One day Harihara was hunting along a bend of the Tungabhadra when he saw something that made no sense. A hare, run down by his hounds, turned around and bit the dogs.1 A prey animal attacking its predators — the king stopped, baffled. Shortly afterward he met a sage named Vidyaranya, who told him to build a city on that exact spot of strange courage, promising it would "prove the strongest in the world."2 So begins the legend of Vijayanagar, the City of Victory, and of the place beneath it the locals already knew by an older, holier name: Hampi, after the river goddess Pampa.3
The hare-bites-hound story is not unique — by legend, Kandy in Sri Lanka was founded the same way, after a prince met another dog-biting hare.4 That a brand-new dynasty would reach for a borrowed miracle to bless its chosen ground tells you exactly what kind of city this was about to become: a place that manufactured its own legitimacy out of whatever materials lay to hand, sacred river goddess and folk omen alike. Within two centuries the spot would hold perhaps 480,000 people and be reckoned second only to Beijing.5 Within two centuries and a few years more, it would be ash, rock and rubble, picked over by tigers.6 This page is about both the splendor and the erasure — and about the stones still standing that quietly demolish the story we tell about why it fell.
Vijayanagar — "City of Victory" — was the capital of the south Indian empire of the same name, founded in the mid-fourteenth century by the five sons of Sangama and ruled, at its height under Krishnadeva Raya, over some 140,000 square miles of the peninsula.7 The site sat on a bend of the Tungabhadra River, on ground already sacred to the river goddess Pampa (from whom the modern name Hampi derives) and to a male deity, Virupaksha — both folded into Devi and Shiva of the Sanskritic mainstream, both adopted as patrons by the founding Sangamas.8
As a settled fact of demography, the numbers are staggering for the era. The capital and its core housed between 100,000 and 250,000 people in the earlier period; by the end of Krishnadeva's reign, the imperial core was home to an estimated 480,000, "making it second only to Beijing in terms of population."9 This was a decentralized state — districts near the capital ruled directly by the Rayas, the wider empire negotiated through feudatories called amaranayakas under a shared-sovereignty system named nayamkara.10
But the deeper significance of Hampi, in Pillai's hands, is not its size. It is the evidence carved into its stone. On the Great Platform in the Royal Centre, among the oldest visible ruins, are reliefs of Muslims — long-nosed, in pointed caps and pointed shoes — riding horses, bearing arms, and dancing for the Vijayanagar king. The Ramachandra temple shows Arabs leading horses. A column in the celebrated Vitthala temple carries a "turbaned Muslim warrior."11 A Hindu capital, in its most ceremonial precinct, chose to depict Muslims not as demons but as cavalry, courtiers, and dancers in the king's own service. Hampi is the physical archive that anchors Pillai's whole argument about the Deccan.
The mechanism here is that buildings cannot lie the way chronicles can. A court poet can recast a war over a riverbank as a holy crusade; a relief panel carved for the king's own platform records what the court actually valued enough to immortalize. And what Vijayanagar immortalized was a cosmopolitan world.
Look at how the period names itself. The king of Orissa was Gajapati, Lord of Elephants; the Bahmani Sultan to the north was Ashvapati, Lord of Horses; and the Vijayanagar Raya was Narapati, Lord of Men — three peers in a single shared scheme, the Muslim sultan slotted in as a respected rival, not an alien.12 Literary sources of the south list the Bahmani Sultans "merely among other dynasties such as the Hindu rulers of Orissa and the old Cholas," depicting Muslims as "respected political rivals, just like other major Hindu powers."13 The Hindus never united as one to face a Muslim bloc; the northern Muslims were never a single consolidated block "baring fangs." The Bahmanis forged Hindu alliances — Telugu Reddy and Velama chiefs stood with them against fellow Hindus in Vijayanagar; the suzerain of Orissa allied with Bidar against the southern Rayas despite their shared Hindu pantheon; and on one occasion a queen in Vijayanagar appealed to a Sultan to come rescue her.14
The internal logic of Hampi as evidence runs like this: if the founding of Vijayanagar had been a "patriotic national" crusade of "the Hindus of the south" against a monolithic Islamic "alien attack," then the founders would not have had to crush "widespread rebellions in Konkan and in the Tamil country" raised by Hindu chieftains in their own early days.15 One of the founders, Harihara, may even have served the Delhi Sultanate.16 Plain realpolitik, dressed in the vocabulary of faith — that is what the carvings and the alliances and the titles all point to. The first Rayas were not shielding Hinduism so much as "taking advantage of a period of public commotion" to build a state and serve their own dynastic interests.17
What Hampi gives the rest of the vault is a hard, datable, physical exhibit for a thesis that is otherwise carried by chronicles and titles. The reliefs and the column are the courtroom evidence behind The Sultan Among Hindu Kings — the moment when Sangama brothers styled themselves Hinduraya Suratrana, "Sultan among Hindu Kings," internalizing a Persianate honorific into Hindu kingship.18 When that page argues the Deccan was cosmopolitan rather than communal, Hampi is where you go to stand inside the proof.
Hampi also emits the founding-legitimacy pattern the vault tracks across new dynasties: a house of "obscure or humble origins" assembling sacred associations after the fact — Pampa and Virupaksha, the sage Vidyaranya who "converted them back to Hinduism," marriages claimed into the defunct Hoysala and Yadava and Pandya lines, the boar emblem borrowed from the lapsed Chalukyas, temples patronized to "purchase the loyalties of local elites."19 That is a reusable model of how capitals are made to look ancient.
And Hampi emits, above all, the destruction case — the city as the thing a battle is for, and the thing that dies when the battle is lost. That hands directly to The Battle of Talikota, 1565, the engagement that turned 480,000 people into refugees and rubble.
Take the single turbaned Muslim warrior on the column of the Vitthala temple and read him slowly, because he carries more argument than any chronicle.20
First, his placement. He is not on a defensive wall or a battle frieze where an enemy might appear. He is on a temple column, in a mandapa, in the most celebrated devotional architecture of the capital. Temple sculpture is curated; it advertises what a king wants his god and his subjects to see. A Muslim warrior carved there is a Muslim warrior the Vijayanagar court wanted permanently associated with its sacred center. Set him beside the Great Platform reliefs — Muslims riding, bearing arms, dancing for the king — and a pattern hardens: these are not captives or caricatured invaders. They are participants in the spectacle of Hindu sovereignty.21
Second, what he proves about the army. Vijayanagar fought the Bahmanis for centuries over Raichur and the Krishna riverbank, yet hired Muslim cavalry, prized Persianate horsemanship, and recorded that prestige in stone. The carving is consistent with the lived fact that the wars of the age "concerned where the boundary between the states lay," not which god ruled — and that "where religious tensions did arise, it was not with Muslims" but between Jains and Vaishnavas, or Shaiva and Madhva Brahmins, disputes the Vijayanagar court itself negotiated and settled.22
Third, what he refutes. The communal reading wants Vijayanagar to be a Hindu bastion holding the line against Islam. The Vitthala warrior makes that reading awkward: you do not enshrine your civilizational enemy on your holiest pillar. Pillai's point is not that there was no "us and them" — the very title Hinduraya Suratrana shows a conception of difference.23 His point is that the difference was political and Persianate-cosmopolitan, not a war of faiths. The column is the difference made visible: a Hindu king proud enough of his Muslim cavalry to carve them into worship.
Fourth, what the column cannot tell us, which is honest to admit. We do not have the carver's instructions or the king's intent in writing; the reliefs are read, interpreted, dated by art historians, and any single panel is mute about motive. Pillai builds his case from the accumulation — Great Platform plus Ramachandra plus Vitthala plus the titles plus the alliances — not from one stone alone. The strength is the convergence.
Stand on the Great Platform in the Royal Centre. It is the third day after the news from Talikota reached the city, and the air has changed.
Below you, the reliefs are still bright in the morning light — the long-nosed Muslims in pointed caps, frozen mid-dance for a king who is no longer here. Ramaraya's brothers came back from the battlefield, and instead of manning the walls, they loaded 550 elephants with gold and diamonds, collected the puppet emperor, and rode out.24 You watched the carts of the rich go first — those with horses and elephants got out. The others trusted in prayers and stayed.25
Now you hear it: the sound of an army that is not coming to fight, because there is no one to fight. The Shahs of the Deccan march in. Orders are given for a comprehensive sack. The temples — Vitthala, Ramachandra, the one with the warrior on its column — the palaces, the houses, the statues: everything is plundered.26 Fires go up in lines across the wards. Thousands are killed where they stand.27 Ferishta will write that the plunder was so great "that every private man in the allied army became rich in gold, jewels, arms, horses and slaves," and his own father, once a tutor at the Nizam Shah's court, will record the killing of Ramaraya "with almost unconcealed delight."28
What you are standing inside is the moment Robert Sewell summed up centuries later: "Never perhaps in the history of the world had such havoc been wrought and wrought so suddenly on so splendid a city; teeming with a wealthy and industrious population in the full plentitude of prosperity one day, and on the next seized, pillaged, and reduced to ruins amid scenes of savage massacre and horrors beggaring description."29 One day the second city of the world after Beijing. The next, rock and rubble. A generation on, a traveller passing through would find nothing here but "Tygres and wild beasts."30
The way Hampi goes wrong is not in stone but in the telling. The diagnostic sign is when the ruins get drafted into a story they actively contradict.
Watch for the crusade frame. When Vijayanagar is described as the noble Hindu rampart that finally fell to Islam, the dancing Muslims on the Great Platform and the warrior on the Vitthala column have to be quietly ignored, because they do not fit. The reliefs are the warning light: a "patriotic national" reading of the founding cannot explain why the founders crushed Hindu rebellions in Konkan and Tamil country, why the Bahmanis fielded Hindu allies, why a Vijayanagar queen once begged a Sultan for rescue.31 To read the clashes of the age "as a collision of religious identities is to ignore the fragmented era in which these confrontations occurred."32
Watch, too, for the opposite failure — flattening the difference entirely. Pillai is careful here, and so should we be: there was an "us and them." The title Hinduraya Suratrana is itself the first time Indian rulers used "Hindu" to mark themselves off from Muslims of foreign origin.33 The honest reading holds both: identities were felt as dissimilar, and the wars were still fought over riverbanks and revenue, not salvation. The misremembering happens when later partisans collapse one of those two truths to win an argument the fourteenth century was not having.
The final diagnostic sign is the romance of the founding origin. There were many origin myths for the Sangama family — Kannada, Telugu, Hindu warriors, Muslim converts — and the temptation is to call this confusion or corruption. It is neither. Different stories were "composed by different authors to be communicated to very different audiences," to sell a new house to a realm full of different languages and cults.34 When a single clean origin gets asserted, that is the sign someone has thrown away the messy evidence in favor of a tidy myth.
The population figures are Pillai's strongest demographic claim and they rest on estimate, not census: 100,000–250,000 in the core earlier, "an estimated 480,000" by the end of Krishnadeva's reign, "second only to Beijing."35 These are the kind of medieval-city numbers that travellers and later historians inflated and deflated by turns; treat the 480,000 as a credible scholarly estimate carrying real uncertainty, not a counted total.
The founding itself sits on contested ground. Pillai openly flags it: the Sangama brothers' origins "are unclear," the descent from the lunar race is a later claim, the conversion-and-reconversion narrative "appears most prominently in local traditions," and even whether they were forcibly converted to Islam or willingly transferred allegiance to Muhammad bin Tughluq is unresolved — Pillai reasons that Tughluq would hardly hand the taming of the south to reluctant converts, but he is inferring.36 The hare-and-hounds omen and the sage Vidyaranya are explicitly legend, weakened further by the identical Kandy story.37
The destruction account leans heavily on Ferishta and on Sewell. Ferishta is a Persianate chronicler whose father served the victorious Nizam Shah — a source with every reason to celebrate the sack, and Pillai quotes the father's gloating to show exactly that bias.38 Sewell's "never perhaps in the history of the world" is rhetoric from a colonial-era historian, vivid and quotable and not a measurement.39 The "Tygres and wild beasts" comes from a later traveller; the precise scale of the months-long sack is, like most such events, beyond firm reconstruction. What is not in serious doubt is the outcome: the greatest city of its age was abandoned by its own rulers and then systematically destroyed.
Open question worth holding: the reliefs are interpreted, not captioned. Art historians read the long-nosed figures as Muslims and date them; the argument's force comes from convergence across the Great Platform, Ramachandra, and Vitthala plus the titles and the alliances. One panel proves little. The accumulation is the case.
Pillai is arguing against a ghost — the communal nationalist reading of the Deccan in which Vijayanagar is the Hindu fortress and the Sultanates are the Islamic flood. His whole rhetorical strategy is to let the evidence embarrass that story, and Hampi is his best witness. His own framing tension is visible: he must concede that an "us and them" genuinely existed (the Hinduraya Suratrana title is his own evidence for it) while insisting the conflicts were political-economic rather than religious. He holds the two together honestly, refusing to resolve the tension into either a fairy tale of harmony or a war of faiths.
He leans on the primary sources while reading against their grain. Ferishta supplies the lurid detail of the sack, but Pillai cites the chronicler's gloating father precisely to mark the source's bias — he uses the celebration as data about the celebrant, not as neutral reportage. Sewell gives him the unforgettable sentence about havoc, and Pillai takes the eloquence while implicitly treating it as rhetoric, not statistics. The convergence across his sources — reliefs, titles, alliance records, court poetry, Persianate chronicle — is what carries the argument; no single one would.
Set against the vault's Laine pages on Shivaji, Shivaji: Legend as Interested Knowledge, a real convergence appears. Laine shows how a later Maratha hero gets recast by interested authors into the figure each audience needs; Pillai shows the same machinery two centuries earlier, at Vijayanagar's founding, where "different authors" composed different origins "to very different audiences." Both authors treat the legend itself as evidence about its makers. The tension is one of sympathy: Pillai writes with affection for the cosmopolitan Deccan he is recovering, while Laine writes as the cooler analyst of how myth hardens. Read alongside the vault's Purandare pages, which take the Maratha vantage as largely settled fact, Pillai's insistence that founding stories are assembled is the corrective the Maratha-vantage material does not supply on its own.
Here is the plain version first: a great city is partly a propaganda machine made of stone, and when it dies you can watch how power both built belief and then erased it.
To behavioral mechanics — Maratha Administrative Governance Model. Vijayanagar ran on nayamkara — shared sovereignty, with amaranayakas (the very word amara borrowed from the Bahmani amir) holding dominions as crown subordinates, and the rich Tamil coast granted autonomy because the capital could not "harmonize the agrarian economy of the dry upland plateau with the commercialized economy" of the coast.40 The Maratha governance page describes a later south-Indian state solving the same control problem — how a center extracts from a periphery it cannot directly administer. The structural parallel is the decentralization-as-necessity trap: both states devolved power not from generosity but from inability to hold it tight, and both stored up the rebellions that devolution invites (Vijayanagar's cultivators of the Kaveri delta "rose up in widespread rebellion against the oppressive taxes"). The insight that neither page reaches alone: the borrowed Persianate vocabulary of office (amir → amara) shows that even a "Hindu" administrative skeleton was assembled from the cosmopolitan Deccan's shared toolkit — the cross-faith borrowing the carvings depict was also happening at the level of how the state was run, not just how it was decorated.
To capital-as-statecraft — Raigad: The Capital Fortress. Both pages treat a capital as an argument in stone. Raigad was built and crowned to make a sovereign; Vijayanagar's reliefs, temples, and borrowed origin myths were assembled to make a new dynasty look ancient and legitimate. The precise tension between them is survival: Raigad was a fortress chosen so the capital could endure a siege, while Vijayanagar's defenders, after Talikota, did not even man the walls — they loaded the elephants and fled, and the unfortified ceremonial heart was sacked over months.41 The insight neither reaches alone: a capital that invests everything in legitimacy-display (platforms, temples, dynastic marriages) and nothing in the brutal logic of defense can be the second city of the world one day and tigers' ground the next. Raigad answers the question Hampi failed: a capital must be survivable, not only magnificent.
To Eastern spirituality — Bhakti as Path. Vijayanagar's legitimacy ran through devotion: the site was sacred to Pampa and Virupaksha, the Sangamas adopted their worship, the Rayas "patronized richly the most important temples of the conquered provinces to win local approval."42 The bhakti page treats devotion as an interior, liberating path; Hampi shows the same devotional energy turned into an instrument of state — temples as loyalty-purchase, gods as legitimacy-assets, the river goddess as the city's title deed. The tension is between devotion as the soul's road and devotion as the crown's ledger. The insight neither reaches alone: in a polity like Vijayanagar the two are not separable — the same temple that housed genuine worship of Virupaksha also advertised the dynasty's right to rule, so that the sack of the city was simultaneously an act of plunder and an act of de-sacralization, killing the gods' patron and the people's faith-machine in one stroke.
Sharpest implication. Hampi is the rare case where the building outlives the argument the building disproves. We tell the fall of Vijayanagar as a tragedy of Hindu civilization overrun by Islam — and the surviving stones of that very civilization show Muslims dancing for the Hindu king, Arabs leading his horses, a turbaned warrior carved into his holiest column. The ruins refute the eulogy written over them. The city's erasure was real and total, but it was the work of a coalition of Deccan Sultanates avenging a political insult, not a holy war — and the stones know it even when the storytellers forget.
Generative questions.