The Mughal Historian Khafi Khan Called Shivaji a Reprobate, Then Wrote That His Troops Were Ordered to Spare Mosques, the Quran and Women

At eight o’clock on the morning of 6 June 1674, the Maratha minister Niraji Pandit brought the English envoy Henry Oxinden into the coronation hall at Raigad. Jadunath Sarkar rebuilt the scene from Oxinden’s letters and memorial in the East India Company’s Surat records. The envoy “bowed from a distance, and his interpreter Narayan Shenvi held up a diamond ring as an offering from the English to the Rajah.” Shivaji “took notice of the strangers and ordered them to come to the foot of the throne, invested them with robes of honour, and then sent them back.”

On 11 June, Sarkar writes from the same records, Niraji Pandit sent word that “the Rajah had granted all our demands and articles, except our money passing current in his country.” On 12 June the eight ministers signed the treaty and handed it to Oxinden at Niraji Pandit’s house.

The outline of Shivaji’s life is already on tuput. This piece goes through the documents and slows down where they disagree.

A mass of gossiping bakhars

In April 1822 Mountstuart Elphinstone wrote to the historian James Grant Duff that his difficulty was to make a rational history “out of a mass of gossiping bakhars and gasconading tawarikhs.” A bakhar is a Marathi prose chronicle and a tawarikh is a Persian history. That letter opens Shivaji and His Times, and Sarkar goes on to name what Grant Duff valued more, the English factory records, “very important for fixing dates.” Stewart Gordon, whose 1993 Cambridge volume is the standard modern account, keeps to incidents confirmed by at least two reliable bakhars or by outside sources.

Every date below follows Sarkar. He gives them in the Old Style, the Julian calendar still used in England then, and says the New Style dates in continental records ran “usually ten days later” in Shivaji’s lifetime. Add ten days to match our calendar.

Torna: 1646 or 1647, and where the money came from

Sarkar dates Shivaji’s first major acquisition to “about 1646,” while the Bijapur court was disordered by the long illness of Sultan Muhammad Adil Shah. Shivaji, he writes, “had taken Torna fort from its Bijapuri commandant by some cunning device without fighting,” and found government treasure of 2 lakh hun, a gold coin, inside. Sarkar’s chronology marks the entry with a question mark and lists it just before the death of the steward Dadaji Kond-dev on 7 March 1647. In Gordon’s telling the taking of Torna was one of Shivaji’s first acts after Dadaji’s death in 1647, done “through stratagem (or perhaps bribery, depending on the text).”

The climb was not clean. In 1656 Shivaji moved against the More family, who held forts south of his lands under Bijapur, and the tactics, Gordon writes, “were brutal, including conscious treachery on Shivaji’s part,” and the bakhars all agree the campaign was “pure conquest and elimination of a rival.” The More father and four sons were largely wiped out.

Pratapgad, 10 November 1659: who struck first

Bijapur sent Afzal Khan with about 10,000 men to crush Shivaji, and after months of manoeuvring the two agreed to meet under truce in a clearing below Pratapgad.

The Marathi chroniclers, whom Sarkar follows, tell the meeting this way. Shivaji wore chain armour under his tunic and a steel cap under his turban. Hidden in his left hand, Sarkar writes, was “a set of steel claws (baghnakh) fastened to the fingers by a pair of rings,” and a thin dagger called the scorpion lay up his right sleeve. Afzal Khan opened his arms to embrace him, then “tightened his clasp, and held Shiva’s neck fast in his left arm with an iron grip,” and stabbed at his side. The armour turned the blow, and Shivaji tore open the Khan’s belly with the claws.

The Persian accounts reverse the order. Bhimsen Saksena, a Kayastha, a Hindu writer caste, in Mughal service who wrote his memoir after retiring in 1707, says: “While the said Khan was embracing Shivaji, the latter struck into his (Khan’s) stomach the lethal weapon known by the name bagh-nakh also called bichava amongst the people of the Dakhan so hard that his intestines came out.” Khafi Khan also has Shivaji strike first, with a weapon hidden in his sleeve, and says Afzal Khan came “without arms.” For Gordon both men came armed, and “the exact sequence of events in the truce tent will never be known.”

An appendix to Sarkar’s book weighs the evidence. An English factory letter from Rajapur dated 10 December 1659 says Afzal Khan was instructed to secure Shivaji by “pretending friendship with him.” A Persian history of the Bijapur court, which Sarkar translates, has the Sultan telling Afzal Khan to ignore any offer of submission. A decade earlier, Gordon adds, Afzal Khan had used a truce to imprison a Hindu general. Sarkar concludes that Afzal Khan struck first and that Shivaji “only committed what Burke calls a ‘preventive murder.’”

Shivaji’s hidden troops then attacked the leaderless camp. The English factors at Rajapur heard that 3,000 men died. Many of the Bijapuri troops, Gordon notes, were themselves Marathas serving their sultan.

The night raid at Pune, 5 April 1663

Shaista Khan, Aurangzeb’s maternal uncle and viceroy of the Deccan, had moved into the house in Pune where Shivaji grew up. The attack came on the night of Sunday 5 April 1663 by Sarkar’s chronology, the sixth day of Ramzan, when the cooks were up preparing the meal before dawn. Shivaji entered the camp with 400 men, the raiders broke through a bricked-up wall into the harem, and in Sarkar’s account Shivaji cut off the Khan’s thumb with one stroke before the lamps in the room were put out.

Khafi Khan heard it from his father, who served Shaista Khan. In his version the Khan seized a bow and shot one attacker, “but he got up to the Amir, and cut off his thumb,” and two slave-girls “dragged him from the scene of strife to a place of safety.” Khafi Khan wrote after 1730, Sarkar notes, and the earliest records are the English factor Gyffard’s letters to Surat of 12 April and 24 May 1663, which carry Shivaji’s own account.

Surat, 1664 and 1670, through English eyes

Shivaji reached Surat at 11 o’clock on the morning of Wednesday 6 January 1664. The Mughal governor shut himself in the fort. The English president, Sir George Oxenden, mustered 150 Englishmen and 60 peons in the factory, and the factors wrote that “it was thought more like Englishmen to make ourselves ready to defend our lives and goods to the uttermost than by a flight to leave money, goods, house to merciless people.”

Shivaji demanded three lakh rupees from the English. Oxenden answered, “We are ready for you and resolved not to go away. But come when you please.” Shivaji did not come. The factors’ letter to the Company of 28 January says a Mr Smith “was present when he cut off more than 26 hands in one day, and as many heads.” Gordon, working from the European accounts, calls the sack “less violent than many” and says news of a Mughal army cut it short.

The English were there to ship Indian cotton cloth to Europe, and Gordon notes they were weaker than the Dutch on this coast in securing pepper and spices. When Shivaji returned in October 1670 he withdrew on 5 October and sent a letter demanding twelve lakh rupees a year, or he would come back and burn what remained. The revenue claim called chauth was, in Gordon’s definition, “protection money, extorted yearly” from territory Shivaji did not rule.

Purandar and Agra, 1665 and 1666

Jai Singh, a Rajput general from Amber, besieged Purandar in the spring of 1665. By the Treaty of Purandar on 12 and 13 June, Gordon writes, Shivaji gave up 23 forts and kept twelve, including Raigad. To Gordon it was “not a surrender” but the result of long negotiation.

Shivaji left for Agra on 5 March 1666. At court on 12 May, Gordon says, he gave the Emperor 1,000 gold mohurs and 2,000 rupees, was not spoken to, and was made to stand among low-ranked nobles. Sarkar’s reconstruction has him say, “You have seen, your father has seen, and your Padishah has seen, what sort of man I am, and yet you have wilfully made me stand up so long. I cast off your mansab,” his rank in the Mughal service. When he asked leave to retire to Benares as a holy man, Aurangzeb replied, “Let him turn faqir and live in Allahabad fort.”

Shivaji escaped on 19 August. The famous version has him sending out baskets of sweets for Brahmans each evening until the guards stopped searching them, then riding out in one basket with his son in another. Sarkar tells it in full from Sabhasad, whose bakhar was written for Shivaji’s son Rajaram, and from Khafi Khan, and he notes that the Alamgirnama, the official history of Aurangzeb’s reign, and the French traveller Bernier give it one sentence each. The court newswriters’ reports, Gordon says, mention no baskets, only speculation about the route, and he thinks Shivaji more likely bribed his guards after borrowing 66,000 rupees from a patron at court. The baskets are best called a tradition, which Sarkar tells as history and Gordon doubts.

The Kshatriya question and the throne

The coronation, Sarkar says, made Shivaji a king in law. Without it he could sign no treaty or grant that would last, and the Maratha families who had once been his equals sneered at him as an upstart. Gordon reads the ceremony as aimed at winning authority over those families, who helped pay for it through a tax, and notes that Oxinden was the only outside observer.

The obstacle was caste. The Bhonsles were known as cultivators, and Brahmans taught that a king must be a Kshatriya, a member of the warrior order. Shivaji sent for Vishweshwar, called Gaga Bhatta, a Benares scholar whom Sarkar describes as “the greatest Sanskrit theologian and controversialist then alive.” Gaga Bhatta accepted a pedigree drawn up by Shivaji’s secretary Balaji Avji that traced the Bhonsles to the Maharanas of Udaipur, the line of Maharana Pratap, and was paid “a huge fee.” Shivaji’s father, Gordon notes, wrote to Bijapur in 1656, “We are Rajputs.”

On 28 May Shivaji was invested with the sacred thread. In Sarkar’s account the assembled Brahmans then refused him the Vedic verses, asserting that no true Kshatriya survived in the modern age, though the sources differ on whether he finally received them. Two Brahmans pointed out that the cities he had burned had cost Brahmans, cows, women and children their lives, and that he could be cleansed for Rs 8,000, a figure Sarkar takes from Dutch records.

On 6 June, in Sarkar’s account, the artillery of every fort fired as he took the throne, and Gaga Bhatta held the cloth-of-gold umbrella over him and hailed him as “Shiva Chhatrapati,” or paramount sovereign. Sabhasad put the cost of the coronation at 1 crore 42 lakh hun, which Sarkar calls “incredible” and caps at 10 lakh hun. The treasury was exhausted, Sarkar writes, and Shivaji had to raid at once. His mother Jijabai died on 18 June, and a second, Tantric coronation followed on 24 September.

What he ordered his soldiers to do

Khafi Khan begins his account of Shivaji’s origins by calling him “the reprobate Sivaji,” and Dowson’s translation has him write that Shivaji “attacked the caravans which came from distant parts.” The same passage goes on: “he made it a rule that wherever his followers went plundering, they should do no harm to the mosques, the Book of God, or the women of any one.” Captured Qurans were treated “with respect” and given to Muslim followers. Later he returns to it. Shivaji “persevered in a course of rebellion, in plundering caravans, and troubling mankind; but he entirely abstained from other disgraceful acts,” and anyone who disobeyed his orders on the honour of Muslim women “received punishment.”

Endowments for Hindu temples, Muslim saints’ tombs and mosques alike are on Sarkar’s list too, and he prints a story from the 91-Qalmi bakhar with a caution, since the woman “is said to have been” the daughter-in-law of a former Bijapuri governor. When Shivaji’s officer Abaji Sondev presented her as a captive, Shivaji said, “If my mother Jija Bai had possessed your beauty, I too should have looked as handsome,” and sent her home under escort of 500 horsemen.

Gordon complicates the picture. Muslims served in Shivaji’s army from the first Pathan unit in 1656, he says, and Sidi Ibrahim was one of ten trusted commanders at Pratapgad. He concludes that Shivaji did not set out to create a universal Hindu rule and “did not represent proto-nationalism.” He also writes that yearly raiding was the ethic of the age, and that oral traditions of the Marathas as destructive raiders survive in areas around Maharashtra.

Aurangzeb’s temple order, the jizya and Shivaji’s letter

The Mughal court chronicler Saqi Mustaid Khan records Aurangzeb’s policy in the Maasir-i-Alamgiri. In April 1669, which Sarkar dates to the 9th, the Emperor, “eager to establish Islam, issued orders to the governors of all the provinces to demolish the schools and temples of the infidels.” In 1679 he ordered that from 2 April the jizya, a tax on non-Muslims, be collected in the capital and the provinces. Summing up the reign, the chronicler calls the tax “this rare piece of good work” and says “the Hindus had not been degraded to such a degree in any other period.”

A Persian letter against the tax, which Sarkar prints, went from Shivaji to Aurangzeb about the middle of 1679, Sarkar says, drafted by his secretary Nila Prabhu. It reminds the Emperor that Akbar reigned 52 years by a policy of “universal harmony (sulh-i-kul).” It then says: “In strict justice the jaziya is not at all lawful. From the political point of view it can be allowable only if a beautiful woman wearing gold ornaments can pass from one province to another without fear or molestation.” It tells Aurangzeb to levy the tax first on Rana Raj Singh, “the head of the Hindus.” The letter also warns that the Emperor’s remaining forts and provinces will be lost, “because there will be no slackness on my part in ruining and devastating them.” A footnote in Sarkar’s book says there is no evidence Shivaji broke the peace with the Mughals in January 1670 as a protest against the temple order, though an English factory letter and the Jedhe chronology place the two events side by side.

Ministers, forts and ships

By 1674, Sarkar says following Sabhasad, the council had eight ministers, the ashta pradhan, and it stayed at eight until his death. It was “in no sense a Cabinet,” Sarkar adds, and Shivaji kept the administration in his own hands. Each fort was held by three officers of equal rank, since “no fort was to be left solely under a haviddar, lest a single traitor should be able to deliver it to the enemy.”

English factory reports, Sarkar writes, never put Shivaji’s fighting vessels above 160 and usually at 60, and Gordon says he never took Janjira, the sea fort of the Siddis, an Abyssinian Muslim family. After Shivaji took Jinji in 1677, the Jesuit Andre Freire wrote that he had “constructed new ramparts round Jinji, dug ditches, raised towers and bastions, and carried out all these works with a perfection of which European skill would not have been ashamed.”

April 1680

On 23 March 1680, Sarkar writes, Shivaji was seized with fever and blood dysentery, and the illness lasted twelve days. The date Sarkar adopts is Sunday 4 April at noon, following Sabhasad and a Bombay letter to Surat of 28 April 1680, and notes that the full moon of Chaitra that year ran over both 3 and 4 April. His own 1920 edition said Sunday 5 April, and so did H. G. Rawlinson in 1915. Maharashtra marks the death anniversary on 3 April, according to a 2025 report from Drishti IAS.

In Sarkar’s record, the Muslim world blamed the curse of a saint of Jalna, and some in Maharashtra whispered that his wife Soyra Bai poisoned him. He treats the poison story as rumour, noting that in Europe at that time “hardly a sovereign died without the event being ascribed to poison.”

The Maasir-i-Alamgiri dates the death Friday 14 May 1680, forty days after the date Sarkar adopts, and says Shivaji “sank down into the pit of hell.” A footnote in Sarkar’s translation puts the death on 4 April.