You have been in the meeting where somebody asks a question they know the answer to.
Sometimes it is a trap, and everybody can feel it. Sometimes it is generous — the senior person asking so the junior one can say the thing and be seen saying it. Sometimes it is procedural: the question has to go on the record, so it gets asked whether or not anyone needed it.
What all three have in common is that the information is not what is moving. The exchange has a purpose, the purpose is real, and it has nothing to do with anybody finding anything out.
Everybody knows this and nobody says it during the meeting, because saying it would end the exchange.
Madhya 20 says it, in the middle of the longest instruction in the book, about that instruction.
Caitanya tells Sanātana that he already knows all of this.
And then explains why he is being asked anyway: "It is the nature of a sādhu to inquire. Although he knows these things, the sādhu inquires for the sake of strictness."1
Take the position that sentence puts the reader in. You are some way into a chapter of instruction. The teacher pauses and tells the student that the student already possesses the content, and that the questions producing the chapter are a matter of form.
The corpus has declared its own longest teaching scene a performance, from inside the scene, and then carried on with it. Nothing about the instruction changes afterwards. The questions keep coming, the answers keep coming, and the disclosure is never revisited.
Worth being precise, because there are two readings and they are not equally interesting.
The weak reading: Sanātana is being flattered. A teacher says you already know this as a courtesy, the way anybody does, and the sentence is politeness.
The strong reading, which is what the text supports: the inquiry is nature, and it is done for the sake of strictness. Neither of those is a compliment. They describe a category of person who asks as a matter of what they are and how things are properly done — which makes the questioning an observance rather than a need.
Under the strong reading, the chapter's question-and-answer form is liturgical rather than informational. The shape is the point. And the corpus is entirely comfortable with this, because it holds elsewhere that the whole scene runs on borrowed capacity anyway: Madhya 20 says outright that Sanātana asks by the Lord's own power.2
So the student's questions are supplied by the teacher, and the student already has the answers, and the exchange proceeds. Nobody in the conversation is doing what the conversation appears to be doing.
For the sake of strictness is the phrase that decides how the page reads, and it is easy to skim.
Strictness is not about truth. It is about doing a thing properly, in the prescribed order, whether or not the shortcut would produce the same result. It is the register of ritual, procedure and craft — the carpenter who marks the line even on a cut he could make by eye.
Which places the sentence somewhere specific in this corpus's own vocabulary. This is the tradition that says intention decides everything and content carries no information; that ranks the interior state above every observable; that licenses outrageous conduct on an inner condition and refuses impeccable conduct on one.3
And here it defends an empty form on the grounds that forms should be kept. The information content of the exchange is nil by the teacher's own statement, and the exchange is justified by propriety.
That is the corpus's disciplinary half speaking inside its doctrinal half, and it is the closest the two come to appearing in one sentence.
Once the exchange is not for the student, the question is who it is for — and the answer is available.
It is for the reader. Sanātana has the content; Caitanya has the content; the only party in the arrangement who does not is the person holding the book. A dialogue between two people who both already know is a device for delivering material to a third who does not, and the third is outside the frame.
Which explains something the chapter would otherwise make inexplicable. If the questions were genuine, their sequence would follow one man's confusion, and confusion is idiosyncratic. The sequence instead follows a syllabus — relationship, then process, then goal — laid out in the same chapter as the organising structure of the whole tradition.4
Genuine questions do not arrive in curriculum order. These do.
So the disclosure is not a lapse in the fiction; it is the fiction being named. And naming it costs the corpus nothing, because the material is delivered either way — while it gains something specific: a reader who has been told the exchange is staged cannot later feel they were being handled.
This belongs to a family, and the family is verified rather than asserted.
Disclaimed agency runs across this corpus at [n≥3]: the author calls himself a wooden doll writing under orders; Rāmānanda calls himself a dancing puppet and a stringed instrument; Caitanya says he does not dance by his own will.5 Here the disclaimer is applied to the inquiry rather than to the answer — the student's questions are not his either.
And the corpus names and endorses the move itself, which is why it is not a tic. It is a doctrine about where capacity comes from, stated by everybody it applies to.
One adjacent reading was tested and failed, and is recorded rather than deleted. The extraction pass proposed that the two great dialogues are a declared matched pair — Rāmānanda answering by the Lord's power, Sanātana asking by it, with the agency deliberately mirrored. Madhya 20 does say exactly that. The test was whether a third instruction scene carried the same explicit borrowed-agency framing, and Madhya 25 does not. Two instances, not a signature, and it is not cited forward by any page in this build.
Layer A gives the disclosure in a line and does not develop it.6
Layer B reads it as a model for the disciple's conduct: one should inquire from the spiritual master even about things one believes one understands, because the asking is itself the discipline.7
That is a fair application and it moves the sentence's subject. The verse describes a category of person — it is the nature of a sādhu — and states a fact about what such a person does. The commentary converts a description into an instruction addressed to the reader.
Worth marking because the two do different work. A description of an advanced practitioner is an observation; the same sentence aimed at a beginner is a rule about deferring to a teacher, and the corpus's own Layer A never issues it.
You are being asked something the asker plainly knows. Work out which of the three it is — trap, generosity, or procedure — because the correct response differs completely and the exchange looks identical in all three.
You are running a teaching dialogue for an audience. The questions will drift into curriculum order and stop being anybody's real confusion. That is not a failure; it is what the form is for. But it does mean the person on stage has stopped representing the audience, and somebody should be checking what the audience would actually have asked.
You have been told outright that a process is a formality. Notice that the telling costs nothing and buys a great deal — a person who has been shown the mechanism cannot later feel they were managed by it. Disclosure is often the cheapest available protection against the charge of manipulation, and it does not slow the mechanism down at all.
You are keeping a form whose content you no longer need. That is what strictness means and it is defensible, but it should be held as observance rather than mistaken for enquiry, because the two produce different behaviour the moment the form and the situation disagree.
The evidence is a single explicit statement inside the Sanātana instruction, applied to the instruction containing it.8
The tension the source doesn't close: a corpus that ranks intention above every observable defends a form here on the ground that forms should be kept. For the sake of strictness is a content-criterion argument in a tradition that says content carries no information.
A second one. The chapter declares its own dialogue staged and continues it unchanged for the rest of its length. Nothing about the instruction is adjusted, which means either the disclosure changes nothing or the instruction was never contingent on the questions — and the corpus does not say which.
Open: whether Sanātana's knowing is a fact or a courtesy. The strong reading makes the whole chapter liturgical; the weak one makes it politeness. Nothing in the range read distinguishes them, and they imply very different accounts of what the Madhya-līlā syllabus is.
This page completes a set with Rāmānanda as Instrument and I Do Not Dance by My Own Will. Those two disclaim the answer and the action; this one disclaims the question. Taken together the corpus has removed agency from every position in its own exchanges — asker, answerer and actor — and still presents the exchanges as instruction.
Against Sambandha, Abhidheya, Prayojana it supplies the evidence for reading that schema as a curriculum rather than as a description. A dialogue whose questions arrive in syllabus order is a syllabus with dialogue on top, and this sentence is the corpus conceding as much.
And it sits beside The Eho Bāhya Ladder as the other great dialogue seen from the opposite end. There the teacher withholds and the student produces; here the student's producing is itself supplied. Both scenes are constructed, and only one of them says so.
Staged Objections as a Rapport Device — that page's mechanism is an author voicing the reader's doubt before the reader finishes forming it. But I'm not qualified for this. The effect is not merely that the objection gets handled; it is a jolt of being known, and the warmth arrives before any argument does.
The parallel is close and the difference is the whole finding.
Both are dialogues in which one party supplies the other party's line. The staged objection is written by the person who will answer it; Sanātana's questions are, on the corpus's own statement, asked by the power of the person answering them. In both, an exchange that appears to have two sources has one.
The staged objection depends on not being noticed as staged. Its effect is the reader's sense of recognition, and recognition cannot survive the discovery that the recogniser wrote both halves. The technique is defeated by disclosure.
The corpus discloses. It tells you the questions are supplied, the student already knows, and the form is being kept for its own sake — and the chapter proceeds without losing anything, because it was never trading on the reader's sense of being read.
Which gives the general finding. A device that runs on the audience's feeling of being known is destroyed by transparency; a device that runs on the audience's need for the content is not. That is a usable test for any constructed dialogue, sales script or teaching set-piece: ask what breaks if the construction is announced. If the answer is everything, the mechanism was the concealment. If the answer is nothing, the form was doing honest work and the concealment was never load-bearing — which is worth knowing before deciding whether to be annoyed at having been staged to.
Sharpest implication. Announcing that an exchange is a formality costs almost nothing and forecloses the charge of manipulation permanently — and the traditions that do it are not the ones with least to hide, but the ones whose mechanism does not depend on the audience feeling met. The test is not sincerity. It is what survives disclosure.
Generative questions.