A boy on the throne, and the men who governed in his name
The reign opens with a minority, and reading its regnal years as biography breaks down at once. Muhammad Shah III was between 9-10 years old when he ascended the throne on 30 July 1463 after the death of his brother, Nizam-Ud-Din Ahmad III [1]. Government continued under the previous reign's arrangement: a regency in which Khwaja Jahan Turk and Mahmud Gawan conducted affairs under the queen-mother's direction. It ended not with the sultan's majority but with an assassination -- Khwaja Jahan monopolised power and kept Gawan on the frontiers, until the queen-mother extracted a promise from her son and the minister was cut down in the king's presence. Only then did Gawan take the vacant title Khwaja Jahan and the office of Wakil-us-Sultanat.
Gawan's position rested on nothing structural. He was an Afaqi -- a newcomer from Iran and Central Asia -- in a nobility split between such foreigners and the Deccanis, the long-settled Muslim families and their Abyssinian allies. Wikipedia gives the standard framing: 'Mahmud, as he hailed from Persia and was likely Shi'a himself, was an Afaqi' [3]. The first half is well attested; the second is not. Ferishta's account of Gawan's character, which he says is taken in part from a life of the minister written by Mulla Abdul Karim Sindi, says the opposite: 'He was by persuasion a rigid Soony.' [19] No source read for this essay offers documentary evidence for the Shi'i attribution, which looks like an inference from Iranian origin and party alignment rather than a finding.
What a Russian merchant saw at Bidar
The reign has something rare for fifteenth-century India: an outsider's description with no stake in the dynasty. Afanasy Nikitin, a merchant from Tver writing for himself rather than a patron, was at Bidar, and his sultan is not the learned prince of the Persian chronicles: 'The sultan (of Beder) is a little man, twenty years old, in the power of the nobles. Khorassanians rule the country and serve in war.' [24] The Khorasanian grandee he names, Melik-Tuchar, keeping 200,000 men, is Malik-ut-Tujjar -- the title Gawan held; but Nikitin gives no personal name, and an earlier holder had been destroyed in the Konkan a generation before, so the identification is inference, not testimony.
His social note is harder than anything in Ferishta: the country people 'are very miserable, whilst the nobles are extremely opulent and delight in luxury' [25], carried on silver beds behind twenty gold-caparisoned chargers. He records Bidar as the chief town of Muslim Hindustan, its trade in horses, cloth, silk and slaves, and a night watch of a thousand mounted men each carrying a light. The dating of the visit is itself disputed, which decides which year of the reign this is: 'Nikitin left Russia in 1468 and spent the years 1471-1474 in India' [5] -- placing the 'little man, twenty years old' in exactly the years of Gawan's reorganisation.
The western war: forts, ports and the horse trade
The campaign that made the state's reputation was fought for trade routes. Richard Eaton states the object plainly: 'Gawan's protracted expedition, which lasted from 1469 to 1472, thus aimed at subduing both the hill-forts and the sea forts from which local chieftains had been harassing strategic trade routes.' [26] Deccan cavalry depended on horses imported by sea, and the Konkan chieftains held the passes and the landing places. Ferishta supplies the mechanics: Gawan sent his cavalry home as useless in mountain country, cut roads through forest with fire and axe, lay five months before Kehlna (Vishalgad) without reducing it, and took the fort next season by stratagem and money.
Goa followed, and it was amphibious -- Gawan moved on the port 'sending an hundred and twenty vessels to attack it by sea, while he marched with his army against it by land ; and before the Ray of Beejanuggur could oppose his design, he made himself master of the place' [10]. The capture took Vijayanagara's principal western outlet out of its hands and gave the plateau a port of its own. Belgaum, reduced in the same phase, is worth separating from the legend attached to it: when the defenders removed by night what was thrown into the ditch by day, Gawan shifted ground and 'began to form trenches and dig mines which till now had not been used in the Deccan' [17]. The source supports siege mining as the untried technique -- not the modern claim that Belgaum saw the Deccan's first use of gunpowder.
Eight provinces instead of four, and what the reform cost
In 1473 Gawan rebuilt the administrative map. Ferishta gives the headline in a sentence: 'The whole kingdom, which had been originally divided into four turufs, or provinces, and placed under four governors, was now distributed into eight.' [11] The usual modern gloss adds that he distributed them 'equally among the Deccanis and foreigners' [2], but no such balance appears in the passage naming the holders, and it should be read as summary rather than testimony. Those holders were a mixed set: Gawil and Mahur carved out of old Berar, Daulatabad to Yusuf Adil Khan, Gulbarga to the Abyssinian eunuch Dastur Dinar, the eastern coast left with Nizam-ul-Mulk Bahri, Bijapur reserved by Gawan himself.
Halving the provinces was the least of it. Governors lost their garrisons: 'one fortress alone was left in the governor's hands, in which he might, if he chose, reside ; but the remainder were entrusted to officers and troops distinctly appointed by the King' [12]. A taraf-holder was thereby converted from a warlord holding a country of forts into a revenue officer with one residence. The pay reform tightened the same screw from the other end: Gawan 'increased the pay of an officer of five hundred to one lack and twenty-five thousand boons' [13], while docking a sum equal to a man's pay for every man short of complement. Pay was raised and tied to muster in one act. Ferishta names the cost beside the praise: order prevailed and the government gained strength, and the scrutiny made the minister hated by every chief whose books it opened.
The college at Bidar and the reach of a Persian secretary
Ferishta, writing around 1610, saw Gawan's college whole: 'a college built by him at Ahmudabad Bidur two years before his death, containing also a mosque and a large square, which at the date of this history were as entire as if only just finished' [16]. UNESCO's tentative-list dossier gives the surviving fabric: 'The Madrasa of Mahmud Gawana occupies a rectangular structure covering 68 by 60 meters and encloses a central quadrangle.' [6] What it housed is remembered in the terms of a madrasa in Herat or Samarqand: free board, lodging and teaching for over 500 students at any given time, including students from outside India, and a founder's library of 3,000 volumes whose fate is unknown [4].
The college is the visible end of a career conducted mostly in prose. Gawan's two surviving works -- the Riyaz al-Insha, 148 personal and official letters, and the Manazir al-Insha -- carry a correspondence list that maps the Persianate world of his century: the poet Jami, the historian Sharaf al-Din Ali Yazdi, the Timurid sultans Abu Sa'id and Husayn Bayqara, the Ottoman Mehmed II. Encyclopaedia Iranica's verdict is compressed and unsentimental: 'He expanded the Bahmanid kingdom with his strategy and statesmanship and introduced a series of reforms in administration; unfortunately his tenure was cut short when he was put to death by Mohammad III in Safar, 886/April, 1481' [7].
April 1481: the killing, and the three accounts of it
Between the wars came a drought that decided everything after it. 'No rain fell during the next year either, and the towns in consequence became almost depopulated. Many of the inhabitants died of famine' [18], and numbers emigrated for food to Malwa and Gujarat. Ferishta makes the famine the direct cause of the eastern war, because the rebel Bhimraj urged Orissa to invade while the Deccan's resources were exhausted; the queen-mother died in the same period, removing the last authority above the minister. In March 1481 the sultan stormed Kanchi, the furthest south a Bahmani army ever reached, and farmans announcing it went out weeks before the execution.
The standard account is a forgery, told alike by Ferishta and the Burhan-i Ma'asir: two courtiers got the keeper of Gawan's seal drunk and had him stamp a folded blank, over which Nizam-ul-Mulk Bahri wrote a letter to the Raya of Orissa -- 'the Deccan may be conquered with little trouble. On the Rajmundry frontier, there is no officer of any character' [14]. Shown it, Gawan said the seal was his but the letter unknown to him, and the king ordered him killed where he stood: 'Thus died Khwaja Mahmood Gawan, in the seventy-eighth year of his age.' [15] Ferishta's most damaging detail is procedural -- the king never asked to see the messenger supposedly taken with the letter, and when he later called for him, none could be produced. That phrase about the age is also the root of a long dispute, Sherwani noting that 'Fer. says that he was 78 at the time of his death on 5.2.886/5.4.1481, so according to this calculation he must have been born in 808/1406.' [20]
Two other witnesses pull opposite ways. The earliest, al-Sakhawi, heard the news in Mecca and knows of no letter at all, only an allegation that Gawan was massing against the royal camp: 'One of the royal slaves struck his sword on the old man's shoulders and went on striking till he was dead. This happened on the 6th of Safar, 886.' [21] Against both stands the state's own paper, a farman issued on the day of the murder asserting the letter as real: 'he chose to send a letter to that most wicked of men, the Raya of Orissa, which purported to be an invitation for an alliance between them' [23]. Sherwani declines to abandon the Persians -- 'the story of the forged letter is more convincing specially as both our two chief Persian authorities agree in its essentials' [22] -- but the fullest and most dramatic version is the latest, the earliest witness knows nothing of a forgery, and the only same-day document is the killer's own justification.
The ruin of the Deccan
The collapse is usually dated to the 1490s, when the provinces declared themselves kingdoms; Ferishta's account of the following weeks argues for 1481. On the night the march was ordered, Imad-ul-Mulk and Khudawand Khan moved their divisions four miles out of camp and answered the king's enquiry by saying that since his favourite servants had destroyed such a minister, they could not but fear for their own safety at court. They returned only in company with Yusuf Adil Khan, and obtained whatsoever they required; escorting the king to Bidar they camped outside and refused to enter, and he, finding his authority unsupported, sent them back to their governments with honours. Iranica records the court poet's verdict on the reign: he 'mourned the death of Mohammad III in a versified chronogram conveyed by the key-phrase karabi-e Dakan (ruin of Deccan)' [8].
The proof is in the map. Every one of the five Deccan sultanates traces to a taraf-holder appointed or confirmed in this reign. Yusuf Adil Khan, whom Ferishta calls a follower Gawan had adopted as a son, founded the Adil Shahi line at Bijapur; Malik Hasan Bahri, chief architect of the plot against his own patron, fathered the founder of the Nizam Shahis; Fathullah Imad-ul-Mulk's Gawil became the Imad Shahi kingdom. Bidar itself passed to a household slave: the Barid Shahi dynasty 'was named after Qasem Barid, a Persianized Turk from the Caucasus who in his youth had been sold as a slave to Sultan Mohammad Bahmani III' [9]. The eightfold division meant to make governors too weak to rebel supplied, within a decade of its designer's death, the exact units of secession.
Dates in this essay
The day of Mahmud Gawan's killing is disputed. Ferishta in Briggs's translation prints 5 Safar 886, with the marginal equivalent 5 April 1481, and Sherwani records that Ferishta and the Burhan-i Ma'asir agree on that day [sherwani-birthdate]. Al-Sakhawi, the contemporary who received the news in Mecca, gives 6 Safar 886 [sherwani-sakhawi]; Sherwani flags the discrepancy rather than resolving it. Encyclopaedia Iranica declines to name a day at all, giving only 'Safar, 886/April, 1481' [iranica-bahmanid-gawan]. This essay dates the killing to April 1481 without asserting the day.
Gawan's year of birth follows from the age at death, and the two are not reconcilable. Ferishta's 'in the seventy-eighth year of his age' [ferishta-execution] yields 808/1406, the figure followed by Aziz Mirza, by Wolseley Haig in the Cambridge History of India III, and by Wajahat Husain. Sherwani rejects it for 813/1411 on the authority of al-Sakhawi, a contemporary against a chronicler writing roughly 150 years later [sherwani-birthdate]; Eaton's chapter title takes 1411 as settled [eaton-cambridge-gawan]. The essay quotes Ferishta's phrase and does not silently adopt either year.
Muhammad Shah III's own death is dated 1 Safar 887, marginally glossed 24 March 1482, in Briggs's Ferishta; Sherwani heads his final chapter 5 Safar 887 / 26 March 1482, and Wikipedia follows 26 March 1482. Iranica gives only 'Safar, 887/March, 1482' [iranica-bahmanid-chronogram]. The month is not in dispute; the day is. The reign is dated here 1463-1482, from the accession of 30 July 1463 [wiki-muhammad-shah-iii] to March 1482.
The year Goa passed to the Bahmanis floats between 1469 and 1472 because the campaign ran across several monsoon seasons. Ferishta places the fall of the port in the season after AH 875/1470 [ferishta-goa-1472]; Eaton dates the whole expedition 1469-1472 [eaton-cambridge-gawan]; English Wikipedia gives February 1472, a Goan diocesan history following P. B. Desai gives 1472, and Encyclopedia.com gives 1471. The essay dates the conquest to the campaign of 1469-1472 rather than to a year.
The end of the Bahmani state itself is dated differently by the sources used here. Iranica's Bahmanid article states that the dynasty ruled until 934/1528, while Iranica's own Barid Shahi article writes the Bahmanid span as 748-944/1347-1538 [iranica-baridshahi-qasim]; English Wikipedia gives 1347-1527. Under the archive's convention -- dating a state to the settlement that extinguished it, not to the loss of effective power -- the terminal date belongs in the 1520s, while this essay argues that effective sovereignty over the tarafs ended in 1481.
Numbering: modern usage, Encyclopaedia Iranica ('Sams-al-Din Mohammad III') and the coin catalogues all number this sultan third, but Briggs's 1829 translation of Ferishta heads the whole reign 'MAHOMED SHAH BAHMUNY II'. Any page reference to Briggs must carry that warning or it will appear to belong to a different ruler.
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
Sayyid Ali Tabataba's Burhan-i Ma'asir -- the second of Sherwani's two chief Persian authorities -- was not read directly for this essay, only through Sherwani's citations; T. S. King's English translation ran in the Indian Antiquary in 1899. It should be read before the essay relies on the detail that the sultan was intoxicated when he gave the order, since Sherwani notes that the Burhan says the king was ill and had gone inside to take a purgative.
Mahmud Gawan's own Riyaz al-Insha (ed. Chand b. Husain and G. Yazdani, Hyderabad 1948) was not opened, so no words of Gawan's are quoted here at first hand. His often-repeated description of Goa as the envy of the islands and ports of India reached this research only at third hand and must be checked against the edition before it can be printed as his.
Ghulam Yazdani's Bidar: Its History and Monuments (OUP, 1947), the standard archaeological account of the madrasa and the Ashtur tombs, was not consulted, nor was Elizabeth Schotten Merklinger's specialist study of the madrasa's plan and tilework in Kunst des Orients 11 (1976). The claim made here that the building's models are Timurid and Khurasani therefore rests on general description rather than on the architectural literature.
Was the eightfold division in fact shared equally between Deccanis and foreigners, as Wikipedia and several reference sites assert? The Ferishta passage listing the eight governors does not say so, and those sites disagree with Ferishta about which old province each new taraf was carved from. A named scholarly source is needed before the balance claim can be repeated.
Did Belgaum see the first use of gunpowder in the Deccan? Ferishta specifies trenches and mines as the untried technique and says nothing of the sort; Wikipedia's own Bahmani article separately dates the first siege use of gunpowder on the plateau to the fort of Machal in 1470 or January 1471. Klaus Rotzer's work on Deccan fortifications would settle the question and was not consulted.
Is the 'Melik-Tuchar' of Nikitin's narrative certainly Mahmud Gawan? The text gives only the title, an earlier Malik-ut-Tujjar had held it and been destroyed in the Konkan a generation before, and the identification here rests on Ferishta's confirmation that Gawan bore the title. Wielhorsky's 1857 translation should also be checked against the modern Russian-Academy edition.
What was the aggregate revenue of the Bahmani state at its greatest extent? Nothing beyond Ferishta's pay rates in hons, Briggs's 1828 valuation of the hon at three and a half rupees, Gawan's household accounts in larees, and the surviving denominations of the Muhammadabad (Bidar) mint could be established, and the numismatic material was read only on a catalogue site. Goron and Goenka's standard corpus would be the proper source.
No inscription corpus was consulted. Z. A. Desai's Topographical List of Arabic, Persian and Urdu Inscriptions of South India is the obvious check on the madrasa's foundation chronogram, still on the surviving front wall, and on any surviving epigraph naming Muhammad Shah III.
Britannica's account of the vizierate -- including its claims that an alliance with Vijayanagara was decisive against Orissa in 1470 and that Kondapalli was taken from Saluva Narasimha in 1481 -- could not be opened and is unverified here. Ferishta's own narrative of the 1471 intervention runs through the Orissan claimant Ambar Rai and mentions no Vijayanagara alliance.
The Khandesh protectorate and the terms of the Kherla settlement with Malwa, by which Kherla was conceded and all other Bahmani territory retained, appear here only through summary and chapter headings; neither was verified from a passage read in full.
Meia Walravens's thesis on Gawan's diplomatic networks (Antwerp/Ghent, 2022) and her articles on the Bahmani-Mamluk Arabic correspondence and on Esfezari's acrostic letter are the current scholarship on Gawan's foreign relations, and only abstract-level material was seen. Any fuller treatment of the Riyaz al-Insha and of Bidar's place in the Persian cosmopolis should start there.
The archive.org OCR file used for Sherwani is catalogued as Mahmud Gawan, the Great Bahmani Wazir (1942) but also contains front matter and chapters from his later The Bahmanis of the Deccan (1953). No page number from that file should be printed without checking a clean copy, and the Ferishta and Nikitin texts are likewise OCR of nineteenth-century scans whose typographic errors have been preserved in the quotations above.