The reform that drew the fault lines
The Bahmani state broke along the administrative map its own most capable minister had drawn. By the 1470s the kingdom had outgrown its inherited provincial structure, and Haroon Khan Sherwani is blunt about the consequence: the incongruity meant that 'the tarafdar of each province had virtually become a small potentate vrilb Ms territory sometimes ready to with¬stand the order of the central govemmcoi itself' [13]. Mahmud Gawan's answer was to split four great tarafs into eight provinces of moderate size, carving royal domain out of each governor's jurisdiction and keeping the forts under garrisons supplied from the court. The reform was meant to cage the provincial commanders; its effect, once the centre lost the will to enforce it, was to hand each of them a compact, defensible, revenue-bearing unit with a fortified capital attached.
The five states that emerge in the 1490s are very nearly Gawan's provinces redrawn as kingdoms. Encyclopaedia Iranica records the outcome without the mechanism: the dynasty 'finally disintegrating into the five independent Muslim kingdoms of the ʿEmādšāhīs of Berār, the Neẓāmšāhīs of Ahmadnagar, the Barīdšāhīs (q.v.) of Bīdar, theʿĀdelšāhīs of Bījāpūr, and the Qoṭbšāhīs of Golconda' [2]. What emerged was not five homogeneous polities but five composite ones: as Roy Fischel puts it, 'the Deccan sultanates incorporated varied Muslim (Foreigners, Deccanis) and non-Muslim (Brahmins, nayakas, desais, Marathas) elites' [22]. The founders bear that out: a Caucasian slave, the son of a convert from the Vijayanagara country, a captured Kannada Brahmin boy, a Qara Qoyunlu Turkoman from Hamadan.
1490: what the governors actually did
The standard date for the first secession is 1490 and the standard authority is Ferishta, writing at Bijapur a century later; his account describes a liturgical act, not a constitutional one. Malik Ahmad Bahri, 'who never forgave the King for the death of his father', refused to attend the royal standard, 'declared his independence at once, founded the city of Ahmudnuggur', and sent envoys to Yusuf Adil Khan and Fathullah Imad-ul-Mulk urging them 'to follow his example and read the Khootba in their own names' [3]. Sovereignty here is tested by khutba and sikka — whose name is spoken in the Friday sermon, whose name is struck on the coin — and the passage describes a negotiation among three men, not a single rupture.
Sherwani rejects the whole picture: 'every one of the fiefholdcrs had great respect for the person of the sovereign and not one of them really declared their independence / in spite of what the later panygyrists ofthdr dynastLcs lilm Burhan-i Ma’astr and Ferishia might have to say' [11]. On that reading 1490 is a back-projection by court historians whose patrons needed a founding secession.
The titles favour Sherwani. At Bijapur, Ibrahim Adil Shah I 'formally claimed the Bijapur sultanate's independence from the Bahmanis in 1538, although it had been in practice independent since 1490' — his father and grandfather having used the granted title Adil Khan instead [26]. At Golconda, Sherwani concludes that 'Qutbu'lmulk never proclaimed his kingship, and in all probability the first ruler of Golkonda to do so was Ibrāhīm Qutb Shāh' — 1550 [25]. Ferishta dates Golconda's break to 918/1512: Sultan Quli 'did not declare his independence till the year 918, when the imbecility of Mahmood Shah Bahmuny s government was apparent to the whole world' [8]. De facto autonomy from about 1490; royal style only between 1538 and 1550. That forty-year gap is the real subject of the period.
Qasim Barid: an office, not a crown
The man who ended up holding the capital never took the throne. Richard Eaton is precise about what he did take: in 897/1491-92 'Qāsem acquired de facto control of the capital city by virtue of his elevation to the office of Bahmanid prime minister (amīr-e jomla). Although he never styled himself king, Qāsem did assume the kingdom’s regalia, and permitted only nominal sovereignty to continue in the hands of weak Bahmanid successors' [1]. Yazdani gives the bargain behind it: 'the king made terms with him, and in a.d. 1492 confirmed him as Amlr-i-Jumla, or Prime Minister. Thereafter the Baihmanls ceased in fact to be a ruling dynasty' [14]. Power at Bidar passed by appointment — which is why the founding date of the Barid state is itself contested, one Indian government teaching module placing it as late as 1531 [21].
Ferishta claims more for Qasim than the evidence will carry. His Bidar chapter has him usurp three forts and, 'leaving to his king only the town and fort of Ahmudabad Bidur, read the public prayers, and coined money, in his own name' [7] — khutba and sikka together, a claim of complete sovereignty. The material record refuses it: the fullest study of Deccan sultanate coinage reports that 'No coin of Barid Shahi is reported before 993' [18] — AH 993 is 1585 CE, some ninety years after Qasim supposedly began minting. A boast of coinage is the sort of claim objects can test, and here they falsify it.
The puppet system was a policy, not an accident
Four Bahmani princes reigned after Mahmud Shah II, and the usual summary — that they 'were mere puppets in the hands of Amir Barid' [23] — misses the reasoning, which Ferishta supplies verbatim: 'Ameer Bereed had but a small territory in his possession, and dreading, should he assume open independence, that the surrounding powers would attack him, placed the Prince Ahmud, son of Mahmood Shah, upon the throne, leaving him the palace, with the use of the royal jewels and a daily allowance of money for his support' [4]. Bidar, smallest of the five, used the Bahmani name as insurance against four larger neighbours. The Burhan-i Ma'asir describes the accumulation from inside the palace: 'Day hy day the power of Malik Barid in the affairs of State became greater and greater, till be brought into his ona hands the wbolo- of the government and the control of the army and the subjects' [10].
Puppets who resisted were removed. Ala-ud-Din II hid armed men near his apartments and invited the minister to the new-moon audience; one of them sneezed, Amir Barid turned back, and the king was deposed and shortly afterwards murdered. Wali-Allah was poisoned when the minister conceived a passion for his wife, a daughter of Yusuf Adil Khan, whom he then married. Kalim-Allah, the last, made the only independent move of the sequence: a petition smuggled to Babur in 1526 offering to cede Daulatabad and Berar — provinces he did not hold — in return for rescue. Babur ignored it; the leak brought closer confinement, then flight to Bijapur and Ahmadnagar. Ferishta closes the story there: 'Kulleem Oolla Shah, however, resided at Ahmudnuggur till his death, and with him ended the dynasty of Bahmuny' [5].
The evidence base, and what it admits about itself
Almost everything the modern literature says about Bidar, Golconda and Berar in this period passes through one chronicler — who warns his readers not to trust him on exactly those three houses. Ferishta prints a disclaimer at the close of his Barid Shahi section: 'the author has failed in procuring any written accounts of the Kootb Shahy, Imad Shahy, or Bereed Shahy dynasties ; he has therefore been compelled to fill up the page with oral traditions, from the most respectable and well informed people of his own times' [6]. For three of the five successor dynasties, then, the base text is transcribed hearsay gathered a century after the events — the single most important fact about the evidence for 1490-1527.
The instability shows. Briggs's Ferishta dates the death of Mahmud Shah Bahmani II — the last sultan to hold real power — twice, irreconcilably. The Bahmani volume gives 4 Zilhijja AH 924, 21 October 1518; the Qutb Shahi volume of the same translation states that 'MahZeehuj 24. mood Shah Bahmuny died on the 24th a. h. 912. of Zeehuj, A. H. 912, in the forty- May 12. seventh year of his age, and in the thirty-seventh a. d. 1507. of his reign' [9] — eleven years earlier, and immediately before the passage listing the governors' secessions. The AH 912 reading is almost certainly a corruption or a displaced passage, but it stands printed in the most widely cited English Ferishta and is flagged rather than silently harmonised.
When did the Bahmani sultanate end?
At least six answers are defensible: 1518 (Mahmud Shah II's death, the textbook terminus), 1527 (Ferishta, with Kalim-Allah at Ahmadnagar), 1528 (Iranica's flat statement that 'The dynasty ruled until 934/1528' [2]), 1538 (Sherwani), 1545 (coins in Kalim-Allah's name). And the numismatists write past all of it: Wagoner and Tandon speak of 'the final collapse of the Bahmani kingdom at the end of the fifteenth century' [17], Fischel of disintegration 'in the 1480s' [22].
Sherwani's case for 1538 is built from objects rather than chronicles. Kalim-Allah supposedly left Bidar in 934/1527-8, yet coins in his name survive from 942 and 943, which could be dismissed as posthumous. Two Bijapur inscriptions block that escape, one dated 943/1537, both still styling Bijapur's ruler by the Bahmani-granted title Majlis-i Rafi' Adil Khan. Sherwani concludes that these 'go to prove almost conclusively that the person whom Ibrahim regarded as his liege lord was living in &+3/i537 but must have died in or Iwfore 945, i. c., some time about 944/1538' [12]. A sultan is dead, on this test, when his successors stop taking their titles from him.
Even 1538 may be too early for the name. Yazdani reported that 'Recently some coins bearing his name and the date 952 h. (a.d. 1545) have been found by the Archaeological Department, Hyderabad. They w'ere apparently issued by the Barid Shahi king 'Ali Barid (a.d. 1542-79)' [15] — coins which, on the modern summary, suggest the last Bahmani outlived Amir Barid [24]. Long after the last sultan fled his capital, the kings of Bidar still found it useful to put his name on their copper. This archive dates a state from the assumption of sovereign power in its own capital to the settlement that extinguished it, placing the terminus between 1527 and 1538; the rival dates are recorded below rather than resolved.
What the coins and the tombs record
The monetary evidence is good for the century before the break-up and poor for the break-up itself. Wagoner and Tandon's study of the Akola hoard shows the small change gone by the mid-fifteenth century: 'only the four largest denominations—6-, 9-, 12-, and 18-māṣas... were regularly minted in quantity, and the smaller denominations were effectively discontinued' [17]. Then the record thins with the state: 'after Mahmud Shah no coins either in gold or silver had been identified. No coins have been assigned to his immediate successors who were merely puppet in the hands of powerful nobles' [19]. Precious-metal issue stops with the last sultan who governed; what continues is copper struck by others in a dead man's name.
The buildings say the same in stone. Yazdani notes that Mahmud Shah's mausoleum at Ashtur is 'a monument of considerable dimensions', erected by the king in his lifetime, while 'The tombs of Ms successors, who were mere puppets in the hands of their BaridI ministers, are insignificant structures and possess no arcliitectural interest' [16]. Tomb scale is a fair index of who commanded revenue. Yet the necropolis kept growing past every proposed end-date: India's UNESCO tentative-list dossier records that 'The 12 tombs of Bahmani rulers in Ashtur were erected from 1436 to 1535' [20]. The Barid Shahis built their own garden necropolis two kilometres from the fort, but their real architectural investment — Ali Barid's domed tomb of 1576, the tiled Rangin Mahal — came long after the dynasty they had displaced was gone [1].
Dates in this essay
End of the Bahmani sultanate: 1518 (death of Mahmud Shah II; the conventional textbook terminus, followed by the e-PG Pathshala module's dating of the five kingdoms 'from about 1518'); 1527 (Ferishta, with whose Kalim-Allah 'ended the dynasty of Bahmuny', followed by Wikipedia); 1528 (Encyclopaedia Iranica, 'The dynasty ruled until 934/1528'); 1538 (Haroon Khan Sherwani, from coins of 942-943 AH and Bijapur inscriptions still using the Bahmani-granted title Majlis-i Rafi' in 943/1537); 1545 (Yazdani, from coins in Kalim-Allah's name dated 952 H found by the Hyderabad Archaeological Department); and 'the end of the fifteenth century' or 'the 1480s' in the numismatic and survey literature (Wagoner and Tandon; Fischel). This archive's convention places the terminus between 1527 and 1538 and records the rest.
Ahmadnagar's independence: 28 May 1490, the Victory of the Garden, in Ferishta and the modern consensus; 901/1496 in Encyclopaedia Iranica's Ahmadnagar entry, following the Burhan-i Ma'asir, which also gives Malik Ahmad's reign as 901-16/1496-1510. The city's foundation is separately dated 1494 or c. 900/1495.
Golconda's independence: 918/1512 in Ferishta's Qutb Shahi section; 7 December 1518 in the standard modern accounts; and never under Sultan Quli at all on Sherwani's reading, which holds that the first Golconda ruler to proclaim kingship was Ibrahim Qutb Shah (r. 1550-80). This archive follows Sherwani on the royal title.
Foundation of the Barid Shahi state: 1489 (Wikipedia's regnal dates for Qasim Barid I and the e-PG module's list); 897/1491-92 (Eaton in Encyclopaedia Iranica, and the title of that article); 1492 (Yazdani, dating the operative event to Qasim's confirmation as Amir-i Jumla, and Wikipedia's Bidar Sultanate infobox); 1527 (Kalim-Allah's flight); and 1531 (the e-PG Pathshala module). Ferishta says only that the Barid Shahi historians dated the dynasty from the reconciliation after Dilawar Khan's death, which he does not date.
Death of Mahmud Shah Bahmani II: 4 Zilhijja AH 924 / 21 October 1518, reign 37 years 20 days (Briggs's Ferishta, Bahmani volume); 24 Zilhijja AH 912 / 12 May 1507, aged 47, in the 37th year of his reign (Briggs's Ferishta, Qutb Shahi volume); 24 Zi'l-hijja AH 924 / 20 December 1518, reign 37 years 2 months (Tabataba, trans. King); Zu'l-hijja 924 / December 1518 (Encyclopaedia Iranica); 27 December 1518 (Wikipedia). The AH 912 reading is probably a corruption but stands printed in the most cited English translation.
Death of Amir Barid I: 1542 at Daulatabad (Yazdani, Wikipedia, the e-PG module and the MANUU coinage thesis, which dates Ali Barid Shah AH 949-987/1542-1579); AH 945 / 1549 with a reign of forty-five years in Briggs's Ferishta, in a section for which Ferishta admits he had no written sources.
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
All quotations here from archive.org full texts (Briggs's Ferishta vols II and III, King's translation of the Burhan-i Ma'asir, Sherwani 1954, Yazdani 1947) are reproduced character-for-character from scanned OCR that contains recognisable errors, such as Sherwani's "&+3/i537" for "943/1537" and "panygyrists ofthdr dynastLcs lilm" for "panegyrists of their dynasties like". Would the printed pages, re-checked against the physical editions, alter any reading used here?
Does Ferishta's original Persian (Gulshan-i Ibrahimi, e.g. the Nawal Kishore Lucknow edition of 1864-65) contain the AH 912 / AH 924 contradiction for Mahmud Shah II's death, or is it an artefact of Briggs's 1829 translation or its typesetting?
What does the fuller Burhan-i Ma'asir say about the rise of Qasim and Amir Barid? J. S. King's 1900 translation covers only the Bahmani portion and stops at 1518; Tabataba's treatment of the Barid Shahis in the Delhi 1936 edition was not opened.
What are the precise Epigraphia Indo-Moslemica references for the four Bijapur inscriptions on which Sherwani's 1538 argument rests, and where were the 942 and 943 AH coins of Kalim-Allah published? The footnote apparatus was not traced, and the Rangin Mahal inscription of Amir Barid I was not read.
What happened to the money supply between 1490 and 1527? The Akola hoard closes about 1485-93, so there is no comparable numismatic evidence for the fragmentation years themselves; Wagoner's 'Money use in the Deccan, c. 1350-1687' (IESHR 51/4, 2014) would supply it and was not opened.
What do European witnesses of the period say? Afanasy Nikitin, who travelled the Bahmani domains, and Duarte Barbosa on the Deccan about 1510-1518 were not read, nor was any Portuguese account of Yusuf Adil Khan, the 'Hidalcao', and the loss of Goa in 1510, which falls inside this period.
Is Krishnadevaraya's title Yavana-rajya-sthapana-charya, 'establisher of the Yavana kingdom', attested in a contemporary Vijayanagara inscription, or is it a later attribution? The claim was found only in a secondary summary sourced to Y. Gopala Reddy's A Comprehensive History of Andhra (1990), and the underlying inscription was not seen.
Is Ferishta's report that Qasim Barid promised the Vijayanagara ray the forts of Mudkal and Raichur in return for invading Bijapur corroborated anywhere else? No other source opened here mentions it.
What does the fullest modern institutional treatment say? Sherwani and Joshi's History of Medieval Deccan (Government of Andhra Pradesh, 1973-74), vol. I on the Barid Shahis and vol. II on the architecture of the Bahmani successor states, exists on archive.org but was not downloaded; Helen Philon's 'Barid Shahis' in the third edition of the Encyclopaedia of Islam is paywalled and was not opened.
What did the Barids, as against the Bahmanis, actually build or repair at Bidar fort during the fragmentation itself? Descriptions of the fortifications and the moat-flooding sluice system were not checked for phasing between 1490 and 1530.
Does any Dakhni text survive from the Bidar court of 1490-1527? The e-PG Pathshala module asserts that Dakhni continued to develop under the Barid Shahis until 1619 but names no work, and no Barid-period Dakhni manuscript was identified here.
What became of Kalim-Allah's son Ilham-Allah, who is said to have left for Mecca in disguise and never returned? No primary attestation of the story was located.