Everybody has watched a group try to reverse one person's decision and fail, and the failure has a recognisable shape.
The first appeal is made by the person closest to the decider, privately, and gets nowhere. Then the group goes together, which feels like it should be more powerful and is in fact less — because a collective appeal turns a judgement into a contest, and now backing down would cost something extra. Then somebody senior and revered tries, and the answer is not an argument but a withdrawal: if you keep on, I will leave.
At that point the appeals stop. Not because anybody is convinced. Because the price of continuing has been moved off the decision and onto the relationship, and nobody is willing to pay it for somebody else.
And then everyone gets on with things, and a year later the person nobody could help is dead.
That is Antya-līlā 2, and it is the most serious material in this corpus.
A note on how this page is written. The corpus's verse text and the 1974–75 purports diverge here more sharply than anywhere else in the book, and the divergence is the point rather than a technicality. It is carried in the body, below, and not in a footnote.
The offence is not what the chapter's title suggests, and getting it wrong makes the rest unintelligible.
Junior Haridāsa was sent — by Bhagavān Ācārya, in Bhagavān Ācārya's name — to beg a measure of rice from Mādhavīdevī.1
Mādhavīdevī is an elderly woman who performs constant austerities. Caitanya himself accepts her as a former associate of Rādhārāṇī. In the corpus's own census of intimate devotees she is the half — the three-and-a-half being Svarūpa Dāmodara, Rāmānanda Rāya, Śikhi Māhiti, and Śikhi Māhiti's sister counted as half a person.2
He fetched the rice. Caitanya ate it and praised it, and then asked who had brought it.
So: an errand, on another man's instruction, in another man's name, to a woman the Lord himself venerates, producing food the Lord himself had just enjoyed. That is the whole of it.
The ruling is one sentence, given to a servant rather than to the man: "From this day forward, do not allow Choṭa Haridāsa to come here."3
No reason is given. Not to Haridāsa, not to the household.
He fasts for three days before anybody can tell him why. And the reason, when it emerges, is not about the errand: "I cannot tolerate seeing the face of a person who has accepted the renounced order of life but who still talks intimately with a woman."4
Two things about that sentence deserve stating.
It is about the Lord's tolerance rather than about the offence. I cannot tolerate seeing the face is a statement of a personal condition, not a finding under a rule — and it is precisely the register in which he elsewhere describes his own vulnerability: "if I even hear the name of a woman, I feel changes in My mind and body."5
And talks intimately is doing enormous work. Nothing in the account describes intimacy. A man asked an elderly saint for rice on somebody else's behalf. The characterisation arrives with the sentence and is never substantiated.
Svarūpa Dāmodara appeals. Then the entire body of devotees. And the answer is: "If you speak this way again, I shall go away, and you will no longer see Me here."6
They cover their ears.
Paramānanda Purī appeals — the most senior figure available — and Caitanya begins packing to leave for Ālālanātha rather than discuss it.
That is not an argument and it was never meant to be. It is the same instrument used to close the King's case and the Gopīnātha Paṭṭanāyaka petition: a threat to withdraw, aimed at the petitioners rather than at the petition, in a community whose entire structure is his presence.7
In a group organised around one person, that instrument is unanswerable by construction. The cost of continuing to advocate is not borne by the advocate — it is borne by everybody, and no individual has standing to spend it.
A full year passes. Junior Haridāsa walks to Prayāga and drowns himself at the confluence.
The strongest evidence that the severity here was chosen rather than required is not an argument. It is another case.
Kālā Kṛṣṇadāsa crossed South India as the Lord's personal attendant, was seduced away by the Bhaṭṭathāris, and was retrieved by Caitanya himself. On arrival at Purī the offence is recounted publicly and the dismissal is absolute: *"Now that I have brought him here, I am asking him to leave. Now he can go wherever he likes, for I am no longer responsible for him."* The man weeps, and Caitanya, "not caring for him, immediately left to take His noon lunch."8
And then the staff void the sentence.
Nityānanda, Jagadānanda, Mukunda and Dāmodara confer. They invent an errand — somebody must carry the news of the Lord's return to Śacī in Bengal — keep Kṛṣṇadāsa in service, and "gave him assurance." The next day they ask permission for a person to be sent.
The answer: "Do whatever you decide."9
Kṛṣṇadāsa goes to Bengal loaded with mahā-prasāda and becomes the man who tells Caitanya's mother her son is home.
Two expulsions, both over offences involving women, one of them far graver on any account. In the first, a pretext is manufactured, the sentence is quietly voided, and the Lord tacitly ratifies it. In the second, every appeal is refused, the appeals are closed with a threat of departure, and a man drowns.
The corpus supplies both and never sets them together.
Told of the drowning, the response is recorded in four words: "He smiled in a pleased mood."10
And then: "If with sensual intentions one looks at women, this is the only process of atonement."
That is not a withheld comment. It is a ratification — the suicide named as the correct remedy for the offence.
And the chapter closes by instructing the reader not to reason about it: *"Please hear the pastimes of Śrī Caitanya Mahāprabhu with faith and confidence. Do not argue, for arguments will produce a contrary result."*11
Which is a striking thing to find in this corpus, because Ādi-līlā 2 says the opposite as an unqualified general instruction: discussion of contested conclusions strengthens the mind and should not be neglected.12 Argument is an exercise everywhere in this book and a danger exactly here.
The instruction should be taken seriously rather than scored off. It is what a tradition says when it knows a passage does not survive analysis and values the passage more than the analysis. That is an honest position and it is not available to a knowledge base.
⚠️ This is where Layer A and Layer B come apart, and the gap is the most consequential in the build.
Layer A supplies no rationale. Kṛṣṇadāsa's verse gives the errand, the sentence, the refused appeals, the year, the drowning, the smile and the instruction not to argue. It states no deterrent purpose, no policy, and no lesson for anybody else. It does not use the corpus's own formula for marking performed severity — "He became angry externally in order to establish the etiquette" — which it had available and used twelve chapters earlier for a man who drank footwater.13
Layer B supplies the rationale the verse withholds. The purport carries seven numbered lessons from Bhaktisiddhānta Sarasvatī. The first is a deterrence argument: the Lord had to expel him because otherwise "pseudo-devotees would have taken advantage of Junior Haridāsa's fault by using it as an excuse to live as devotees and at the same time have illicit sexual connections," which would have "demoralized the cult." The second makes it institutional policy — the episode "set the standard for ācāryas, or the heads of institutions propagating the Caitanya cult."14
Layer A records a smile. Layer B builds an administrative precedent on it.
That is not a matter of emphasis. A text that gives no reason and a commentary that gives a policy reason are making different claims, and only one of them is four hundred years old. A twentieth-century movement in rapid expansion had a use for this episode that the sixteenth-century verse does not display, and the honest thing is to say so and to leave the verse's silence intact rather than filled.
You are appealing a decision and the answer is a threat to withdraw. The appeal is over, and not because you were wrong. Recognise the instrument: the cost has been moved off the decision and onto the relationship, and no individual advocate can spend a collective asset. If it is going to be reversed, it will be by manufactured pretext rather than by argument.
You are watching a severe ruling and wondering whether it was required. Look for the same offence handled differently by the same authority. That comparison, not the reasoning, is what tells you whether severity was chosen — and it is almost always available somewhere in the record.
Somebody supplies a policy reason for an act the record leaves unexplained. Separate them. A silence and a filled silence are different objects, and the filling is nearly always more useful to whoever is doing it than to anybody else.
You are told not to reason about something. That instruction is information. It is usually honest — it marks a passage that its own tradition knows will not survive analysis — and it tells you exactly where to look.
The evidence is a complete chapter read to its end, with the offence, the sentence, four refused appeals, the year, the death, the response and the closing instruction all in Layer A.15
The tension the source doesn't close: the corpus has a formula for marking staged severity and does not use it here. It uses it for a man who drank footwater in the wrong building. Whether the formula is diagnostic or is only ever attached to trivial breaches is the question the whole comparison turns on, and the corpus does not answer it.
A second one. Kālā Kṛṣṇadāsa's dismissal is voided by manufactured pretext and tacitly ratified; this one is defended against every appeal with a threat of departure. Same authority, comparable offences, opposite outcomes, and no passage compares them.
Open: what talks intimately refers to. Nothing in the account supports it, and the characterisation arrives with the sentence rather than with the evidence.
This page and Haridāsa's Chosen Death are a declared pair and the consolidation pass refused to merge them, because the insight lives entirely in the contrast and needs two pages to exist. One man dies at a time of his own choosing with the Lord seated before him, is carried on the Lord's own lap, buried by his hands, and has a stretch of sea consecrated around him. The other drowns alone after a year of unappealable rejection and is met with a smile. The corpus has an implicit ranking of deaths and never states it.
Against The Bengali Who Drank the Footwater this is the pivot of the whole build. That chapter proves the corpus can mark a severity as performed. This one does not.
And it must be read with The Prerogative Doctrine, which is the same question answered in the opposite direction within three chapters. Rāmānanda Rāya oils and bathes two young dancing girls daily and is exempted by name; Junior Haridāsa fetches rice from an elderly saint and is expelled to death. The corpus supplies both rulings and treats them as consistent.
Deterrence Strategies — that page's subject is the fight you win by never having it: making the prospective cost of an action visible enough, and credible enough, that the action is not attempted. Deterrence is a communication problem before it is a punishment problem.
Run Layer B's rationale through that and it fails on its own terms, which is the finding.
Deterrence requires an announced rule. Nothing was announced. No reason was given at the time of the sentence, and the household spent three days unable to tell the condemned man why he had been excluded. A penalty whose ground is not stated cannot alter anybody's expectations about their own conduct.
Deterrence requires consistency. Twelve chapters earlier the same authority dismissed a man for a graver offence of the same kind and allowed the sentence to be voided by an invented errand. Two outcomes with no stated distinguishing rule teach the community that the variable is not the conduct.
And deterrence requires proportionality to be legible, or the penalty reads as arbitrary and produces concealment rather than compliance — which is the standard finding wherever punishment outruns its stated rule.
So the commentary's own justification does not hold up in the commentary's own terms. What the episode actually establishes is not a deterrent but a demonstration of unappealability — that this authority's rulings do not respond to collective pressure, seniority or time.
That is a real effect and a different one, and it is worth naming precisely because it is the effect such cases usually have. Deterrence is what institutions say they are producing; unappealability is what they are usually producing, and the two are distinguishable by a single test — whether anybody could state, in advance, what conduct would trigger the same result.
Sharpest implication. The strongest evidence about a severe ruling is never the reasoning offered for it; it is the same authority handling a comparable case differently. That comparison exists here, twelve chapters earlier, and it shows the sentence being quietly voided by manufactured pretext with tacit approval — which means the severity in this case was chosen, whatever else is true about it.
Generative questions.