Amirapu Natarajan states the classical fourfold policy set in a sentence that removes every trace of statecraft from it:
*"All the clever ideas we use to manage human beings may be classified into four categories. They are assurance, appeasement, discord and punishment."*1
⭐ That is sāma · dāna · bheda · daṇḍa, the standard four-upāya set, rendered as workplace advice. Not the four instruments of royal policy. The clever ideas we use to manage human beings.
⚑ And the claim attached to it is a completeness claim. Four categories, and everything goes in one of them. A taxonomy that says all is making a much stronger promise than a list of tactics — and the book then runs the taxonomy twice, in two scenes, in two entirely different modes.
Natarajan's glosses are plain and it is worth holding them in his words rather than the Sanskrit's, because the English shifts them.1
Assurance — sāma. Conciliation: talking someone into alignment by showing them that their interest and yours already coincide.
Appeasement — dāna. Giving. Natarajan's English is the tell: not gift, not inducement, but appeasement, which imports a note of buying-off that the Sanskrit does not carry.
Discord — bheda. Splitting: separating the target from the people whose support makes them formidable.
Punishment — daṇḍa. Force, in the plainest sense.
⚠ Notice there is no fifth box for do nothing. ⭐ The classical tradition has one — upekṣā, indifference — and whether Natarajan knows that and dropped it, or was working from a four-item source, the book does not say. Its absence matters for the second application, where the option of walking away is never considered.
The first run is on the tradition's most awkward internal problem: Aṅgada, on the southern shore, contemplating not going back.1
Hanumān opens with bheda. The argument is a split, executed in two moves:
First, the followers. The monkeys will not die for you. A claim about the durability of Aṅgada's support under pressure, aimed at the belief that makes rebellion feel possible.
Second, the leadership. Jāmbavān, Nīla, Suṣeṇa and I are Sugrīva's ministers, and will never leave him. ⭐ He names himself in the list. ⚑ The splitter includes his own name among the people the target cannot have — which is what makes this bheda rather than a threat: he is not saying he will oppose Aṅgada, he is saying Aṅgada has miscounted.
Then the pivot, which Natarajan files under daṇḍa and which reads as its opposite: you inherit the kingdom anyway, Sugrīva has no children.
⚠ That is not punishment in any ordinary sense. ⭐ It is the removal of the motive — the thing Aṅgada would be rebelling for, handed to him on a timeline. The classification is Natarajan's and the build records it as loose, because the move is closer to sāma: showing the target that their interest and the existing order already coincide.
⚑ What survives the mislabelling is the sequence. Split first, then supply the future. The split removes the means; the inheritance removes the reason.
The second application is where this becomes a method rather than a list, because Hanumān does not choose a tool. He rules three of them out, one at a time, each on a stated property of the target.1
Assurance fails on a rākṣasa. Conciliation presumes a shared frame of interest, and the premise is that he does not have one. Appeasement fails on the extremely wealthy. ⭐ Dāna is a transfer of value, and its purchasing power is a function of what the target lacks. Rāvaṇa lacks nothing. Discord fails because army and ministers are wholly under his control. Bheda requires a seam, and the assessment is that there is not one.
Which leaves one: *"Only by showing my strength will I be able to control him."*
⭐⭐⭐ The conclusion is not a preference. It is a remainder.
⚑ And that is the whole difference between a taxonomy and a procedure. A list of four options invites you to pick the one you are best at — and everyone has a favourite, which is why most people's influence repertoire is one tool applied to every situation. An elimination argument runs the other way: it asks what is true about this particular target, and lets the answer discard options you might have preferred.
⭐ Each elimination is grounded in a property, and the properties are different in kind — a disposition (rākṣasa nature), a resource state (wealth), and an organisational fact (control of the ministers). ⚠ None of the three is I am strong, which is the property he actually ends up using. The chosen tool is chosen by exclusion, not by inventory of his own capabilities.
Put the two applications beside each other and notice who is on the receiving end.
Aṅgada is a colleague. The crown prince of the allied kingdom, on the same mission, under the same commander, wavering. ⭐ Rāvaṇa is an enemy head of state holding a hostage.
⚑ The same four tools are applied to both, in the same vocabulary, with no change of register and no acknowledgement that anything has changed.
⭐⭐ And the framing sentence guarantees it. All the clever ideas we use to manage human beings. One class, one set, and its scope is human beings rather than adversaries — which is what allows a splitting operation to be run on a wavering ally as naturally as on a hostile court.
⚠ This is the eligibility gate arriving as an absence. Nothing in the passage asks whether the target is somebody you owe anything to. ⚑ Aṅgada is split from his own followers by a man who is, by his own account in the same speech, one of Aṅgada's colleagues — and the manoeuvre is filed under management rather than under betrayal because the outcome was good.
⭐ The build's reading: a taxonomy that names its domain as human beings has already decided the question of who may be operated on, and has decided it in the widest possible terms.
The framing sentence is worth returning to, because it is doing more work than a gloss.
"All the clever ideas we use to manage human beings." ⭐ A doctrine of royal policy — what a king does to a rival kingdom — has been relocated into ordinary interpersonal life, and the book's whole premise justifies the move: it is a manual for building personality, addressed to young readers, drawing its examples from an epic.1
⚑ Two things happen at once in that relocation, and they are not the same thing.
It becomes usable. ⭐ Most people have no occasion to conduct statecraft and every occasion to manage a difficult colleague, and the four boxes are genuinely a better map of that territory than the vocabulary most people bring to it.
⚠ And it loses its restraints. A doctrine constituted for kings comes bundled with a theory of when a king may act at all — provocation, standing, proportion, the interests of the realm. Natarajan carries none of that across. ⭐ What arrives in the reader's hands is four instruments and an elimination procedure, with the eligibility question left behind in the original context.
⚑⚑ This is the build's recurring finding arriving in a new place: the corpus takes the technique and leaves the limit. Here the corpus doing it is not the modern extension one — it is a 2008 devotional publisher, and the limit was dropped in the act of making the thing accessible.
Three problems, and the second is the serious one.
Every elimination rests on an unevidenced premise about the target. ⚠ Rākṣasas cannot be conciliated. The wealthy cannot be bought. Rāvaṇa's ministers are wholly loyal. Three assertions, none argued, and the third is false within days — Vibhīṣaṇa defects, and Mālyavān is thrown out of court for the same counsel.2 ⭐ The seam bheda required was there. The assessment missed it, and the Carit narrates the defection two chapters later without anyone noticing that the earlier elimination was wrong.
The procedure has no null outcome. ⚑ If all four had been eliminated, the argument produces nothing — and there is no fifth box, no upekṣā, no report back and await instruction. ⭐ An elimination argument over a closed set always terminates in an action, which is exactly the property you do not want in a method for deciding whether to use force.
And the conclusion was reached before the scene it justifies. ⚠ Hanumān is a bound envoy in a hostile court when he supposedly runs this reasoning. What he actually does in that court, in the Carit's fuller telling, is neither daṇḍa nor anything on the list: he refuses Rāvaṇa's premise about himself, delivers a compressed Vedānta lesson, and offers terms under which Rāvaṇa keeps Laṅkā.3 ⭐ The elimination concluded only strength; the envoy tried instruction first, at length, twice over. 🔴 The taxonomy and the scene do not agree, and the volume prints both.
Run the eliminations before the selection, and write down the property each one turns on. ⭐ The discipline is not the four boxes; it is being made to say why not three times. A person who cannot state the property that rules out conciliation has not ruled it out — they have skipped it.
Then check the property, especially the one that eliminates splitting. ⚠ This is where the Carit's own reasoning failed, and it is where it will usually fail, because their people are loyal is the assessment most often made from outside and least often tested. ⚑ The seam is rarely visible from the enemy's side of the wall. Vibhīṣaṇa was writing Rāma's name on his walls while the elimination was being run.
Treat appeasement as a function of what the target lacks, not of what you can offer. ⭐ This is the sharpest single line in the passage. A gift's power is measured in the recipient's deficits, which means the tool is unavailable against anyone sufficiently supplied — and it means the standard mistake, offering more of what the target already has, is not a matter of degree but of category.
Name yourself in the split rather than threatening with yourself. ⚑ Jāmbavān, Nīla, Suṣeṇa and I will never leave him is an inventory, not a warning. It gives the target a corrected count instead of an adversary, and it is the reason Aṅgada could climb down without having lost to anybody.
And supply the future after removing the means. ⭐ A split that leaves the target with the same goal and no way to reach it produces a cornered person. The inheritance line is what converts a blocked rebellion into a waiting one — and the Carit's Aṅgada goes on to serve.
Natarajan's book for the taxonomy and both applications,1 the Carit for what actually happened in the court and for the defection that falsifies one elimination,23 and two of this build's pages for the corpus's own escalation material.45
Bounded. ⚠ Natarajan cites no text for the fourfold set. He presents it as a classification of "clever ideas we use to manage human beings", with no attribution to the Arthaśāstra, the Manusmṛti, the Mahābhārata or any other source, and the build does not claim one for him. [PLAUSIBLE — needs corroboration] [LOCATOR MISSING] [EXTRACTION — verify against print]
⚠ A second boundary. The upekṣā observation above is the build's, drawn from the fourfold set's ordinary form, and Natarajan neither includes it nor excludes it. The claim here is only that his list has four items and no null option — not that he suppressed a fifth.
🔴 A contradiction preserved. The elimination concludes that only daṇḍa can work on Rāvaṇa. The Carit's court scene has him attempt sustained instruction, refuse Rāvaṇa's self-description, and offer a settlement that leaves Laṅkā in Rāvaṇa's hands.3 ⚑ Two accounts of the same envoy's approach to the same target, in one build, and neither text knows the other exists.
🔴 A second, internal to the reasoning. Bheda is eliminated on the ground that Rāvaṇa's ministers are wholly under his control. ⭐ Vibhīṣaṇa, Mālyavān and the omen-reporting ministers are all documented dissenters in the very next chapter.2 The elimination was not merely unproven; it was wrong.
Open question. ⭐ Does the volume anywhere state an eligibility condition for using the four? The build has read 123 teaching essays and the whole non-tattva site and has found the tools taught and never gated. If the condition is genuinely absent across both corpora, then a persuasion doctrine has been transmitted through this tradition for at least fifty years with no answer to when may I do this to someone — and that is the gap the book's extension will have to fill itself.
Two practitioner corpora, both building escalation logic, arriving at opposite shapes.
The extension corpus builds a hierarchy — a graded ladder of responses with a stated order of resort.4 ⭐ Its logic is ordinal: try the gentle thing, then the next, then the next. It assumes escalation is a sequence you walk.
Natarajan's is a partition. ⚑ Nothing is tried and abandoned; three options are ruled out in advance on the target's properties and the fourth is used first. On this model escalation is not a path but a diagnosis, and the gentle options were never available at all.
⚠ The two produce different failures. A hierarchy wastes moves against a target immune to the early ones — and telegraphs your position while doing it. A partition acts on the strongest tool immediately, on the strength of assessments that, as this very scene demonstrates, can be wrong.
⭐ What neither has is a re-entry rule. ⚑ When Vibhīṣaṇa defects and the bheda elimination is falsified, no one revisits the conclusion. The partition has no provision for updating, and the hierarchy has no provision for going back down — and the Carit supplies the falsifying event without either corpus registering it.
The Sermon in the Enemy's Court. That page holds the envoy's speech to Rāvaṇa as an argument tailored to one listener under maximum constraint. Set against the elimination, it produces a finding neither passage yields alone: the taxonomy's output was overridden in the field, by the same agent, within hours. ⭐ The reasoning concluded that only a display of strength could work. The speech that was actually delivered opens by denying Rāvaṇa's own account of his species, runs a compressed Vedānta argument, and closes with terms — surrender Sītā, keep Laṅkā. ⚑ That is sāma at full extension, attempted against the target on whom sāma was ruled out. ⚠ And the build should be honest about which one to believe. The elimination is a tidy piece of doctrine reported in a 2008 summary; the speech is the scene. ⭐ What the pairing establishes is that this corpus's operators do not follow its own decision procedures — which is either a defect in the procedures or evidence that the procedures were reconstructed afterwards from an outcome, and the second is more likely.
The Conflict-Resolution Hierarchy. That page records the extension corpus's graded ladder of responses with a stated order of resort. The four tools are the same problem solved by partition instead of by sequence, and the pairing exposes what each shape hides. ⭐ A hierarchy makes the order explicit and the criteria implicit — it tells you what comes next and never why this rung fits this person. A partition makes the criteria explicit and the order irrelevant — Rāvaṇa's wealth is named as the reason dāna fails, and nothing is tried first. ⚑ What the pairing produces is that these two are complements rather than rivals, and each corpus has exactly the half the other lacks. ⚠ A usable version is short: run the partition to eliminate on stated properties, then walk the hierarchy through whatever survives, and treat any falsified property as an instruction to re-run the partition. Neither corpus has the third clause, and it is the one that would have caught Vibhīṣaṇa.
Sharpest implication. ⭐ The elimination procedure is the best decision method either corpus contains, and the volume's own narrative falsifies it twice in the space of two chapters — once when the eliminated tool turns out to have been available, and once when the envoy ignores the conclusion and tries the eliminated approach anyway. ⚑ The interesting question for the book is not whether the method is sound but why a tradition preserves a procedure alongside the scenes that break it, without noticing. The build's answer, on the evidence of ten phases: teaching essays and narrative live in different sections, are written by different people, and are never read against each other — which is the same mechanism that hid the report format and the scarcity speech.
Generative questions.
What would the fifth box contain? ⭐ A set with no null option cannot output this is not yours to act on. Adding upekṣā would change the Rāvaṇa reasoning entirely — an envoy with no mandate for force reaches the end of the eliminations and has somewhere to go. The extension corpus needs this box more than it needs any of the four.
Is appeasement is a function of the target's deficits generalisable to the rest of the set? ⚑ If each tool has a resource it draws on — shared interest for sāma, deficit for dāna, seam for bheda, asymmetry for daṇḍa — then the elimination procedure becomes mechanical, and the four assessments are four questions rather than four judgements. Natarajan supplies the pattern for one of the four and never states it as a pattern.