Here is a question people answer without noticing they have answered it.
Would you break a rule for your own benefit? Would you break the same rule for somebody else's?
Almost everybody's answers differ, and almost nobody has examined why. The person who will not fiddle their own expenses will fiddle them for a colleague in trouble. The one who will not ask for a favour will ask for one on a friend's behalf, and find it easy. The rule did not change. What changed is who collects.
That asymmetry is usually treated as a nice thing about people, a bit of ordinary decency, and left there.
Antya 10 has a servant state it as a principle, out loud, with the arithmetic done — and the statement is more precise than the doctrine his master spent a chapter establishing.
Caitanya lies down across the doorway of his room. Govinda, whose daily duty is to massage his legs, cannot get in.1
And stepping over the Lord's body is a grave offence.
"Kindly turn on one side. Let me pass." — "I don't have the strength to move My body."
Again. And again.
"I want to massage Your legs." — "Do it or don't do it. It depends upon your mind."2
Read the second answer. It is not a refusal and not a permission. It is a transfer of the decision, made in a way that gives Govinda no information and no cover: whatever he does next, it was his.
Govinda spreads the Lord's own wrapper over him and steps across. He performs the massage. And then, unable to leave the same way, he stays — misses his meal entirely, and keeps massaging all afternoon.3
The corpus gives his thinking rather than summarising it, which is why this is a page.
"My duty is to serve, even if I have to commit offenses or go to hell. I would not mind committing hundreds and thousands of offenses for the service of the Lord, but I greatly fear committing even a glimpse of an offense for my own self."4
Take that apart, because every clause is doing work.
The offence is conceded. He is not arguing that stepping over is permissible. He accepts that it is an offence and accepts the consequence — even if I have to go to hell.
The quantities are asymmetric to the point of comedy. Hundreds and thousands, against a glimpse. He is not weighing one against the other; he is stating that they are not on the same scale at all.
And the variable is not the act. It is who the act is for. Identical conduct, identical offence, and the sorting is done entirely by direction of benefit.
Which is Ādi-līlā 4's criterion — kāma and prema distinguished by whose senses are served, with no content criterion whatever.5 A cook has just applied the corpus's central doctrine to a doorway, and applied it more sharply than the chapter that stated it.
And there is a fourth clause, easy to take as hyperbole, which is not.
Even if I have to go to hell. Govinda has put his own final destination on the table and spent it, for an afternoon's massage.
In most devotional systems that would be incoherent — the whole point of the practice is where you end up, and no act of service is worth forfeiting it. This system can say it, and the reason is elsewhere in the corpus. Its exemplary devotees decline the five liberations when offered, on the ground that every one of them is a promotion out of the position where service happens.12 If the seat is better than the destination, then hell-with-service beats heaven-without it, and Govinda's sentence is not a flourish but an entailment.
The servant has done the arithmetic the theology implies and nobody else states. Rādhā declines liberation in a chapter about metaphysics; Govinda offers to be damned in a corridor, for a specific job, on a specific afternoon — and it is the same calculation with a price attached.
The abstract version has a known weakness, and this scene closes it.
Ādi 4's criterion is unfalsifiable from outside: whose pleasure is this for? is a question only the actor can answer, which is exactly why the corpus's disciplinary chapters produce the results they do.6 Stated abstractly, the direction test is an invitation to self-justification.
Govinda's version runs it in the costly direction. He does not use the criterion to license something he wanted. He uses it to accept a penalty he could have avoided by doing nothing — the offence, the risk, and the missed meal, none of which arrive if he simply goes away and returns tomorrow.
That is the test being applied against the applier's interest, and it is the only form in which the test can be trusted. A criterion that only ever excuses is not a criterion. Here it accuses, and the man applying it accepts the accusation and proceeds.
The corpus never draws this contrast. It states the doctrine in Ādi 4 and shows the honest use of it four hundred pages later, in the mouth of a servant nobody would think to consult on doctrine.
And then the narrator supplies the fact that reorganises the scene.
"The Lord is very interested in manifesting the exalted qualities of His devotees, and that is why He engineered this incident."7
The doorway was arranged. The immobility was not immobility. It depends upon your mind was not a hand-off of a real decision — it was a prompt, delivered by somebody who knew what Govinda would do and wanted it visible.
Which changes what kind of thing has just been recorded. It is not an incident that revealed a quality; it is a demonstration constructed to reveal one, with the subject unaware.
Three observations follow, and the corpus draws none of them.
Govinda's reasoning is not diminished by the staging. He did not know. The thinking was his, the cost was real, and the fact that somebody arranged the occasion does not touch what he did in it.
But the whole exchange was conducted in bad faith on one side. I don't have the strength to move My body is false, said three times, to a subordinate, in order to produce a situation. The corpus reports this approvingly and does not experience it as a difficulty.
And this is the corpus's own house style. The footwater ejection was staged and disclosed; the inquiry of a student who already knows is declared a formality; the King's admission runs on a fiction of ignorance.8 Four constructed scenes, all disclosed to the reader, none disclosed to the participant.
Worth staying with Govinda rather than with the mechanism, because the corpus does not.
He misses his meal. He works an entire afternoon out of position. He accepts, by his own statement, a burden of offence he expects to pay for. And he does all of it inside a situation that was manufactured, on the basis of information that was withheld from him.
The demonstration cost him and it was not for him. The audience is the reader, and to some extent the household; the beneficiary is the corpus's argument about the quality of its devotees. He paid for an exhibit whose value accrues elsewhere.
None of which makes the scene cynical. A person's qualities cannot be exhibited without an occasion, and occasions have to come from somewhere; a tradition that wants to show what its members are made of has no other instrument.
But the ledger should be stated plainly, because the corpus states only one side of it: the man was used, generously, in a way he would have consented to had he been asked, and he was not asked.
Layer A supplies the doorway, the three refusals, Govinda's verbatim reasoning, the missed meal and the narrator's disclosure of the staging.9
Layer B expands his statement into an instruction about the servant's proper attitude to duty and about the primacy of service over personal spiritual accounting.10
Worth marking the direction, because it inverts the passage's own emphasis. Layer A's Govinda is reasoning — weighing, deciding, accepting a cost with the arithmetic explicit. Layer B's Govinda is exemplifying, held up as a model of surrender rather than as somebody who worked something out.
That is a real loss. The verse gives a man doing moral reasoning of a quality the surrounding theology does not reach; the commentary gives a good servant. The first is transferable and the second is only admirable.
You are about to break a rule. Ask who collects. It is the only version of the question that tracks anything real, and your answer will differ depending on the answer — which means you already hold the criterion and have never stated it.
You are using the direction test to justify something you wanted anyway. Then you are not using it. A criterion that only ever licenses is decoration; the test of whether you hold it is whether it has ever cost you a thing you could have had.
You are engineering a situation to reveal what somebody is made of. The reveal is for you and the audience, and the cost is theirs. That may be entirely justified. Do the ledger anyway, and consider whether they would have consented — because you are about to publish the result either way.
Somebody hands you a decision with no information attached — do it or don't, it's up to you. That is rarely neutral. It is usually a prompt from somebody who already knows what they expect, and the absence of guidance is the instrument rather than an oversight.
The evidence is a short domestic scene with the servant's reasoning quoted in full and the narrator's account of the staging supplied in the same passage.11
The tension the source doesn't close: the corpus's most precise application of its central criterion is spoken by a household servant and is never connected to the doctrine. Ādi 4 states it abstractly and unfalsifiably; Govinda states it in the costly direction; the corpus does not put them together.
A second one. The scene is disclosed as engineered and the engineering involved three false statements to a subordinate. The corpus records this as evidence of the Lord's interest in his devotees' glory and registers no difficulty with the method.
Open: whether Govinda ever learns. The disclosure is made to the reader; nothing in the range read indicates the participant was told, and the corpus's other staged scenes follow the same pattern.
This page belongs directly beside Kāma and Prema: The Direction Test and should be read immediately after it. Ādi 4 states the criterion; Govinda applies it, in a doorway, at his own expense, and the application is more precise than the statement. The corpus's most abstract doctrine and its best worked example are four hundred pages apart and unconnected.
Against The Bengali Who Drank the Footwater it is the same construction with the opposite valence. There a scene is staged to establish a boundary; here one is staged to exhibit a virtue. Both are disclosed to the reader and not to the participant, and the corpus treats the method as unremarkable in both.
And it is the honest counterweight to The Two Formidable Enemies. That rule protects a practitioner by removing them from risk; Govinda's principle accepts risk in order to serve, and the two cannot both be the operative ethic. The corpus supplies both and praises both.
Master–Servant Influence — that page's mechanism is downward transmission: servants mirror their master's disposition toward authority, so a master who is loyal and disciplined produces servants who are, and a disloyal one produces the opposite. Behaviour cascades from the top.
The scene fits the model and then does something the model cannot account for, which is where it earns its place.
The cascade is real here. Govinda's whole framing — service above self-accounting — is the household's ambient ethic, and he did not invent it.
But the refinement is his. Nothing in the master's stated doctrine distinguishes offences by beneficiary in the way Govinda's sentence does. Ādi 4 distinguishes lust from love by direction of benefit; it does not say that offences committed for another are a different order of thing from offences committed for oneself, nor does it supply the ratio — thousands against a glimpse. The servant has produced a corollary the master never stated, and produced it under load, in a doorway, in about the time it takes to decide.
Which gives the finding neither page reaches. A cascade model predicts fidelity and cannot predict elaboration, and elaboration is what a transmitted ethic actually does in the hands of somebody who has to apply it to a situation the source never addressed. The master supplies the axiom; the person at the point of contact supplies the theorem, and the theorem is usually sharper because it had to survive a real case.
The practical consequence is uncomfortable for anybody running an organisation on stated values. The most precise version of your principle will be produced by somebody junior, in a corridor, and you will not hear it — because there is no route by which a cook's corollary travels back up to the chapter where the doctrine is stated.
Sharpest implication. A criterion is only load-bearing if it has cost the person applying it something they could have kept. Ādi 4's direction test, stated abstractly, is an invitation to self-justification; the same test in a doorway, used to accept an offence rather than to excuse one, is the only version anybody should trust — and the corpus produces both without noticing which is which.
Generative questions.