Between 1656 and 1659, a rising rebel and the most dangerous prince in India wrote to each other in the language of loyalty and service. Shivaji offered to hold his captured Bijapur lands "on behalf of the Mughals." Aurangzeb wrote back about pardons, conditions, and a robe of honour. Read on the surface, it looks like a junior begging entry into a great court. Read underneath, it is two predators circling — each using formal courtesy to learn what the other actually wants. The strangest thing about the exchange is that we can only hear half of it. Not one of Shivaji's letters survives; we have only Aurangzeb's replies, and through them we reconstruct what the Maratha must have been saying.1 This page is about correspondence as combat by other means, and about what it looks like when both correspondents already know the other is lying.
Start with the gap in the archive. The handful of letters Shivaji sent Aurangzeb in this period are gone — most likely burnt with the rest of his papers when the Mughals invaded his territories after his death.2 What remains is one-sided: Aurangzeb's responses, which act like a mirror held up to letters we cannot see. When Aurangzeb writes about "requests and demands" and spells out "conditions in relation to loyalty and service," we infer the shape of Shivaji's offers from the contour of the replies.3
The exchange begins not even between the two men directly. In July 1656 Aurangzeb wrote to Multafat Khan, the governor of Ahmadnagar, telling him to "keep the path of correspondence with Shivaji open" and to write in a way that would "encourage him to offer his loyalty" — making Shivaji "more and more willing" to serve the Mughals.4 So the channel itself was a Mughal instrument, opened deliberately to draw the rebel in. Shivaji used it for the opposite purpose: to buy time and cover.
The engine of the whole correspondence is a single proposal with two buried purposes. After pushing into Bijapur's domains, Shivaji wrote asking to keep those captured areas on behalf of the Mughals — and, if granted a mansab and revenue rights, to merge them with the empire and stay on "the path of allegiance and service."5 On its face, submission. Underneath, Purandare names the twin objectives Shivaji left unstated: deny Bijapur its claims, and block immediate Mughal aggression against himself.6 One offer, two enemies neutralized, neither openly named.
The mechanism is the false-submission gambit. You dress a defensive land-grab in the costume of loyalty, so that accepting your "service" forces your two largest rivals to leave you alone — Bijapur because the land is now Mughal-claimed, the Mughals because you are now their loyal agent. It works only if the other side cannot see through it. Aurangzeb saw through it almost at once. That is what makes the exchange a contest rather than a con: Shivaji played the gambit knowing it might be read, and Aurangzeb returned the courtesy knowing it was a gambit.
This page is the vault's clearest specimen of letters as a probing weapon, and it feeds three neighbours. It gives Diplomatic Correspondence as Statecraft a case where the documentary record is deliberately one-sided and the analyst must reconstruct one party from the other's replies — correspondence not as transparent record but as a mirror with one face missing. It supplies The First Notice to the Mughals 1657 its diplomatic backdrop: the Junnar and Ahmadnagar raids land inside this letter-writing window, so the same Shivaji professing loyalty in ink was scaling Mughal walls with rope-ladders. And it pairs with The First Maratha Ships and the Viegas Brothers as the inland twin of a coastal story — both show Shivaji read through the small administrative gesture, the harbour permit and the mansab term, rather than the grand declaration.
The decisive scene is not any single letter but the gap between Aurangzeb's public replies and his private ones — and it is in that gap that you watch a man read his enemy correctly while pretending not to.
Begin with the public face. After Shivaji's submissive letters of late 1657 and early 1658, Aurangzeb's responses run the formal course. To Nasiri Khan he approves the reply sent to "the defeated and disgraced Shivaji," allowing that if the rebel "sends a trusty agent" and his demands prove "worthy of being reported," they may be forwarded.7 In April 1658, writing directly to Shivaji from Ujjain, he is grander still: "Your offences are too many and don't deserve any pardon. But you have demonstrated your intention of being loyal... Ours is not a durbar that persecutes, so I have pardoned your offences."8 Then, after his second coronation, the 14 July 1659 letter — the high courtesy peak: God has adorned his banners with victory, Shaista Khan is the new Deccan subahdar whose orders Shivaji must obey, and "a robe of honour is conferred on you for your exaltation."9
Now the private face. Even while sending these letters professing acceptance of Shivaji's loyalty, Aurangzeb wrote to Mir Jumla a warning that strips off the courtesy entirely: attend to the district, "as the son of a dog is waiting for his opportunity."10 And he wrote to the Adil Shah — the same Bijapur ruler who had just signed peace with the Mughals — to "Expel Siva who has sneaked into the possession of some forts," advising that if Bijapur wanted Shivaji's services it should park him with jagirs in the Karnatak, "far from the imperial dominions, so that he may not disturb them."11
Set the two faces side by side and the case study writes itself. The robe of honour and the "son of a dog" letter were composed by the same hand in the same season. Aurangzeb was not deceived for a moment; he had decided early that Shivaji "was not as innocuously out to toe the line... as he pretended to be."12 The earliest warning came in the very first phase: even as he welcomed the on-behalf-of-the-Mughals offer to his wazir Muazzam Khan, he attached the threat that if Shivaji did not follow orders, "they would be trampled underfoot by our powerful army."13 The correspondence, then, is not one man fooling another. It is two men who have each seen through the other, continuing to write in the polite register anyway — because the polite register is the weapon. Every courteous letter bought Shivaji consolidation time and gave Aurangzeb a paper trail of a rebel's professed submission to cite when he finally moved.
You are handed Aurangzeb's file on Shivaji and told to find out what the prince really thinks. Do not read the letters to Shivaji first. Read them last. Start with the notes to his own men — to Multafat Khan, to Muazzam Khan, to Mir Jumla, to the Adil Shah. There you find the unguarded sentences: keep the channel open to draw him in; trample him if he disobeys; the son of a dog is waiting; expel Siva. Now turn to the letter that grants the robe of honour and read it again. It does not read like trust anymore. It reads like a man laying down a paper record of a rebel's promises so that breaking them will be the rebel's crime, not the emperor's.
Hold both layers at once. The letter Shivaji received says pardon and exaltation; the letter Mir Jumla received says he is waiting for his chance. Neither is the truth alone — the truth is the distance between them. When you read any correspondence between rivals, find the private channel first. The public letter tells you what each man wants the other to believe; the private note tells you what he is preparing to do. A diplomat who reads only the addressed letter has read the costume and missed the body underneath.
The exchange fails — as probing — when one side believes the courtesy. The diagnostic signs are clear. You take the robe of honour as proof of trust rather than as a placeholder. You file the submissive letters as a settled fact instead of as a stalling move with an expiry date. You assume that because a rival has written "I will stick to the path of loyalty," his next act will be loyal — when in fact, between two of these very letters, Shivaji scaled the walls of Mughal Junnar with rope-ladders and rode to the gates of Ahmadnagar.14
The deeper failure sign is mistaking the salutation for the strategy. Aurangzeb's "defeated and disgraced Shivaji" and his makhu epithet — the Perso-Arabic suffix meaning "one who has invoked someone's wrath," which he attached to "Siva" — are not idle insults; they are tells that the courteous parts of the same letters are tactical, not sincere.15 If you cannot hold contempt and ceremony in the same document at once, you will read this correspondence wrong, and you will be surprised by the war that both men already knew was coming.
The whole reconstruction rests on Purandare, a [POPULAR SOURCE], working from a structurally lopsided archive. Aurangzeb's letters survive — to Multafat Khan, Muazzam Khan, Nasiri Khan, Mir Jumla, the Adil Shah, and directly to Shivaji — but Shivaji's originals are gone, "likely... part of the papers burnt by the Mughals" after his death.16 [SINGLE SOURCE on Shivaji's side] Every claim about what Shivaji wrote is an inference from what Aurangzeb wrote back; the page should be honest that we are reading the rebel's voice as an echo, never directly.
That asymmetry cuts in a specific direction. The surviving record is the Mughal record, so the framing — "defeated and disgraced," "son of a dog," makhu — is the victor-adjacent framing of a dynasty that later fought Shivaji for decades. The contempt may be genuine, may be performative court-Persian convention, or may be both; the source does not let us cleanly separate them. Open questions: What exactly did Shivaji's lost letters claim, and did he ever expect the submission to be believed, or only to be useful? And did the makhu epithet and the robe of honour truly sit in the same letters, or has the compression of a popular narrative collapsed a longer, messier exchange into a tidy irony?
Purandare is the lone modern narrator, so the tension that matters is the one inside the documentary record he draws on — and it is a tension of register within a single author, Aurangzeb, rather than between rival chroniclers. The Mughal letters converge on one undeniable fact: Aurangzeb understood Shivaji early. Purandare states it plainly — Aurangzeb "was among the earliest to have understood Shivaji's ideas and objectives," seeing that the Maratha was "not as innocuously out to toe the line... as he pretended to be."17 On that, the public letters and the private notes agree. The split is in tone, and it is severe. The directly-addressed letters extend pardon, recognition, and a robe of honour; the side-channel notes to Mir Jumla and the Adil Shah call Shivaji a son of a dog and order his expulsion. What the divergence reveals about the concept is that, in elite statecraft, courtesy and contempt are not contradictory positions but coordinated tools deployed by the same mind at the same moment. The robe of honour was not insincere instead of being contemptuous; it was contemptuous and useful — a way of binding Shivaji to professed promises Aurangzeb fully expected him to break. Because we lack Shivaji's letters, we cannot run the same two-register test on him, but the inference is symmetric: a man who scaled Junnar's walls between submissive letters was practising the identical doubleness from the other side. Two men, one method, mutually understood — that is the convergence hiding under the apparent imbalance of the surviving record.
Letters between rivals who have each seen through the other are not really about information transfer; they are about influence, signal-reading, and the management of an opponent's perception — which is why this diplomatic episode reaches into behavioral influence and into the intelligence tradition.
Sun Tzu — Intelligence and the Five Spies. Sun Tzu treats foreknowledge of the enemy's intentions as the supreme advantage, obtained by reading the opponent rather than trusting his declarations. Aurangzeb's correspondence is a working model: he keeps "the path of correspondence open" not to communicate but to gather — every submissive letter from Shivaji is an intelligence sample, and his private "son of a dog is waiting for his opportunity" is the analyst's conclusion drawn from the pattern, not the words.18 The structural parallel is precise: both treat the enemy's professed loyalty as evidence to be interpreted against his behaviour. The tension Sun Tzu's frame surfaces is that intelligence without the freedom to act on it is frustration — and Aurangzeb's letters drip with exactly that, scolding his own generals for slow progress while he reads Shivaji perfectly. The insight neither domain produces alone: in this exchange the channel itself is the spy. Aurangzeb did not need to plant an agent in Shivaji's camp; he opened a correspondence and let the rebel's own letters, read against his raids, betray him. Diplomacy here is espionage conducted in the open, with the target's consent.
Religious Tolerance as Political Philosophy. Aurangzeb frames his pardon in a striking phrase — "Ours is not a durbar that persecutes" — even as he prosecutes a war of faith elsewhere, boasting that God has crushed "enemies of the Faith."19 The handshake is a tension, not a parallel: the language of clemency and the language of religious war coexist in the same correspondence, deployed toward the same rebel. The political-philosophy frame asks whether tolerance is a principle or a tactic; this exchange answers that, in Aurangzeb's hands at this moment, the merciful-durbar pose is an instrument of binding a useful rival, not a creed. The insight neither the diplomatic nor the religious frame yields alone is that the rhetoric of tolerance can be a control mechanism — a way of placing the rebel in your debt — entirely detachable from any underlying disposition to tolerate. Read beside the genuine question of religious tolerance as governing philosophy, Aurangzeb's "not a durbar that persecutes" warns that the words and the practice can move in opposite directions within one letter.
Strategic Patience and Calibrated Retreat. Shivaji's submissive letters of late 1657 and 1658 are not sincere capitulation; they are timed concessions inside his own stated rhythm — two steps forward, one step back, then four forward and two back, always the net gainer.14 The handshake to calibrated retreat is a structural parallel: a written submission functions exactly like a battlefield withdrawal, a temporary yielding bought to consolidate ground already taken. The tension is that calibrated retreat usually describes physical maneuver, while here the "retreat" is a piece of paper professing loyalty to the very power Shivaji had just raided. The insight neither the diplomatic record nor the maneuver-doctrine produces alone is that a letter can be a retreat — that the same leader who pulls his cavalry back from Junnar pulls his rhetoric back into the language of service, and both moves serve the identical purpose of trading apparent ground for real time. Diplomacy and tactics are not separate departments in Shivaji's method; the submissive letter and the calibrated withdrawal are the same instrument written in two media, which is why his correspondence cannot be read apart from his campaign calendar.
The Sharpest Implication. The unsettling thing this correspondence says is that mutual understanding does not de-escalate conflict — it can run perfectly in parallel with it. Both men saw through each other early and kept writing courteous letters anyway, because the courtesy was the contest. The reader who assumes that "if we just understood each other we would cooperate" is reading these letters backwards. Aurangzeb and Shivaji understood each other with great precision, and that understanding sharpened the war rather than softening it. Clarity about an opponent is a weapon, not a peace offering.
Generative Questions. (1) When the only surviving half of a correspondence is the victor's, how much of the loser's "voice" are we actually reconstructing, and how much are we ventriloquizing through the winner's framing — what would Shivaji's burnt letters have shown that the replies cannot? (2) Is the false-submission gambit — dress a land-grab as loyalty to force two rivals to leave you alone — a repeatable structure across history's rising powers, and what distinguishes the cases where the larger power "accepts" it as a delaying convenience from the cases where it simply crushes the upstart?