Somewhere in the middle of a company's twentieth-anniversary book, between the founding story and the photographs, there is a paragraph about the early years that reads slightly wrong.
Some of those who worked with us in that period took a different direction. Their approach, which departed from the founders' vision, did not prosper.
No names. No account of what the different direction was. And the sentence is doing a great deal of work — it explains an absence, forecloses a question, and settles a dispute the reader did not know had happened, in a book nobody would think to check for a dispute.
Ādi 12 has that paragraph, and it is written in tree.
"After the disappearance of Śrī Caitanya Mahāprabhu, some of the branches, for unfortunate reasons, deviated… Some branches did not accept the original trunk… Thus Lord Caitanya did not sprinkle upon them the water of His mercy, and they gradually withered and died."1
Read it as botany and it is gentle. Branches wither; that is what happens to branches; nobody did anything.
Read it as what it is — a passage about people, written after their deaths, by a partisan of the surviving line — and every clause is a ruling. Deviated is a verdict. Did not accept the original trunk is the charge. Did not sprinkle upon them the water of His mercy assigns the outcome to the founder rather than to the writer's own party. And gradually withered and died describes what happened to their claim as though it were weather.
A dispute about legitimate succession is, in plain terms, a fight. Somebody claimed and somebody denied, and one side eventually had the numbers.
The tree makes it a natural process.
That is the whole function, and it is worth spelling out because the device is everywhere and rarely noticed. Botanical metaphors convert human decisions into biology, and biology has no agents to hold responsible. Nobody expelled anybody. There was no meeting, no ruling, no faction. There was a tree, and some branches were not watered, and the not-watering is attributed to the founder — who was dead.
Compare what the same corpus does when it wants an act to be visible. Junior Haridāsa is expelled by a named person, in public, with the refusals of intercession individually recorded.2 The corpus is perfectly capable of narrating exclusion as exclusion. Here it chooses horticulture.
The positive half of the ruling: only Acyutānanda's line is legitimate, and "those who did not follow this path simply scattered."3
Scattered is the second botanical verb doing the same work as the first. Seeds scatter. Nobody scatters them.
And there is a real historical situation underneath this. Advaita Ācārya had several sons; the question of which of them carried his authority was live and contested in the generations after his death, and this chapter is one party's settlement of it, written into a devotional catalogue where it will be recited rather than argued with.
That is the most effective possible place to put a succession ruling — not in a polemic, where it invites a reply, but inside a list of names that the same chapter says produces liberation when recited.4
Then the reach of it:
Anyone opposing the cult is "an atheist subject to be punished by Yamarāja" — and the qualification is explicit: "be one a learned scholar, a great ascetic, a successful householder or a famous sannyāsī."5
That list is not decoration. It names the four categories of person who could plausibly mount a credible rival claim: intellectual authority, ascetic authority, social authority, institutional authority. All four are pre-emptively excluded.
Which is a strong move and a familiar one. A rule that says your credentials do not protect you is aimed precisely at people with credentials, and it is issued at the moment when the founder's personal authority is no longer available and credentials are the only currency left.
The uncomfortable thing about this chapter is that nothing new had to be invented for it.
Three chapters earlier the tree was a generous image. It was the founder admitting he could not reach everywhere, commissioning every man within this universe, and giving distributed members a place to attach.6 Nobody reading Ādi 9 would call it an instrument of exclusion.
But a structure that can tell a member where they attach can also tell them they don't. The trunk that made the distribution legible is the trunk that later defines deviation. The apparatus of inclusion and the apparatus of purge are the same apparatus, and the corpus supplies both uses within three chapters, in one metaphor, without any sense that a second thing has happened.
That is not hypocrisy and it is not a later corruption. It is what organisational structure is: a shape that says where things belong necessarily also says what does not belong, and the second capability is dormant only until someone needs it.
A limit worth stating plainly.
Nothing on this page establishes that the excluded lines were treated unjustly. They may have deviated in ways the corpus would have been right to reject; sixteenth-century Bengal produced plenty of claimants worth rejecting, and the corpus's own later chapters record a case where a man of flawless external conduct was refused because his position was genuinely incompatible.7
What can be said is narrower and firmer. We have one side's account, delivered in a form designed not to be argued with, in a document written by the surviving party's partisan. The other side's version does not exist in this corpus, and this vault holds no page representing it, because the source does not supply one.
Layer A gives the withering, the single legitimate line and the sanction, embedded in a devotional catalogue.
Layer B supplies extensive historical and genealogical detail about these families and lines — considerably more than the verse — and is generally firm on the legitimacy question.8 The commentary is on the same side as the verse, which is unsurprising, and it adds the specificity a modern reader would otherwise lack.
Worth recording that the two layers do not diverge here at all. Where the corpus is settling a succession dispute, the twentieth-century commentary is a continuation of the sixteenth-century argument rather than an observer of it.
You read that a group of people drifted away or did not continue or took a different direction. Ask who wrote the sentence and what would have happened to them if the other side had won. Passive and botanical verbs in an institutional history are almost always covering an agent.
You are writing the history of something you were part of and there were people who left badly. Name it as a decision, including your own part in it. The gardening version is easier to write and it is the version that will make a later reader distrust everything else on the page.
Your organisation adopts a structure that makes it clear where everyone attaches. That is a genuine good and it is worth having. Notice that you have also, in the same act, built the mechanism that will one day be used to say somebody doesn't — and that the mechanism will feel entirely natural when it is used, because it will just be the structure doing its job.
Somebody tells you that credentials, seniority or track record will not protect you from a rule. That rule is aimed at people who have those things, and you have just been told where you stand.
The evidence is a short, explicit passage: the deviation, the failure to accept the original trunk, the withering attributed to withheld mercy, the single legitimate line, and the sanction with its four named categories.9 Nothing has to be inferred.
The tension the source doesn't close: the same tree is a commission and a court. Ādi 9's image justifies universal distribution and Ādi 12's adjudicates legitimacy, and the corpus never notices the second use is a use.
A second one. The commission was every man within this universe. The exclusion applies to people who accepted that commission and then, in the surviving party's judgement, attached to the wrong trunk. A universal commission and a restricted lineage are both canonical, and no rule is given for how a commissioned distributor becomes an illegitimate branch.
Open: the corpus supplies one side. What the excluded lines actually held, and whether did not accept the original trunk is a fair characterisation of it, is not recoverable from this source, and this page makes no claim about it either way.
This page is the far end of The Bhakti-vṛkṣa and the two must be read as one movement. Three chapters separate a founder's generous, capacity-justified commission from a ruling on who is legitimate, and the metaphor does not change in between. That continuity is the finding.
It sits directly against Remembering the Names Liberates, which is the same chapter's other face. The catalogue is a practice that liberates anyone who recites it and asks nothing about who they are — and it is a register of membership with a boundary. One enumeration, one chapter, doing an unconditional thing and a conditional thing at once.
And it supplies the institutional consequence of The Pañca-tattva's structure. A configuration with real authority at five points has no mechanism for settling a disagreement among its heirs; this chapter is what arrived instead of a mechanism.
Otto Strasser Expulsion 1930 — Loyalty Extraction — that page turns on one exchange in a Berlin hotel room. Strasser: "A Leader must serve the Idea… since it is eternal whereas the Leader passes and can make mistakes." Hitler: "For us, the Leader is the Idea, and each party member has to obey only the Leader."
That is the same dispute as did not accept the original trunk, put in the two parties' own words, and having it stated so plainly makes visible what the botanical version conceals.
Strasser's position is the one the withered branches presumably held in some form: there is a thing we are all serving, and it is separable from any particular person or line, and fidelity to it can require departing from them. Hitler's is the trunk position: the person is the thing, so separating them is not a disagreement about doctrine but a defection.
Neither position is unreasonable and only one of them can be institutionally survivable. A movement that agrees the Idea is separable from the Leader has licensed every future member to claim they are serving the Idea better — which is exactly what a succession dispute is. A movement that fuses them has no such vulnerability and no way to correct a leader either.
What the pair yields is a test rather than a verdict: watch what happens to the first person who says the thing is bigger than the lineage. How that is handled tells you which of the two structures you are actually inside, and it tells you before the stakes are high. In Berlin it produced an expulsion narrated as a betrayal. In Bengal, eighty years after the founder's death, it produced a paragraph about branches that were not watered.
The difference is only in the register. Strasser's expulsion is narrated with agents, in a room, with quoted speech — which is why it can be argued about five hundred years from now. Ādi 12's is narrated as weather, inside a list that is meant to be recited rather than examined, and that is the more durable of the two techniques by a considerable margin.
Sharpest implication. The safest place to put a contested ruling is inside something nobody reads critically — a catalogue, a liturgy, a list of names the same text says produces liberation when recited. A ruling stated in a polemic invites a reply. A ruling recited as devotion becomes a fact about the world within a generation.
Generative questions.