Parīkṣit interrupts twice in these five chapters. The first, here at 10(29), asks how women whose minds were full of guṇas and who loved him as a paramour could possibly be liberated. That is a question about mechanism, and Śuka answers it with the Śiśupāla precedent and the five affects.
The second, at the close of 10(33), is much harder and is about conduct: "By associating with the wives of others, how could he then act in this perverse way?"
Different questions, different answers, and only the second produces the non-imitation clause. This page carries the first frame; the second belongs with the clause it generates.
The strongest support is the five-limb prohibition. Nothing in a strīdharma argument requires banning meditation and chanting. That clause is over-built for its stated purpose, and its over-building is what shows the volume knew it was staging its own strongest objection rather than a plausible one.
The tension, preserved. The lecture is orthodox and correct within the volume's own framework, and the volume never retracts it. It is answered, not withdrawn. Later chapters continue to assume conventional household dharma — the debt-to-parents doctrine at 10(45), the householder's day at 10(70), the eleven modes of worship at 11(11). A position defeated in argument at 10(29) is still operating at 11(11), and the text does not treat that as a problem.
A second tension, and it is the cluster's. The gopīs win by redefinition — he is the ātman, therefore serving him is strīdharma. That argument, if sound, is available to anyone. Four chapters later the volume installs a non-imitation clause specifically to prevent anyone using it. The rebuttal that wins here is the one 10(33) exists to fence off.
Open question. The "great sinner" threshold is stated and never used by either side. Does the volume anywhere invoke it, or any comparable limit on a duty presented as absolute? Not found in the range read; the āpad-dharma material at 11(17) is a different mechanism — permission to switch livelihood under distress, not a limit on a marital obligation. No claim is made about Volumes 1–2.
Set against Bhakti Transcends Dharma: the Gopī Defense, this chapter complicates the standard framing in a way worth stating precisely. That defence has bhakti overriding dharma — the higher claim trumping the lower.
The gopīs do not argue that. They never say dharma is outranked. They say the dharma was misaddressed, and that they are the ones keeping it. That is a materially different and much stronger position, because it does not require a hierarchy of claims — it requires only that he is what he says he is.
The transcendence reading is the tradition's, and it is doing work the chapter does not need. What the chapter supplies is not transcendence but correct reference.
Against The Only Thing Anyone Has Ever Loved, the two passages lock together and were written fifteen chapters apart. 10(14) establishes that everything is loved because it belongs to the self, and that Kṛṣṇa is that self. 10(29) is the gopīs cashing that in. Their winning move is unavailable without the earlier chapter's argument, and neither passage cites the other.
Sympathize, Socialize, Select, Show, Secure — Hughes collapses every objection into four types — indifference, disbelief, confusion, obstacles — and supplies one formula that handles any of them without rebuilding.
The rāsa scene runs that machinery with the roles reversed, and the reversal is the useful part. The person raising the objection is the one with all the power, and the people handling it have none. Kṛṣṇa's objection is type four, obstacles — your relatives will search, your children are crying, the night is dangerous — wrapped around a type-two core, this is not permitted.
And the gopīs' answer skips every step of the formula except one. No sympathising, no socialising, no showing. They go straight to reframing the premise, which in Hughes's taxonomy is not a step at all.
⚑ What neither field gives alone: the objection-handling literature assumes the objection is a barrier between the parties, when the strongest handling treats it as a claim the objector has staked and must now honour. The formula manages the objector's state — reduce the friction, make the yes easy. The gopīs do the opposite: they make his own stated rule bind him. He cannot retreat from the supreme dharma for a woman is to serve without abandoning the ground he chose to fight on. That is a fifth move the formula has no slot for, and it is only available when the objector has argued from principle rather than from preference — which is precisely why arguing from principle is the more dangerous way to say no.
Ahimsa as Dharma Bedrock, With Carved Exceptions — Aśoka's one unmodulated principle, held consistently across regions and languages, and the exceptions he nonetheless carved into it: the general problem of an absolute principle meeting a world in which absolute adherence is not possible.
The strīdharma lecture has the same architecture and hides it better. Six tolerances are named — evil, unfortunate, aged, foolish, diseased, poor — which reads as the rule proving its own severity. Then "as long as he is not a great sinner", which is a carved exception doing exactly what Aśoka's do.
⚑ The insight: an enumerated list of tolerances is how an exception gets smuggled past the audience. Reciting six hard cases the rule survives makes the rule sound absolute, and the listener's attention is spent on the severity rather than on the qualifier that follows. Aśoka's exceptions are visible because his principle is stated bare; this one is nearly invisible because it arrives after a demonstration of toughness. And the tell is that neither party in the scene uses it. The gopīs had a clean legal route — argue the threshold — and took the harder philosophical one instead, which suggests the clause was not salient to them either. A carved exception nobody notices is functionally an absolute rule, and this is what that looks like from inside.
Sharpest implication. The volume builds the case against its own central episode at full strength — orthodox, correct, and extending to a total ban on devotion itself — prints what it costs the women to hear it, and then lets them win on his own premise rather than against it. They do not transcend the rule. They apply it to the right object, and the right object is standing in front of them.
Generative questions.