HISTORICAL ANALYSIS

Temple desecration and Indo-Muslim states

Why, after the rise of pre-modern Indo-Muslim states, were some Hindu temples desecrated, some protected, and others constructed anew?

RICHARD M. EATON

PART II

IN the first part of this essay (published in the previous issue) it was argued that from 1192 on, the annexation of territory by Indo-Muslim states was often accompanied by the desecration of temples and/or images associated with defeated monarchs, a tr adition well-established in Indian political practice from at least the sixth century. This being so, one might legitimately ask: what happened once the land and the subjects of former enemies were integrated into an Indo-Muslim state?

TEMPLE PROTECTION AND STATE MAINTENANCE

On this point, the data are quite clear: pragmatism as well as time-honoured traditions of both Islamic and Indian statecraft dictated that temples lying within such states be left unmolested. We learn from a Sanskrit inscription, for example, that in 13 26, thirteen years after he annexed the northern Deccan to the Tughluq empire, Sultan Muhammad bin Tughluq appointed Muslim officials to repair a Siva temple in Kalyana (in Bidar district), thereby facilitating the resumption of normal worship that had b een disrupted by local disturbances.¹ According to that sultan's interpretation of Islamic law, anybody who paid the poll-tax (jizya) could build temples in territories ruled by Muslims.²

Figure 8: Govind Deva Temple at Brindavan, built by Raja Man Singh, 1590. Courtesy of Catherine B. Asher.

Such views continued to hold sway until modern times. Within several decades of Muhammad bin Tughluq's death, Sultan Shihab al-Din (1355-73) of Kashmir rebuked his Brahmin minister for having suggested melting down Hindu and Buddhist images in his kingdo m as a means of obtaining quick cash. In elaborating his ideas on royal patronage of religion, the sultan referred to the deeds of figures drawn from classical Hindu mythology. "Some (kings)," he said,

have obtained renown by setting up images of gods, others by worshipping them, some by duly maintaining them. And some, by demolishing them! How great is the enormity of such a deed! Sagara became famous by creating the sea and the rivers.... Bhagiratha obtained fame by bringing down the Ganga. Jealous of Indra's fame, Dushyanata acquired renown by conquering the world; and Rama by killing Ravana when the latter had purloined Sita. King Shahvadina [Shihab al-Din], it will be said, plundered the image of a god; and this fact, dreadful as Yama [death], will make the men in future tremble.³

About a century later, Muslim jurists advised the future Sikandar Lodi of Delhi (reign: 1489-1517) that "it is not lawful to lay waste ancient idol temples, and it does not rest with you to prohibit ablution in a reservoir which has been customary from a ncient times."⁴

The pattern of post-conquest temple protection, and even patronage, is especially clear when we come to the imperial Mughals, whose views on the subject are captured in official pronouncements on Sultan Mahmud of Ghazni, one of the most controversial fig ures in Indian history. It is well known that in the early eleventh century, before the establishment of Indo-Muslim rule in north India, the Ghaznavid sultan had made numerous, and very destructive, attacks on the region. Starting with the writings of h is own contemporary and court poet, Firdausi (death 1020), Mahmud's career soon became legend, as generations of Persian poets lionised Mahmud as a paragon of Islamic kingly virtue, celebrating his infamous attacks on Indian temples as models for what ot her pious sultans should do.⁵

Figure 9: Interior of the Govind Deva Temple at Brindavan, built under Mughal patronage by Raja Man Singh, 1590. Courtesy of Catherine B. Asher.

But the Ghaznavid sultan never undertook the responsibility of actually governing any part of the subcontinent whose temples he wantonly plundered. Herein lies the principal difference between the careers of Mahmud and Abu'l-fazl, Akbar's chief minister and the principal architect of Mughal imperial ideology. Reflecting the sober values that normally accompany the practice of governing large, multi-ethnic states, Abu'l-fazl attributed Mahmud's excesses to fanatical bigots who, having incorrectly represe nted India as "a country of unbelievers at war with Islam," incited the sultan's unsuspecting nature, which led to "the wreck of honour and the shedding of blood and the plunder of the virtuous."⁶

Indeed, from Akbar's time (reign: 1556-1605) forward, Mughal rulers treated temples lying within their sovereign domain as state property; accordingly, they undertook to protect both the physical structures and their Brahmin functionaries. At the same ti me, by appropriating Hindu religious institutions to serve imperial ends - a process involving complex overlappings of political and religious codes of power - the Mughals became deeply implicated in institutionalised Indian religions, in dramatic contra st to their British successors, who professed a hands-off policy in this respect. Thus we find Akbar allowing high-ranking Rajput officers in his service to build their own monumental temples in the provinces to which they were posted, as in the case of the Govind Deva Temple in Brindavan, patronised by Raja Man Singh.⁷ (See Figures 8 and 9.)

Akbar's successors went further. Between 1590 and 1735, Mughal officials repeatedly oversaw, and on occasion even initiated, the renewal of Orissa's regional cult, that of Jagannath in Puri. By sitting on a canopied chariot while accompanying the cult's annual car festival, Shah Jahan's officials ritually demonstrated that it was the Mughal emperor, operating through his appointed officers (mansabdar), who was the temple's - and hence the god's - ultimate lord and protector.⁸ Such actions in effect projected a hierarchy of hybridised political and religious power that descended downward from the Mughal emperor to his mansabdar, from the mansabdar to the god Jagannath and his temple, from Jagannath to the sub-imp erial king who patronised the god, and from the king to his subjects. For the Mughals, politics within their sovereign domains never meant annihilating prior authority, but appropriating it within a hierarchy of power that flowed from the Peacock Throne to the mass of commoners below.

Such ideas continued in force into the reign of Aurangzeb (1658-1707), whose orders to local officials in Banaras in 1659 clearly indicate that Brahmin temple functionaries there, together with the temples at which they officiated, merited state protecti on. "In these days," he wrote in February of that year,

information has reached our court that several people have, out of spite and rancour, harassed the Hindu residents of Banaras and nearby places, including a group of Brahmins who are in charge of ancient temples there. These people want to remove those B rahmins from their charge of temple-keeping, which has caused them considerable distress. Therefore, upon receiving this order, you must see that nobody unlawfully disturbs the Brahmins or other Hindus of that region, so that they might remain in their t raditional place and pray for the continuance of the Empire.⁹

Figure 10: The Babri Masjid in Ayodhya: as it stood prior to its demolition on December 6, 1992. Photo: Subir Roy.

By way of justifying this order, the emperor noted: "According to the Holy Law (shari`at) and the exalted creed, it has been established that ancient temples should not be torn down." On this point, Aurangzeb aligned himself with the theory and th e practice of Indo-Muslim ruling precedent. But then he added, "nor should new temples be built" - a view that broke decisively from Akbar's policy of permitting his Rajput officers to build their own temple complexes in Mughal territory. Although this o rder appears to have applied only to Banaras - many new temples were built elsewhere in Mughal India during Aurangzeb's reign¹⁰ - one might wonder what prompted the emperor's anxiety in this matter.

TEMPLE DESECRATION AND STATE MAINTENANCE

It seems certain that rulers were well aware of the highly charged political and religious relationship between a royal Hindu patron and his client-temple. Hence, even when former rulers or their descendants had been comfortably assimilated into an Indo- Muslim state's ruling class, there always remained the possibility, and hence the occasional suspicion, that a temple's latent political significance might be activated and serve as a power-base to further its patron's political aspirations. Such conside rations might explain why it was that when a subordinate non-Muslim officer in an Indo-Muslim state showed signs of disloyalty - and especially if he engaged in open rebellion - the state often desecrated the temple(s) most clearly identified with that o fficer. After all, if temples lying within its domain were understood as state property, and if a government officer who was also a temple's patron demonstrated disloyalty to the state, from a juridical standpoint ruling authorities felt justified in tre ating that temple as an extension of the officer, and hence liable for punishment.

Thus in 1478, when a Bahmani garrison in eastern Andhra mutinied, murdered its governor, and entrusted the fort to Bhimraj Oriyya, who until that point had been a loyal Bahmani client, the sultan personally marched to the site and, after a six-month sieg e, stormed the fort, destroyed its temple, and built a mosque on the site (no. 40; Map 2). A similar thing occurred in 1659, when Shivaji Bhonsle, the son of a loyal and distinguished officer serving the `Adil Shahi sultans of Bijapur, seized a governmen t port on the northern Konkan coast, thereby disrupting the flow of external trade to and from the capital. Responding to what it considered an act of treason, the government deputed a high-ranking officer, Afzal Khan, to punish the Maratha rebel. Before marching to confront Shivaji himself, however, the Bijapur general first proceeded to Tuljapur and desecrated a temple dedicated to the goddess Bhavani, to which Shivaji and his family had been personally devoted (no. 63; Map 3).

MAP 1

We find the same pattern with the Mughals. In 1613 while at Pushkar, near Ajmer, Jahangir ordered the desecration of an image of Varaha that had been housed in a temple belonging to an uncle of Rana Amar of Mewar, the emperor's arch enemy (See Table and Map 3: no. 56). In 1635 his son and successor, Shah Jahan, destroyed the great temple at Orchha, which had been patronised by the father of Raja Jajhar Singh, a high-ranking Mughal officer who was at that time in open rebellion against the emperor (no. 5 8). In 1669, there arose a rebellion in Banaras among landholders, some of whom were suspected of having helped Shivaji, who was Aurangzeb's bitter enemy, escape from imperial detention. It was also believed that Shivaji's escape had been initially facil itated by Jai Singh, the great grandson of Raja Man Singh, who almost certainly built Banaras' great Vishvanath temple. It was against this background that the emperor ordered the destruction of that temple in September, 1669 (no. 69).¹¹

About the same time, serious Jat rebellions broke out in the area around Mathura, in which the patron of that city's congregational mosque had been killed. So in early 1670, soon after the ring-leader of these rebellions had been captured near Mathura, A urangzeb ordered the destruction of the city's Keshava Deva temple and built an Islamic structure (`idgah) on its site (no. 70).¹² Nine years later, the emperor ordered the destruction of several prominent temples in Rajasthan that had become associated with imperial enemies. These included temples in Khandela patronised by refractory chieftains there; temples in Jodhpur patronised by a former supporter of Dara Shikoh, the emperor's brother and arch rival for the Mughal throne; and the royal temples in Udaipur and Chitor patronised by Rana Raj Singh after it was learned that that Rajput chieftain had withdrawn his loyalty to the Mughal state (nos. 71-74).

Considerable misunderstanding has arisen from a passage in the Ma'athir-i `Alamgiri concerning an order on the status of Hindu temples that Aurangzeb issued in April 1669, just months before his destruction of the Banaras and Mathura temples. The passage has been construed to mean that the emperor ordered the destruction not only of the Vishvanath temple at Banaras and the Keshava Deva temple at Mathura, but of all temples in the empire.¹³ The passage reads as follows:

MAP 2

Orders respecting Islamic affairs were issued to the governors of all the provinces that the schools and places of worship of the irreligious be subject to demolition and that with the utmost urgency the manner of teaching and the public practices of the sects of these misbelievers be suppressed.¹⁴

The order did not state that schools or places of worship be demolished, but rather that they be subject to demolition, implying that local authorities were required to make investigations before taking action.

More importantly, the sentence immediately preceding this passage provides the context in which we may find the order's overall intent. On April 8, 1669, Aurangzeb's court received reports that in Thatta, Multan, and especially in Banaras, Brahmins in "e stablished schools" had been engaged in teaching "false books" and that both Hindu and Muslim "admirers and students" had been travelling over great distances to study the "ominous sciences" taught by this "deviant group." ¹⁵ We do not know what sort of teaching or "false books" were involved here, or why both Muslims and Hindus were attracted to them, though these are intriguing questions. What is clear is that the court was primarily concerned, indeed exclusiv ely concerned, with curbing the influence of a certain "mode" or "manner" of teaching within the imperial domain. Far from being, then, a general order for the destruction of all temples in the empire, the order was in response to specific reports of an educational nature and was targeted at investigating those institutions where a certain kind of teaching had been taking place.

MAP 3

In sum, apart from his prohibition on building new temples in Banaras, Aurangzeb's policies respecting temples within imperial domains generally followed those of his predecessors. Viewing temples within their domains as state property, Aurangzeb and Ind o- Muslim rulers in general punished disloyal Hindu officers in their service by desecrating temples with which they were associated.

How, one might then ask, did they punish disloyal Muslim officers? Since officers in all Indo-Muslim states belonged to hierarchically ranked service cadres, infractions short of rebellion normally resulted in demotions in rank, while serious crimes like treason were generally punished by execution, regardless of the perpetrator's religious affiliation. No evidence, however, suggests that ruling authorities attacked public monuments like mosques or Sufi shrines that had been patronised by disloyal or re bellious officers. Nor were such monuments desecrated when one Indo-Muslim kingdom conquered another and annexed its territories.

On the contrary, new rulers were quick to honour and support the shrines of those Chishti shaikhs that had been patronised by defeated enemies. For example, Babur, upon seizing Delhi from the last of the city's ruling sultans in 1526, lost no time in pat ronising the city's principal Chishti tomb-shrines. ¹⁶ The pattern was repeated as the Mughals expanded into provinces formerly governed by Indo-Muslim rulers. Upon conquering Bengal in 1574, Mughal administrators showered their most lavish patronage on the two Chishti shrines in Pandua - those of Shaikh `Ala al-Haq (d. 1398) and Shaikh Nur Qutb-i `Alam (d. 1459) - that had been the principal objects of state patronage by the previous dynasty of Bengal sultans.¹⁷ And when he extended Mughal dominion over the defeated Muslim states o f the Deccan, the dour Aurangzeb, notwithstanding his reputation for eschewing the culture of saint-cults, made sizable contributions to those Chishti shrines in Khuldabad and Gulbarga that had helped legitimise earlier Muslim dynasties there.¹⁸

TEMPLES AND MOSQUES CONTRASTED

Evidence presented in the foregoing discussion suggests that mosques or shrines carried very different political meanings than did royal temples in independent Hindu states, or temples patronised by Hindu officers serving in Indo-Muslim states. For Indo- Muslim rulers, building mosques was considered an act of royal piety, even a duty. But all the actors, rulers and the ruled alike, seem to have recognised that the deity worshipped in mosques or shrines had no personal connection with a Muslim monarch. N or were such monuments thought of as underpinning the authority of an Indo-Muslim king, or as projecting a claim of sovereign authority over the particular territory in which they were situated. One can hardly imagine the central focus of a mosque's ritu al activity, the prayer niche (mihrab), being taken out of the structure and paraded around a Muslim capital by way of displaying Allah's co-sovereignty over an Indo-Muslim ruler's kingdom, in the manner that the ritual focus of a royal temple, th e image of the state-deity, was paraded around pre-modern Hindu capitals in elaborate "car" festivals.

Instances of Temple Desecration

This point is well illustrated in a reported dispute between the Emperor Aurangzeb and a Sufi named Shaikh Muhammadi (d. 1696). As a consequence of this dispute, in which the shaikh refused to renounce views that the emperor considered theologically devi ant, Shaikh Muhammadi was ordered to leave the imperial domain. When the Sufi instead took refuge in a local mosque, Aurangzeb claimed that this would not do, since the mosque was also within imperial territory. But the shaikh only remarked on the empero r's arrogance, noting that a mosque was the house of God and therefore only His property. The standoff ended with the shaikh's imprisonment in the Aurangabad fort - property that was unambiguously imperial.¹⁹

This incident suggests that mosques in Mughal India, though religiously potent, were considered detached from both sovereign terrain and dynastic authority, and hence politically inactive. As such, their desecration would have had no relevance to the bus iness of disestablishing a regime that had patronised them. Not surprisingly, then, when Hindu rulers established their authority over the territories of defeated Muslim rulers, they did not as a rule desecrate mosques or shrines, as, for example, when S hivaji established a Maratha kingdom on the ashes of Bijapur's former dominions in Maharashtra, or when Vijayanagara annexed the former territories of the Bahmanis or their successors.²⁰ In fact, the rajas of Vijayanagara, as is well known, bu ilt their own mosques, evidently to accommodate the sizable number of Muslims employed in their armed forces.

By contrast, monumental royal temple complexes of the early medieval period were considered politically active, inasmuch as the state-deities they housed were understood as expressing the shared sovereignty of king and deity over a particular dyna stic realm. Therefore, when Indo-Muslim commanders or rulers looted the consecrated images of defeated opponents and carried them off to their own capitals as war trophies, they were in a sense conforming to customary rules of Indian politics. Similarly, when they destroyed a royal temple or converted it into a mosque, the ruling authorities were building on a political logic that, they knew, placed supreme political significance on such temples. That same significance, in turn, rendered temples just as deserving of peace-time protection as it rendered them vulnerable in times of conflict.

TEMPLE DESECRATION AND THE RHETORIC OF STATE BUILDING

Much misunderstanding over the place of temple desecration in Indian history results from a failure to distinguish the rhetoric from the practice of Indo-Muslim state-formation. Whereas the former tends to be normative, conservative, and rigidly ideologi cal, the latter tends to be pragmatic, eclectic, and non-ideological. Rhetorically, we know, temple desecration figured in Indo-Muslim chronicles as a necessary and even meritorious constituent of state-formation. In 1350, for example, the poet-chronicle r `Isami gave the following advice to his royal patron, `Ala al-Din Hasan Bahman Shah, the founder of the Bahmani kingdom in the Deccan:

If you and I, O man of intellect, have a holding in this country and are in a position to replace the idol-houses by mosques and sometimes forcibly to break the Brahminic thread and enslave woman and children - all this is due to the glory of Mahmud [of Ghazni].... The achievements that you make to-day will also become a story tomorrow.²¹

But the new sultan appears to have been more concerned with political stability than with the glorious legacy his court-poet wished him to pursue. There is no evidence that Bahman Shah violated any temples. After all, by carving out territory from lands formerly lying within the Delhi Sultanate, the founder of the Bahmani state had inherited a domain void of independent Hindu kings and hence void, too, of royal temples that might have posed a political threat to his fledgling kingdom.

Unlike temple desecration or the patronage of Chishti shaikhs, both of which figured in the contemporary rhetoric on Indo-Muslim state-building, a third activity, the use of explicitly Indian political rituals, found no place whatsoever in that rhetoric. Here we may consider the way Indo-Muslim rulers used the rich political symbolism of the Ganga River, whose mythic associations with imperial kingship had been well established since Mauryan times (321-181 B.C.). Each in its own way, the mightiest imper ial formations of the early medieval peninsula - the Chalukyas, the Rashtrakutas, and the Cholas - claimed to have "brought" the Ganga River down to their southern capitals, seeking thereby to legitimise their claims to imperial sovereignty. Although the Chalukyas and the Rashtrakutas did this symbolically, probably through their insignia, the Cholas literally transported pots of Ganga water to their southern capital.²²

And, we are told, so did Muhammad bin Tughluq in the years after 1327, when that sultan established Daulatabad, in Maharashtra, as the new co-capital of the Delhi Sultanate's vast, all-India empire.²³ In having Ganga water carried a distance o f forty days' journey from North India "for his own personal use," the sultan was conforming to an authentically Indian imperial ritual. Several centuries later, the Muslim sultans of Bengal, on the occasion of their own coronation ceremonies, would wash themselves with holy water that had been brought to their capital from the ancient holy site of Ganga Sagar, located where the Ganga River emptied into the Bay of Bengal.²⁴

No Indo-Muslim chronicle or contemporary inscription associates the use of Ganga water with the establishment or maintenance of Indo-Muslim states. We hear this only from foreign visitors: an Arab traveller in the case of Muhammad bin Tughluq, a Portugue se friar in the case of the sultans of Bengal. Similarly, the image of a Mughal official seated in a canopied chariot and presiding over the Jagannath car festival comes to us not from Mughal chronicles but from an English traveller who happened to be in Puri in 1633.²⁵

Such disjunctures between the rhetoric and the practice of royal sovereignty also appear, of course, with respect to the founding of non-Muslim states. We know, for example, that Brahmin ideologues, writing in chaste Sanskrit, spun elaborate tales of how warriors and sages founded the Vijayanagara state by combining forces for a common defence of dharma from assaults by barbaric (mleccha) Turkic outsiders. This is the Vijayanagara of rhetoric, a familiar story. But the Vijayanagara of prac tical politics rested on very different foundations, which included the adoption of the titles, the dress, the military organisation, the ruling ideology, the architecture, the urban design, and the political economy of the contemporary Islamicate world. ²⁶ As with Indo-Muslim states, we hear of such practices mainly from outsiders - merchants, diplomats, travellers - and not from Brahmin chroniclers and ideologues.

CONCLUSION

One often hears that between the thirteenth and eighteenth centuries, Indo-Muslim states, driven by a Judeo-Islamic "theology of iconoclasm," by fanaticism, or by sheer lust for plunder, wantonly and indiscriminately indulged in the desecration of Hindu temples. Such a picture cannot, however, be sustained by evidence from original sources for the period after 1192. Had instances of temple desecration been driven by a "theology of iconoclasm," as some have claimed,²⁷ such a theology would hav e committed Muslims in India to destroying all temples everywhere, including ordinary village temples, as opposed to the highly selective operation that seems actually to have taken place.

Rather, contemporary evidence associates instances of temple desecration with the annexation of newly conquered territories held by enemy kings whose domains lay on the path of moving military frontiers. Temple desecrations also occurred when Hindu patro ns of prominent temples committed acts of treason or disloyalty to the Indo-Muslim states they served. Otherwise, temples lying within Indo-Muslim sovereign domains, viewed normally as protected state property, were left unmolested.

Finally, it is important to identify the different meanings that Indians invested in religious monuments, and the different ways these monuments were understood to relate to political authority. In the reign of Aurangzeb, Shaikh Muhammadi took refuge in a mosque believing that that structure - being fundamentally apolitical, indeed above politics - lay beyond the Mughal emperor's reach. Contemporary royal temples, on the other hand, were understood as highly charged political monuments, a circumstance t hat rendered them fatally vulnerable to outside attack by Hindu or Muslim invaders. Therefore, by targeting for desecration those temples that were associated with defeated kings, conquering Turks, when they made their own bid for sovereign domain in Ind ia, were subscribing to, even while they were exploiting, indigenous notions of royal legitimacy.

The fundamentally non-religious nature of these actions is reflected in the fact that contemporary inscriptions in Sanskrit or in regional languages never identified Indo-Muslim invaders in terms of their religion, as Muslims, but generally in terms of t heir linguistic affiliation (most typically as Turk, "turushka"). That is, they were construed as but one ethnic community in India amidst many others. ²⁸ B.D. Chattopadhyaya finds in early medieval Brahminical discourse an essential urge to legitimise any ruler so long as he was both effective and responsible - meaning, in particular, that he protected Brahmin property and temples. Th is urge was manifested, for example, in the perception of the Tughluqs as legitimate successors to the Tomaras and Cahamanas; of a Muslim ruler of Kashmir as having a lunar, Pandava lineage; or of the Mughal emperors as supporters of Ramarajya (th e "kingship of Lord Rama").²⁹ Indo-Muslim policies of protecting temples within their sovereign domains certainly contributedto such perceptions.

In sum, by placing known instances of temple desecration in the larger contexts of Indo-Muslim state-building and state-maintenance, one can find patterns suggesting a rational basis for something commonly dismissed as irrational, or worse. These pattern s also suggest points of continuity with Indian practices that had become customary well before the thirteenth century. Such points of continuity in turn call into serious question the sort of civilisational divide between India's "Hindu" and "Muslim" pe riods first postulated in British colonial historiography and later reproduced in both Pakistani and Hindu nationalist schools. Finally, this essay has sought to identify the different meanings that contemporary actors invested in the public monuments th ey patronised or desecrated, and to reconstruct those meanings on the basis of the practice, and not just the rhetoric, of those actors. Hopefully, the approaches and hypotheses suggested here might facilitate the kind of responsible and constructive dis cussion that this controversial topic so badly needs.

Richard M. Eaton teaches South Asian history at the University of Arizona, U.S., and is the author of several books on pre-modern India, including The Rise of Islam and the Bengal Frontier, 1204-1760 (New Delhi: Oxford University Press, 1997) a nd Essays on Islam and Indian History (New Delhi: Oxford University Press, 2000).

2 Ibn Battuta, Travels in Asia and Africa, 1324-1354, trans. H.A.R. Gibb (1929; repr. New Delhi: Oriental Books Reprint Corporation, 1986), 214.

3 S.L. Sadhu, ed., Medieval Kashmir, Being a Reprint of the Rajataranginis of Jonaraja, Shrivara and Shuka, trans. J.C. Dutt (1898; repr. New Delhi: Atlantic Publishers & Distributors, 1993), 44-45.

4 Nizamuddin Ahmad, Tabaqat-i-Akbari, trans. B. De, 3 vols. (Calcutta: Bibliotheca Indica, 1927-39), 1:386.

5 A useful discussion of Mahmud, his legend, and the question of iconoclasm prior to the establishment of Islamic states in India is found in Richard Davis, Lives of Indian Images (Delhi: Motilal Banarsidass, 1999), chs. 3 and 6.

6 Abu'l-fazl `Allami, A'in-i Akbari, vol. 3, trans. H.S. Jarrett, ed. Jadunath Sarkar (2nd edn. Calcutta: Asiatic Society of Bengal, 1927; repr. New Delhi: Oriental Books Reprint Corp., 1977- 78), 377.

7 Catherine B. Asher, "The Architecture of Raja Man Singh: a Study of Sub-Imperial Patronage," in Barbara Stoler Miller, ed., The Powers of Art: Patronage in Indian Culture (Delhi: Oxford University Press, 1992), 183-201.

8 P. Acharya, "Bruton's Account of Cuttack and Puri," Orissa Historical Research Journal 10, no. 3 (1961), 46.

9 Journal of the Asiatic Society of Bengal (1911), 689-90. Order to Abu'l-Hasan in Banaras, dated 28 Feb., 1659. My translation. The "continuance of the empire" was always forefront on the minds of the Mughals, regardless of what religious functi onary was praying to which deity.

10 See Richard M. Eaton, The Rise of Islam and the Bengal Frontier, 1204-1760 (New Delhi: Oxford Universtiy Press, 1997), 184- 85, 263.

11 Surendra Nath Sinha, Subah of Allahabad under the Great Mughals (New Delhi: Jamia Millia Islamia, 1974), 65-68; Catherine B. Asher, Architecture of Mughal India, vol. I:4 of The New Cambridge History of India (Cambridge: Cambridge University Press, 1992), 254, 278; Saqi Must`ad Khan, Ma'athir-i `Alamgiri (Calcutta: Bibliotheca Indica, 1871), 88.

12 Saqi Must`ad Khan, Maasir-i `Alamgiri, tr. J. Sarkar (Calcutta: Royal Asiatic Society of Bengal, 1947), 57-61; Asher, Architecture, 254.

13 See Sita Ram Goel, Hindu Temples: What Happened to Them, vol. 2: The Islamic Evidence (New Delhi: Voice of India, 1991), 78-79, 83; Sri Ram Sharma, The Religious Policy of the Mughal Emperors (2nd edn.: London: Asia Publishing Hou se, 1962), 132-33; Athar Ali, The Mughal Nobility under Aurangzeb (Bombay: Asia Publishing House, 1966), 98n.

14 Saqi Must`ad Khan, Ma'athir-i `Alamgiri, text, 81. My translation. Cf. Saqi Must`ad Khan, Maasir-i-`Alamgiri: a History of the Emperor Aurangzeb-`Alamgir, trans. Jadunath Sarkar (Lahore: Suhail Academy, 1981), 51-52.

15 Ma'athir-i `Alamgiri, text, 81. Cf. Jadunath Sarkar, trans., Maasir-i-`Alamgiri (Lahore: Suhail Academy, 1981), 51.

16 Wheeler M. Thackston, tr., The Baburnama: Memoirs of Babur, Prince and Emperor (New York: Oxford University Press, 1996), 327.

17 Eaton, Rise of Islam, 176-77.

18 Asher, Architecture of Mughal India, 260.

19 Muzaffar Alam, "Assimilation from a Distance: Confrontation and Sufi Accommodation in Awadh Society," in R. Champakalakshmi and S. Gopal, eds., Tradition, Dissent, and Ideology: Essays in Honour of Romila Thapar (Delhi: Oxford University Press, 1996), 177n.

20 Examples of mosque desecrations are strikingly few in number. There is evidence that in 1680, during Aurangzeb's invasion of Rajasthan, the Rajput chief Bhim Singh, seeking to avenge the emperor's recent destruction of temples in Udaipur and elsewhere , raided Gujarat and plundered Vadnagar, Vishalnagar, and Ahmedabad, in the latter place destroying thirty smaller mosques and one large one. See Raja-sumudra-prasasti, XXII, v. 29, an inscription composed ca. 1683, which appears in Kaviraj Shyama ldas, Vir Vinod (Udaipur: Rajayantralaya, 1886); cited in R.C. Majumdar, ed., The Mughal Empire (Bombay, Bharatiya Vidya Bhavan, 1974), 351.

21 Agha Mahdi Husain, ed. and trans. Futuhu's-salatin, or Shah Namah-i Hind of `Isami (London: Asia Publishing House, 1967), 1:66- 67.

22 Davis, Lives, 71-76.

23 Mahdi Husain, trans., The Rehla of Ibn Battuta (India, Maldive Islands and Ceylon) (Baroda: Oriental Institute, 1953), 4.

24 Sebastiao Manrique, Travels of Fray Sebastien Manrique, 1629-1643, trans. E. Luard and H. Hosten (Oxford: Hakluyt Society, 1927), 1:77.

25 P. Acharya, "Bruton's Account of Cuttack and Puri," in Orissa Historical Research Journal 10, no. 3 (1961), 46.

26 See Phillip B. Wagoner, "'Sultan among Hindu Kings': Dress, Titles, and the Islamicisation of Hindu Culture at Vijayanagara," Journal of Asian Studies 55, no. 4 (Nov. 1996), 851-80; idem., "Harihara, Bukka, and the Sultan: the Delhi Sult anate in the Political Imagination of Vijayanagara," in David Gilmartin and Bruce Lawrence, eds., Beyond Turk and Hindu: Rethinking Religious Identities in Islamicate South Asia (Gainesville: University Press of Florida, 2000).

27 See Andre Wink, al-Hind: the Making of the Indo-Islamic World, vol. 2: The Slave Kings and the Islamic Conquest, 11th-13th Centuries (Leiden: Brill, 1997), 294-333.

28 See Cynthia Talbot, "Inscribing the Other, Inscribing the Self: Hindu-Muslim Identities in Pre-Colonial India," Comparative Studies in Society and History 37, no. 4 (Oct., 1995), 701.

29 Brajadulal Chattopadhyaya, Representing the Other? Sanskrit Sources and the Muslims (8th-14th century) (New Delhi: Manohar, 1998), 49-50, 53, 60, 84.