Ask what the avatāra is for and the answer comes back instantly, because everyone has heard it. He comes when dharma declines. He comes to destroy the wicked and protect the good.
So you read the Tenth Skandha expecting a list of bad men, and the volume does supply one. Kaṃsa. Jarāsandha. Śiśupāla. Naraka. Pauṇḍraka. Śālva.
Then at 10(50), in the middle of a campaign report, the text says what the burden actually is, and it is not a list of men:
"Act so as to remove the earth's burden, in the form of these twenty-three armies."1
⚑ Not the kings. The armies. The burden is a number of soldiers, and it is stated as a quantity that has to come down.
Once you have seen that sentence, the rest of the campaign reads differently, because the campaign has always behaved like a quantity problem and now you know why.
Jarāsandha arrives at Mathurā with twenty-three akṣauhiṇīs. The number is repeated. It is the number Jarāsandha himself uses later when he gives his own account of the war: "In encounters with Shouri, with twenty-three akshouhinis, I have been defeated seventeen times."2
Nobody in this exchange is counting battles won or territory taken. Both sides are counting units.
And the thing about a quantity target is that it is indifferent to who supplies the quantity. A wicked king who brings no army is not the target. A decent man who brings four akṣauhiṇīs is. The moral frame and the operational frame are not the same frame, and the volume states the operational one out loud and then never mentions it again.
Here is where the stated aim stops being an abstraction.
Jarāsandha is defeated seventeen times. Seventeen times he is captured or routed, and seventeen times he is let go. The text gives the reasoning without embarrassment:
"Since he will make efforts to raise another army again, the king of Magadha should not be killed."1
⚑ The enemy is preserved because he is productive. He is the one man in the region who can raise twenty-three akṣauhiṇīs out of nothing, repeatedly, and walk them into range. Kill him and the recruiting stops. Let him go and the burden keeps arriving on its own legs.
Read it as a farmer would. You do not eat the seed stock. Jarāsandha is not the crop; he is the field.
And what makes this hard rather than merely clever is that seventeen armies' worth of men die in the harvesting. Each one of those seventeen releases is a decision that the men who will be raised next year should be raised, and then killed.
The volume then supplies the war's origin, and it is domestic.
Asti and Prāpti are Jarāsandha's daughters. They were married to Kaṃsa. Kaṃsa is dead. So they "went to their father's house" and told him what had happened.1
That is it. The longest war in the Skandha opens with two widowed women reporting to their father, and the father's obligation to them is the engine that raises twenty-three armies seventeen times over.
⚑ The volume is consistent about this. Its late antagonists are almost all discharging a debt to someone already dead, and they say so — Dvivida for Naraka, Sudakṣiṇa for his father, Śālva for Śiśupāla, Dantavakra in his own words, Viduratha for Dantavakra. Jarāsandha is the earliest and the largest. The volume's violence is kinship accounting, not appetite — and that finding belongs to the antagonist material and is cited here rather than rebuilt.
What it does to this page is sharper. If the burden is armies, and the armies exist because a father is discharging an obligation to his daughters, then the thing being removed from the earth is produced by the ordinary operation of family duty. Nobody in the causal chain is doing anything the text elsewhere condemns.
Then a second front opens — Kālayavana, with an army of mlecchas, from another direction entirely — and Kṛṣṇa's response is not tactical. It is administrative.
He abandons Mathurā. He has a city built in the sea, twelve yojanas across, laid out "in conformity with principles of vastu", with treasuries, storehouses, upper-floor shrines and a population "from the four varnas." And then he moves the people there by yoga, in their sleep, all at once.3
⚑ A god facing two armies at once relocates a civilian population and builds them housing. No battle is fought over Mathurā. The city is simply subtracted from the board.
That is the same mind that keeps Jarāsandha alive. Both moves are indifferent to honour and entirely attentive to the sum. Keep the recruiter; move the non-combatants; take the armies as they arrive.
Your objective is stated as a number. You notice, six months in, that you have stopped asking whether the individual items are the right items, because the number does not care and neither, now, do you.
You could end a recurring problem permanently by removing the person who generates it. You decline, because the problem is how your operation gets fed. You tell yourself this is patience.
Somebody explains why the campaign has to continue, and their reason is a family obligation to someone who died years ago. Notice that this reason is never treated as insufficient, by them or by you.
The pressure comes from two directions at once and you cannot hold both. So you stop defending the position and move everyone out overnight — and afterwards nobody calls it a defeat, because there was no battle to lose.
The definition at 10(50) has one consequence the volume follows all the way through, thirty-one chapters later, and it is the reason this page matters beyond the Jarāsandha campaign.
At 11(1) the Yadus themselves are marked for destruction, and the stated reason is the same reason: they are an army. Kṛṣṇa's own people, having won, are now the surviving military mass, and "the world will be destroyed by their surging flow" if they remain.4
⚑ The war aim is depopulation of the military class on both sides, and the Yadus are the last army to go.
That is what converts a throwaway campaign clause into the volume's structural spine. If the burden had been wicked kings, the Yadus would be safe — they are the protagonist's own lineage, they fought on the right side, and nothing in their conduct at 10(50) marks them. They are not destroyed for wickedness. They are destroyed for mass.
And the volume knows this is the reading, because it stages the Yadu destruction as a drunken brawl among friends rather than as a punishment. Nobody is judged. The number simply comes down.
Three things are absent here, and their absence is the finding.
There is no reconciliation with the moral frame. The volume never says the armies are the burden because armies are made of wicked men. It could have. It does not. The two accounts of what the avatāra is for sit in the same Skandha without being brought into contact.
There is no accounting of the soldiers. Twenty-three akṣauhiṇīs, seventeen times, is a figure the text supplies and never converts into people. The freed kings at 10(73) come closest, confessing that "we were extremely cruel and killed our own subjects" — but that is the kings' own account of what their wars did, not the volume's account of what its war did.
And there is no discomfort. The release policy is stated as sound practice, in a clause, once. ⚠ Nothing in the chapter treats it as a hard case, and any reading that finds anguish in it is supplying the anguish.
That third absence is worth holding against what the volume does find worth an argument. It stops the story to answer a chronology objection about whether a deed done at five was reported at six. It stops it again, twice, so Parīkṣit can raise the morality of the rāsa. It interrupts its own narration at 10(77) to deny that its protagonist could have grieved. The text has an appetite for auditing itself and a listener who exists to object — and neither of them raises a word about the seventeen releases.
⚑ So the silence here is not the silence of a text that never questions anything. It is the silence of a text that questions constantly and does not consider this a question. What a book declines to interrogate is a stronger statement of its priorities than what it defends, and the burden clause is never defended because it is never attacked.
The strongest support is that the definition is operational rather than rhetorical. It appears inside an instruction — act so as to remove — with the object specified as a quantity and a form. It is the kind of sentence a commander says to a subordinate, and the volume prints it as one.
A second support. The release policy has a stated cause clause (since he will make efforts to raise another army again) that only makes sense under the quantity reading. Under the wicked-kings reading, sparing the wicked king seventeen times is unintelligible.
The tension, preserved. The volume holds two incompatible accounts of what its protagonist came to do — the standard one about dharma and the wicked, and this one about mass — and never sets them against each other. ⚑ Per the vault's standing rule, neither is deleted. The page states that the text supplies both and offers no bridge.
A second tension. Jarāsandha, the man kept alive as a supply chain, is also the volume's most articulate speaker on equanimity: seventeen defeats and "I never rejoice or grieve… everything in this world is driven by time and destiny." The doctrine goes to the enemy and the logistics go to the hero, and the text does not remark on it.
Open question. If the burden is mass rather than wickedness, what makes any given death just? The volume's answer everywhere else is absorption — the killed enter him. Whether that answer is meant to cover soldiers, or only named antagonists, is not settled by the range read.
Set against He Argued Karma to Abolish a Rival, the same operator is visible at both ends of the Skandha. There he dismantles a rite by arguing that no deity confers fruits, and then redirects the entire offering to a mountain he is standing inside. Here he dismantles a war by redefining its object. ⚑ Both moves work by changing what the thing under discussion actually is, and in both the redefinition is stated once and never revisited.
Against The Throne He Would Not Take, the political picture completes itself. He refuses the crown at Mathurā on an inherited curse and installs Ugrasena; he keeps a hostile king alive as a recruiting engine; he relocates a civilian population overnight; he later washes guests' feet at somebody else's sacrifice. The volume's protagonist holds no office and runs the whole theatre, and this chapter is where the running is most visible.
Against The Galleries Convicted the Assembly, there is a real friction to keep open. The women in Kaṃsa's arena convict a room for permitting an unfair contest, and the volume prints their rule: one must never remain in a place where adharma is on the rise. ⚑ Nobody applies that rule to a policy of releasing a warlord seventeen times. The text has the vocabulary for the objection and never turns it in this direction.
The Enemy as Untapped Resource — Talleyrand, needing one partner to move against Napoleon, picks the man who tried to have him killed. The enemy is not an obstacle to be removed but an asset nobody else is using, and the relationship's hostility is precisely what makes it valuable.
Jarāsandha is that asset held for seventeen cycles. What he supplies is not intelligence or leverage — he supplies the material the campaign exists to consume, and he supplies it reliably, at scale, on his own initiative, at his own expense.
⚑ What neither field gives alone: the strategic version keeps the enemy alive because he is useful to you, and this one keeps him alive because he is useful against himself. Talleyrand's arrangement is a transaction with a living counterparty who knows what he is getting. Jarāsandha does not know he is being farmed. He believes he is prosecuting a war of obligation for his widowed daughters, and every army he raises out of that grief is delivered on schedule to the thing that will destroy it.
That is a harder mechanism than the strategic literature usually names, because it needs no cooperation and no deception. Nothing is said to Jarāsandha. He is simply not killed, and the not-killing does the work. The most efficient use of an enemy in this volume requires only that you decline to finish him.
Logistics and Extended Campaign: Sustaining Indefinite War — Hannibal, 1,500 miles from Carthage with no possible resupply, adopts a doctrine in which the army lives off the territory it occupies. The campaign's sustainability is solved by relocating the supply problem onto the enemy's ground.
Kṛṣṇa's version runs one level further out. He does not live off the enemy's territory; he lives off the enemy's capacity to regenerate. The thing being consumed is armies, and the producer of armies is deliberately maintained.
⚑ The insight the pairing produces: predatory logistics has a terminal condition and renewable logistics does not. Hannibal's doctrine degrades — every year the countryside is stripped further, allies defect, the base thins. A doctrine that farms the enemy's recruitment instead of his granaries has no such ceiling, which is exactly why the volume needs an external stopping rule and supplies one: twenty-three armies, a stated number, after which it ends.
And when the number is reached, the doctrine turns on its own side. The Yadus are the last harvest. Predatory logistics ends when the land is empty; this one ends when the ledger balances, and the ledger does not distinguish whose soldiers they were.
Sharpest implication. The volume states its war aim as a quantity of soldiers, not a list of wicked men — and everything else follows: a warlord released seventeen times because he is the only reliable producer of armies, a war originating in a father's duty to two widowed daughters, a city evacuated rather than defended, and finally the protagonist's own people destroyed for being the last army left. Nobody on that list is destroyed for being wicked.
Generative questions.