The khanqah accession, and the question of who tells it
Abdullah Qutb Shah died on 21 April 1672 with no son and three sons-in-law, and the succession was settled inside the palace. Sarkar records that Sayyid Ahmad, husband of the eldest daughter and the long-serving wazir, claimed the throne, and that Ma Sahib ‘armed her foreign slave-girls and stood within the palace sword in hand to defend her rights’. The prize went instead to Abul Hasan, a distant kinsman with no following of his own, carried in by the general Sayyid Muzaffar — though Fryer, Tavernier, Khafi Khan and the Ma’asir-i-Alamgiri, whom Sarkar lines up against one another, already disagreed about who had chosen him and what he was [2].
Onto that scramble Deccan tradition laid a providential story, which Sarkar takes from a Persian text he names as the Qutb-numa: Abul Hasan had spent sixteen years as a disciple of the saint Sayyid Raju Qattal, and was seized from the hermitage and married to Abdullah’s third daughter in wedding clothes cut for another man, his master having ‘laughingly daubed Abul Hasan with red clay, saying “Thus I dye your fingers with henna like a bridegroom’s.”’ [1]. The same outline survives in modern Hyderabadi retelling [24]. Nothing contemporary from Golconda confirms it, and it converts a palace coup into a saint’s prophecy. The hostile framing is no less interested: the same sentence of Sarkar reports that ‘On account of his low tastes, dirty habits and debauched life, the Court and the gentry had shunned him’ [1], a verdict drawn from Mughal-side material written after the conquest.
Madanna and Akkanna: a coup, then a fiscal machine
The most consequential thing Abul Hasan did was an act of self-defence, not of tolerance. Sayyid Muzaffar, having made the king, set out to reduce him to a cypher, and the king answered by buying Muzaffar’s own establishment: ‘Abul Hasan bought over Madanna, the Brahman factotum of Muzaffar, and through him corrupted most of the captains of his personal followers, so that one day Muzaffar was quietly deprived of his wazir-ship, which was conferred on Madanna, with the title of Surya Prakash Rao’ [3]. That is about 1673; Akkanna rose with his brother, and the ministry ran until both were killed in 1686. Golconda’s Brahmin administration began as the instrument by which a king with no faction broke the faction that had enthroned him.
What the brothers held is contested. Sarkar’s Persian-derived tradition makes Madanna wazir and Akkanna commander-in-chief; Gijs Kruijtzer finds no source giving Madanna a Peshva title, identifies his real position as majmu’adar shahi — king’s collector, bookkeeper — and notes that the contemporary court miniatures caption both brothers as chief bookkeepers. The reform is legible whatever the titles were: the elite came off revenue-farmed jagirs onto treasury salaries, civil tax-collection rights were sold to civilians for a lump sum paid in advance, and policing stayed Muslim and treasury-paid while fiscal administration opened to Brahmins and Baniyas — the same ministry collecting jizya from the Hindu majority [19].
The argument they were making survives in Akkanna’s own words, through a Dutch East India Company channel in 1683: their government was ‘fullheartedly devoted to the welfare of the country’, he wrote, ‘while that of the Moors is only to the end of becoming rich and then to leave for those places which they consider to be either their fatherland or holy’ [19]. Kruijtzer reads that both ways at once, as early patriotism and as an attack on a Muslim elite as extractive outsiders, and this entry carries both.
Patronage and its limits: Kuchipudi and Bhadrachalam
Two Telugu-country episodes fix the reign in modern Andhra and Telangana memory, and both show what a Golconda king could do rather than what he believed. The first is the tax-free grant of Kuchipudi village to its Brahmin dancer-families after the sultan watched them perform, dated by one account to 1675, when ‘the Kuchipudi village was given as an agraharam (grant of land) to the Brahmin artists of the village’ [21], and by the standard accounts to 1678, when he ‘granted the dancers lands around the Kuchipudi village, with the stipulation that they continue the dance-drama’ [22]. Neither quotes the firman, so the discrepancy stands.
The second runs the other way. Kancherla Gopanna, remembered as Bhadrachala Ramadasu, was Madanna’s nephew and revenue officer at Bhadrachalam. Sarkar records that he spent six lakhs of pagodas of government money on halls, shrines and corridors for the god Ram, was arrested by royal order and put to torture, and was released, in the devotional tradition Sarkar passes on, through the miraculous intervention of Ram and Lakshman, who appeared as men and paid the deficit on his behalf [3]. The state that endowed Kuchipudi jailed a Hindu officer for diverting revenue to a temple, and made his uncles collect jizya from Hindus [19]. Tolerance is the wrong frame; a fiscal order that recruited widely and collected ruthlessly is the right one.
The Maratha subsidy and the manufacture of a pretext
By the mid-1670s Golconda’s security was a purchased commodity. Madanna had contracted a subsidiary alliance with Shivaji, and when Shivaji reached Hyderabad in February 1677 with some fifty thousand men the terms Sarkar reports are exact: the Sultan was to pay ‘a subsidy of 3,000 hun a day’, five thousand men and a train of artillery, in return for ‘such parts of his conquests in the Kamatak as had not belonged to his father Shahji’ [4]. The itemisation comes from the Mughal-Persian and English-factory side; Grant Duff, writing from Maratha material, states that no authentic record of the secret compact was preserved by the Mahrattas [20].
The alliance became the charge sheet. The Ma’asir-i-Alamgiri lists against Abul Hasan his state’s Shi’i faith, its support of Hindu practices and its support of Maratha depredations in Mughal territory [17], while Sarkar shows the policy from the other end: an insolent envoy was sent to goad the king into defiance and furnish a pretext, and he declined to be provoked, asking how, if he was no Padishah, Aurangzeb could be Padishah of Padishahs [2]. The English at Fort St George afterwards recorded the same accusation, that he had offended ‘in encouraging the Brahmans and discouraging the Moors’ [12]. A first settlement came in 1685: a crore and twenty lakhs of rupees for past dues, two lakhs of hun a year, the surrender of Malkhed and Serum, and, as the second clause, ‘He must dismiss Madanna and Akkanna’ [5]. What followed five months later was the enforcement of a treaty article.
How the brothers died: four witnesses, no verdict
March 1686 produced four incompatible accounts of one murder, and Sarkar prints them without choosing. The imperial chronicle makes the king its author: Abul Hasan ‘had cut off and sent to Shah Alain the head of Madanna Brahman as a proof of his own submission’ [6]. Khafi Khan, whom Sarkar rates as well informed about Hyderabad, says the killing was done without the king’s knowledge, and that ‘Abul Hasan put off arresting the two brothers till his army lost patience and rose against them’ [7]; Bhimsen, in Shah Alam’s camp, agrees with him.
The European witnesses split the same way. Havart, the Dutch eyewitness of 1693, has a crowd act with royal sanction: ‘Under orders from the Sultan they murdered the two brothers in a most ignominious way’ [8]. Manucci, in Golconda that year, makes the king act first and personally — ‘the first thing he did was to kill the Brahmans and fling them down from the fortress walls’ [13]. A Telugu antiquarian tradition gives a fifth shape, a plot encouraged by members of the Qutb Shahi family and executed by an African slave named Jamsheed [25]. Each witness had a reason to place or withhold blame; all agree that Golconda lost the administration that had raised its revenue for thirteen years, ten months before the besieging army arrived.
The siege of 1687: mines, famine, and a purchased postern
Siege operations opened on 28 January 1687 against a fort that held until 21 September. Aurangzeb brought roughly fifty thousand infantry, comparable cavalry and about a hundred siege guns, yet the three great mines of some thirty-seven thousand pounds of powder each achieved nothing: two were countermined, the first backfiring and killing over a thousand of the besiegers, and the third never ignited. The besiegers were also fighting the weather — northern Coromandel had been failing for want of monsoon since October 1685, famine and epidemic had spread across the coast by 1687, and by early June the Mughal army had itself run out of food in a countryside where planting had stopped with the invasion [17]. The conquest landed on a region already broken by climate.
It ended by purchase. Sarkar names the seller as ‘An Afghan soldier of fortune, named Abdullah Pani, surnamed Sardar fvhan’, by then one of the two highest officers in the fort, who opened a postern before dawn; one officer refused the money, Abdur Razzaq Lari riding alone into the incoming flood and reaching the gate of the Bala Hisar covered in wounds [9]. Manucci reports independently that he held the place until the garrison was bought, and gives him eighteen wounds where Sarkar gives seventy [14].
The capture of the king is reported twice, irreconcilably. Khafi Khan, through Sarkar, has him console his wives, order breakfast at the usual hour, greet Ruhullah Khan first and invite his captors to eat, explaining his composure by a God who had made him a beggar, then a king, and now a beggar again; Ishwardas has him merrymaking with his dancing-girls [10]. Manucci, an eyewitness, has him dragged out amid shrieking women, kept more than an hour in the sun, at first refused water by his guards, and later beaten ‘unmercifully’ [14].
Daulatabad, the reckoning, and what survives
The kingdom was worth taking. Encyclopaedia Iranica puts the motive without ornament: Qutb Shahi wealth ‘included fertile agricultural tracts, maritime trade, and diamond mines, excited the cupidity of the Mughals’ [15]. Khafi Khan, via Sarkar, gives the spoils as nearly seven crores of rupees in cash besides plate and jewels, and the annual revenue of the conquered kingdom as ‘two krores and 87 lakhs of Rupees’ [11] — a yearly yield about two-fifths of the one-off haul. The dominant modern reading holds that the annexation deepened the empire’s jagir crisis, the rewards needed to entice enemy officers over exceeding what the annexed lands produced [17]. Aurangzeb never returned to Delhi.
Abul Hasan was sent to Daulatabad, by Manucci’s account after refusing to account for his spent treasure, to what he calls ‘a close prison on a great eminence’; the emperor married two of the king’s daughters to two of his own officials to show how little he esteemed that family [14]. Sarkar, who walked the site in October 1916, has ‘the most luxurious king of the Deccan’ sighing out his captive life there ‘on a pension of Rs. 50,000 a year’ [10], a figure for which he gives no source. When the king died is unsettled: Nile Green lists among the royal burials at Khuldabad ‘the last king of Golkonda Abu al-Hasan (d. 1111/1702)’ [23], while the standard administrative monograph puts his death about three years after the fall.
What survives is legible in two registers. UNESCO’s tentative-list dossier records within the citadel of Golconda the ‘Akkanna-Madanna Offices’ and ‘Ramdas Jail’ — the ministry and its most famous prisoner both still standing — while the necropolis holds the tombs of five of the seven sultans, ending with Abdullah; Abul Hasan has none there [16]. After 1687 the Mughals destroyed the Qutbshahi palaces and moved the regional capital to Aurangabad [15]. The name went the same way: an honorific meaning child saint or pious king became, in Urdu and Hindi, the word for ‘an autocratic or despotic ruler’ [18]. Whether the sensualist behind that inversion existed is disputed — Sarkar keeps the stereotype, Sajjad Shahid calls it a Mughal propaganda campaign that turned a mild and much adored monarch into a characterless wretch and licensed war on a fellow Muslim sovereign [24], a framing the imperial sadr is said to have objected to [17], and Grant Duff reports him ‘notorious for dissipated habits in his youth, was, on ascending the throne, completely reclaimed’ [20]. The split is the content.
Dates in this essay
Fall of Golconda: Sarkar, on the Ma'asir-i-Alamgiri and Khafi Khan, dates the entry through the postern to about three in the morning of 21 September 1687, and the archive takes that capitulation as the terminal date of the sultanate. Wikipedia's siege article agrees on 21 September in its text but runs the reign to 22 September in its infobox. William Irvine, annotating Manucci, converts the Ma'asir's 24 Zu'l Qa'dah 1098 AH to 1 October 1687 New Style, and the Fort St George Diary of 12 November 1687 likewise reports the fort taken on the 2nd of the previous month New Style, about midnight. Most of the spread is Old Style against New Style. Manucci's own narrative, placing the end in February 1687, is simply wrong.
Death of Abul Hasan in captivity: four rival dates, none from a Golconda document. Nile Green ('Auspicious Foundations', 2004) lists him among the Khuldabad royal burials as d. 1111/1702; J. F. Richards (Mughal Administration in Golconda, pp. 67-68), as reported in Wikipedia's siege article, has him dead about three years after the fall, c. 1690; Sajjad Shahid gives 1700 after fourteen years of captivity; the Wikipedia biography says twelve years, i.e. 1699, and flags its own sentence as uncited. The two Tier-A authorities, Green and Richards, are the furthest apart. The archive records the spread and does not choose.
Killing of Madanna and Akkanna: the Ma'asir-i-Alamgiri places the murder early in March 1686 and the audience at which the head was produced on 16 March 1686; Irvine's note to Manucci gives the head as sent in on 1 Jamada I = 26 March 1686; Live History India dates the ambush itself to 16 March 1686; Wikipedia's siege article says March 1686, while its article on the brothers says 1685 in its lead and reports a third version placing a public killing in November 1685, resting on a Bijapur report of 16 November 1685 that Sarkar judges unlikely to be fact and reads as a garbled account of what the king had only promised.
Ministry of Madanna: Sarkar dates the transfer of the wazir-ship from Sayyid Muzaffar to Madanna to about 1673 and runs the ministry to the assassination in 1686; no Golconda document fixing the appointment was located, so the start date rests on Sarkar's Persian sources alone.
Grant of Kuchipudi village: Sahapedia gives 1675, Wikipedia and the standard reference accounts (and Live History India) give 1678. No source consulted quotes the copper-plate firman, so the discrepancy stands open.
Reign dates: the archive dates the reign from Abul Hasan's assumption of sovereign power at Golconda after Abdullah's death on 21 April 1672 to the capitulation of the fort in September 1687, with the New Style reckoning of 1 October 1687 noted. Encyclopaedia Iranica's article gives the dynasty's span as 918-1098/1490-1686, ending a year early against its own statement that the conquest occurred in 1098/1687.
Where authorities disagree about when a state began or ended, the archive follows one rule — a polity is dated from the assumption of sovereign power in its own capital, and to the settlement that extinguished it — and records the alternatives rather than hiding them.
Open questions
What is the Qutb-numa, the Persian text Sarkar cites (pp. 90-94) for the entire hermitage-and-henna story of Abul Hasan's marriage? Its date, authorship and distance from 1672 are the whole transmission question, and neither the text nor any scholarly description of it could be opened for this entry.
What does Daniel Havart's Op en Ondergang van Cormandel (Amsterdam, 1693) actually say, in the original, about the rise and death of Madanna and Akkanna, and what do his portraits of the brothers and his plate of their bodies dragged through the streets show? It was read here only at second and third hand, through Sarkar quoting the Madras Annual Report on Archaeology 1914-15 and through Irvine's footnote to Manucci.
What is the full argument of Gijs Kruijtzer, 'Madanna, Akkanna and the Brahmin Revolution' (JESHO 45.2, 2002)? The Brill page opens but the article is paywalled; only the abstract and secondary summaries of its findings were read, and this is the single most important modern scholarly treatment of the ministry.
Does the Golconda chapter of Kruijtzer's Xenophobia in Seventeenth-Century India (Leiden University Press, 2009), or the 2008 dissertation behind it, settle the office Madanna held and whether Akkanna ever commanded troops? Both are marked open access at the Leiden repository, but the PDF did not resolve.
What exactly does J. F. Richards, Mughal Administration in Golconda (Clarendon, 1975) argue about Abul Hasan's death date (pp. 67-68), the siege, and the post-conquest administrative overhaul? It is cited throughout the secondary literature used here and was not accessible.
What does H. K. Sherwani, History of the Qutb Shahi Dynasty (1974), p. 652, say about the etymology of 'Tana Shah', and how does his dynastic narrative differ from Sarkar's? Not accessible.
Can the collapse of food security be dated precisely from Archisman Chaudhuri's 'The El Nino of 1685-1687 in Golconda and Northern Coromandel' (2022)? Only the abstract and its VOC-archive endnotes were read; the rainfall table and famine chronology in the chapter body are paywalled.
What does the Kuchipudi copper-plate firman itself say, and does it date the grant to 1675 or 1678? Every account paraphrases it and none quotes it.
Did Abul Hasan in fact issue bilingual Persian-and-Telugu edicts, as asserted by Sajjad Shahid? No inscription corpus or epigraphical report confirming bilingual farmans under him was located.
Is the grant of the revenues of Melupaka village in perpetuity to Chandramaulishwara at Kanchipuram, attributed to Abul Hasan in a newspaper column, supported by any inscription or archival record? It could not be verified from anything read here and is not relied on in this entry.
Is Abul Hasan's grave at Khuldabad identified and marked today, and what is its relation to the shrine of Razi al-din Raju Qattal? Wikipedia describes a modest grave within that shrine and Nile Green lists him among Khuldabad's royal burials, but no site survey or ASI notification was consulted.
What are the true dimensions and form of the Shah Raju Qattal dargah at Misrigunj, and is its construction in fact attributable to Abul Hasan? Popular sources variously report over 90 feet on 110 pillars, 165 feet on 110 pillars, and 50 feet on 112 pillars; no measured survey was found, and the accompanying claim that it is the tallest tomb in India is unverified and is not repeated here.
On what authority does Sarkar give the captive king a pension of Rs. 50,000 a year at Daulatabad? He cites nothing in the passage read, and no corroboration was found.
What did the 1656 treaty clause actually stipulate, by which Abdullah's second daughter was married to Aurangzeb's son Muhammad Sultan with a proviso about the succession, and what became of that prince? The treaty text would settle whether a live Mughal claim to the Golconda throne existed in 1672.
Are Sarkar's Abdullah Pani, surnamed Sardar Khan, and the Sarandaz Khan named in Wikipedia's siege article the same man under variant transliteration, or two different officers? The identity of the man who opened the postern in September 1687 is unresolved.
What did Abul Hasan himself compose? Sarkar credits him with musical skill of no mean order and with settling craftsmen whose manufactures were famous throughout India, but names no work, no craft and no workshop, and no composition attributed to him was located.
Where are the Golconda court miniatures whose captions Kruijtzer reads as identifying Madanna and Akkanna as chief bookkeepers, and the BnF album painting of the Sufi singer Shir Muhammad visiting Abul Hasan (ca. 1720)? No museum catalogue record was opened, so no accession numbers or collection details can be given.