INTRODUCTION
Why and to what extent did Muslim armies destroy Hindu temples in premodern India? For nearly a century this contentious issue fed British imperial propaganda, inflamed Hindu passions, and raised Muslim anxieties, and today it continues to present scholars of South Asia with a vexed set of questions about history, memory, and trauma.Footnote 1 On one side is the position that we must distinguish rhetoric from action and not accept at face value every boast of iconoclasm found in the Indo-Persian sources.Footnote 2 On the other is the observation that in the wake of the Turkish conquests of the eleventh and twelfth centuries the size and number of temples declined, especially in urban North India, even if the chief reason was not their wanton destruction but the weakening of Hindu royal patrons as Turkic courts and retinues took over the life of many cities.Footnote 3 This leaves us with two emphases, one that looks for acts of accommodation and the other that gives voice to those who lost when various “Hindu” political forms had to make room for “Muslim” ones.
This essay expands the scope of inquiry in two ways. First, instead of treating Turkic and Persian peoples coming into South Asia from the eleventh to the eighteenth centuries as a uniform cultural group, that is, as Muslims with fixed notions of iconoclasm or religious engagement, it asks how these communities and concepts changed over the centuries. Second, rather than imagining these inbound groups only as makers of South Asian history, it also treats them as objects shaped by similarly disruptive developments in Iran and Central Asia. Such a perspective yields a third and distinct emphasis, which, by locating the issue outside the emotionally charged imaginaries of today, has the potential to refocus the scholarly debate. To move in this direction, I propose adding a comparative dimension to my opening question: Was the destruction of Hindu temples in India during times of war comparable to the attacks on Muslim saint shrines in Iran and Central Asia? The answer, I hope, will shed new light on an old scar of history.
TEMPLES AND THE ENACTMENT OF INDIC KINGSHIP
Well before the advent of Turkic and Afghan rulers in South Asia, Hindu kings attacked temples associated with rival sovereigns. The “royal temple” first appeared in North India in the Gupta era, circa the fifth century CE, during the formative phase of what later came to be known as Hinduism, when the older aniconic rites of Vedic sacrifice were subsumed by the veneration of divine images.Footnote 4 As Shiva, Vishnu, and the Goddess became paramount divinities, and their worship aligned with Brahmanical and Vedic norms, their images began to grace the temple's inner sanctum. Now built as freestanding stone structures, temples evoked the height and cosmic significance of the mountains, the ancient abodes of the gods. While the great majority of temples were not associated with royalty, the rising significance of temple divinities did transform the institution of Indic kingship.Footnote 5 By building a temple, a Hindu monarch centered his kingdom within the cosmos and linked his power with that of a major deity. But this was not primarily an act of piety or ecumenical patronage; it was, as Ronald Inden puts it, an effort “to fashion a new chain of being, one that stretched out below him on earth and reached up into the heavens above.”Footnote 6 Thus a new political geography took shape in early medieval India in which grand temples became places for the enactment of universal imperium.
Upon the fragmentation of the Gupta realm in the fifth century, its temple-centered sovereignty provided a charter for the regional polities that followed. Over the next seven hundred years, until the rise of the Turkic and Muslim Delhi Sultanate, South Asia saw a political equilibrium evolve among four geographically differentiated spheres in which no ruler was able to extend his influence over an external region in a durable fashion.Footnote 7 Nonetheless, these regional rulers adopted the totalistic idiom of kingship based on the Gupta template. The result was an era of interregional political competition that spawned an impulse toward universal dominion that was aspirational but unachievable. In an agonistic performance of sovereignty, Hindu kings erected grand temples in their own territory and attempted to loot or tear down those built in their rival's country. Over centuries such practices evolved into a paradigm.Footnote 8 Thus a tenth-century Rashtrakuta inscription celebrated how King Indra III “knocked askew by the strokes of his rutting tuskers” a temple of his enemies.Footnote 9 Such acts were expected of a conqueror, as instructed in a twelfth-century Chalukya advice manual: “The enemy's capital city should be burned—the palace of the king, beautiful buildings, palaces of princes, ministers and high ranking officers, temples, streets with shops, horse and elephant stables.”Footnote 10
When, beginning in the eleventh and twelfth centuries, Muslim rulers from Afghanistan and Central Asia ventured across the river Indus and expanded their presence in South Asia, they also began to enact such rites. As Richard Eaton argues, after the initial flurry of iconoclastic destruction abated, Turkish and Afghan armies mainly targeted temples along a moving frontier of conquest, and especially those that harbored icons linked with local rulers who offered resistance.Footnote 11 In other words, Muslim warriors coming into India learned that the destruction of royal temples was a way to unravel the power of local kings. In times of peace, by contrast, Indo-Muslim rulers permitted the construction and repair of temples and could even play the ritual role of the “Hindu” sovereign in temple festivals. But their Persian chroniclers, constrained by conventions of genre to uphold the iconoclastic ideals of Islam, understated these acts of participation, exaggerating instead their Muslim patrons' deeds of conquest.
Eaton's thesis is widely accepted, but is also qualified at times with the observation that temple desecration by Muslim kings was only partially analogous to similar acts by Hindu rajas.Footnote 12 Muslim monarchs in India destroyed temples as a rite of conquest but did not build them. They built mosques instead. To be sure, early mosques in North India were built from temple spolia by non-Muslim craftsmen who used local Indic techniques and decorations. These buildings may have blended in the landscape as temples of another kind, a phenomenon echoed in Ghaznavid coin inscriptions in India that translated the Arabic word for prophet, rasul, with the Sanskrit word avatara, a divinity descended to earth in human form.Footnote 13 In the other direction, the Perso-Islamic architectural motifs of domes, arches, and aniconic vegetal ornamental patterns provided a new stylistic impulse for Indic and especially Jain temples and shrines.Footnote 14 But even so, the coming of Turkic and Afghan rulers sent a ripple across Indic kingship. Mosques built by Muslim kings did not become sites for Brahmanical rites of sovereignty, but were instead used to fashion a chain of being which, to use Inden's phrase, stretched out below on earth and into the heavens above but also along worldly and cosmic axes that veered away in other directions.
TRANSCENDENT RELIGIONS AND IMMANENT KINGS
Is it right to assume that Muslim kings had the same need to fashion a chain of being as did Hindu kings? Were the former not adherents of a fiercely transcendent and iconoclastic doctrine that was at odds with the concrete and iconophilic ritualism of the latter? To answer such skepticism and soften the contrast between Islam and Hinduism, it helps to view the two sacred traditions in the longue durée, world-historical framework proposed by the eminent sociologist of religion Robert Bellah.Footnote 15
Bellah draws upon the work of Max Weber and his associate Karl Jaspers to argue that post-Vedic Hinduism and the late-Abrahamic traditions—Islam being the last in the series—were products of the “Axial Age,” which, beginning in the first millennium BCE, produced an enduring shift in the nature of organized religion across Eurasia. As part of this transformation, “archaic” religious forms, which had focused primarily on ritual communion with the immanent divine, gave ground to “historic” religions that promised to guide the human self—itself a significant new concept—through the perilous maze of a moral-soteriological order of existence. The result was a set of “world religions” grounded in universal ethics and text-based doctrines of salvation. Not all immanence was lost, though, and not every religion was equally transcendent. There remained some distance between the Abrahamic religions with their strident insistence on a singular deity and Puranic Hinduism with its luxuriant pantheon. Yet, in terms of deep structure, what these religious systems had in common were doctrinal ideals of salvation that were universal, otherworldly, and—unlike the “magical” rituals of archaic religions—relatively impervious to empirical threats.Footnote 16 They possessed at their core what Alan Strathern calls “transcendental intransigence,” a built-in mechanism of boundary maintenance and resistance to conversion.Footnote 17
Most significant for this essay's argument are Bellah and Strathern's perspectives on how these historical transformations in religious thought and practice affected the nature of “divine” and “sacred” kingship. As these religious systems developed independent clerisies with direct doctrinal recourse to an authority higher than the king, the rather typical acceptance of kings as embodying divinity or with supreme priestly access to the gods was abandoned, qualified, or reinterpreted.Footnote 18 The monarch was now constrained, in soteriological theory at least, to be “righteous.” Again, this institutional and intellectual change manifested itself in relative strength, more forcefully in the history of kingship where Abrahamic religions held sway and less so in the history of Indic kingship. But it was in the sphere of kingship that archaic concepts and “immanentist” practices continued to thrive within Eurasian societies dominated by the religions of salvation. Put simply, kings refused to give up their cosmic powers. The reasons why this happened are many sided and much debated, but it is undeniable that the strong persistence of what Francis Oakley terms “cosmic kingship” eventually forced even Christian theologians—and, as we shall see, Muslim jurists—to arrive at a series of accommodations that admitted into doctrine what had widely endured in practice: the sacred and divine nature of the royal body.Footnote 19
The enduring tension created by the formal separation of the king and the clerisy is the crux of the issue. For our analysis, it implies that, in terms of doctrine, both Islam and Hinduism possessed transcendental intransigence.Footnote 20 Yet kings of both religions resisted the policing power of their respective clerisies because they depended on immanentist rituals to establish themselves within the cosmos—to fashion a chain of being. There existed, non-intuitively, a greater potential for cross-cultural exchange in the “enactive” domain of kingship fueled by agonistic competition than in the “theoretic” sphere of doctrinal religion driven by antagonistic argument.Footnote 21 Thus Muslim kings had good cause to assume the customs of Hindu rulers and vice versa, but the clerical classes of both religions had few incentives to rationalize such interactions or record their occurrence. That such exchanges depended on objects and rites more than texts and speech is perhaps why scholars focused on material culture—Finbarr Flood, Phillip Wagoner, and Richard Eaton—have studied them more than scholars of religious texts.Footnote 22
With these insights, we can pick up the thread of our historical analysis and ask: Was the mosque the primary site for the performance of Muslim kingship—the main anchor of the Muslim chain of being? At first this may seem a superfluous question. The building of a congregational mosque where on Friday a public sermon included a prayer for the Muslim ruler was an established custom of sovereignty by the time Muslims reached India. Yet the history by which this came to be and the changes that it underwent merit a closer and comparative look.
THE MUSLIM CHAIN OF BEING
In the eleventh century, when Turkic Muslim armies arrived in North India, the “real” sovereign in Islamic political imagination was not the one who held worldly authority, the sultan, but rather the one who maintained the Muslim community's link with the divine, the caliph. Mahmud of Ghazni (d. 1030), who famously led raids from Afghanistan into India to loot and destroy temples, derived his authority from the ‘Abbasid caliph, having received a caliphal robe of investiture along with a title, “The Right Hand of Sovereignty” (Yamin al-Dawla).Footnote 23 Alongside other rituals of incorporation, the ancient custom of robing had become a mainstay of Muslim religious and royal practices.Footnote 24 A robe that had embraced the sacred body of the caliph would imbue its wearer with sovereign charisma. Such a tactile connection enabled Turkic upstarts like Mahmud—whose father had been captured from Central Asia, converted to Islam, and sold into military slavery to the Samanids in eastern Iran—to partake of caliphal sacredness. The Buyids (934–1055) and the Seljuqs (1037–1194), western neighbors and Muslim competitors of the Ghaznavids (963–1186), also enacted such rites. But in these cases, too, the chief aim was neither piety nor ecumenical outreach but the ritual performance of kingship—an attempt to participate in the cosmology of the ‘Abbasid caliphate.Footnote 25
Crucially, the ‘Abbasids became a wellspring of sacred sovereignty only after their loss of political and doctrinal authority. It was the dissolution of ‘Abbasid power in the second half of the ninth century and the political competition it engendered among their Persian and Turkic successors across the eastern Islamic lands that made “contact” with the caliph such a desired ritual for newly wrought shahs and sultans. This was also the milieu in which conversion to Islam reached its apogee, and a number of newly Islamized warrior groups took political control. One after the other, powerful Muslim rulers declared themselves “slave” (mamluk), “servant” (khadim), and “most lowly devotee” (aqall al-‘abid) of the caliph.Footnote 26 But the steep rise in the formal respect shown to the caliph was inversely proportional to the political or doctrinal authority he wielded. In the seventh century, the powerful caliph had been addressed as an ordinary man, but by the tenth century, the powerless figure was treated like a god with sultans kissing the ground where he had stood.Footnote 27 Muslim kings now declared themselves his “seedlings.”Footnote 28 The caliph had begun to function as a pure icon—in the sense not of an insubstantial symbol but of a substantial object like a holy relic or sacred chalice—in a ritual effort to seed multiple centers of Muslim sovereignty. It would appear, then, that in the tenth and eleventh centuries the body of the ‘Abbasid caliph enveloped in his robe had become for Muslim monarchs in the eastern Islamic lands what for Hindu kings in India was the iconic divinity ensconced in a temple.
The person of the ‘Abbasid caliph was the chief but not the sole icon of Islamic sovereignty in the tenth century. Farther west, the Fatimids (909–1171) in Egypt and North Africa had set up a rival caliphate. They saw themselves as the true heirs of the Prophet through his daughter Fatima and her husband ‘Ali (d. 661), the first cousin of Muhammad and the fourth caliph of Islam. More broadly, the descendants of ‘Ali had since his death in the seventh century staked a claim to earthly and spiritual sovereignty, a privilege that drew much popular support. In their own bid for sovereignty, the ‘Abbasids had used ‘Alid messianic propaganda in eighth-century Iran to foment rebellion against their Umayyad rivals, raising black banners as a sign of the imminent rise of an ‘Alid savior.Footnote 29 And at the height of ‘Abbasid power, the caliph Mamun (d. 833) had pretended to step down in favor of a descendant of ‘Ali, the eighth imam of the Shi‘a, ‘Ali Reza (also “al-Riḍa” in Arabic transliteration, d. 818). Mamun had kept the ‘Alid leader in his retinue until the latter's death and then buried him in Tus in northeastern Iran, next to the grave of Mamun's father, the caliph Harun al-Rashid (d. 809). Centuries later, the Imam's mashhad, or martyrium, evolved into a major shrine in Iran with a city, Mashhad, named after it, a point of significance to which I shall return.
For now, the observation should suffice that even in its early centuries sovereignty in Islam was articulated in a corporeal idiom, one that shared much with the practices of pre-Islamic Iran and India.Footnote 30 These customs became a part of Islam in the eighth and ninth centuries as it spread outside Arabia and new Muslim rulers embraced an ancient and thriving culture of kingship in which the divine was made manifest in the body of the sovereign. From such a perspective, the competition between tenth- and eleventh-century Muslim kings over which caliphal icon to draw sovereignty from is comparable to the way medieval Hindu rulers sought universal imperium via the ritual embrace of either Shiva or Vishnu. But this development remained at odds with the transcendental doctrines of Islam, and the way this tension worked itself out shaped the way Muslim rulers used the caliph and the mosque as icon and space, respectively, to fashion their chain of being.
THE CALIPH AND THE MOSQUE
By the eleventh century, around the time when the caliph became a sacred icon, a consensus developed among jurists in the eastern lands of Islam that a Muslim ruler was a necessary condition for holding Friday prayers in the mosque.Footnote 31 If there was no Muslim ruler, they postulated, it would lead to chaos (fitna) since different groups would offer Friday prayers at different times. But in what seems like a paradox, these same jurists hesitated in giving their support to the custom of praying for the sovereign by the Friday congregation. Although this prayer, in which the living caliph and the ruling sultan were named together, had become a widely enacted rite, most jurists neither sanctioned nor condemned it in their opinion books. A few termed it permissible but “undesirable,” while most simply left the matter unmentioned, an act that Norman Calder describes as “deliberate abstention.”Footnote 32
Calder sees these jurists’ recalcitrance as an attempt to preserve their independence: by not letting earthly practice muddy the waters of divine doctrine, the keepers of Islamic law could maintain a space from where to criticize those in power.Footnote 33 But the jurists' silence can be interpreted another way: that the Muslim clerisy was powerless to prevent this heavy intrusion of immanentist kingship into the transcendental space of the mosque.Footnote 34
The naming of the sultan alongside the caliph during the Friday congregational prayer was a rite not unlike, if less concrete than, the act of robing. Neither custom was openly sanctioned by the Quran or laws derived from it. Yet such rituals were enacted widely because they enabled leaders of newly Islamized groups to insert themselves into the Muslim chain of being. This situation in Islamic history can be compared to the early development of temple Hinduism, when Brahman ritual specialists enabled the aniconic and sacrificial rituals of the ancient Vedas to be adapted for use in icon-centered temples, thereby making temple Hinduism vital to kingship and attractive to rising warrior groups in India. In the Iranian case, a well-studied example of such a development is that of the tenth-century Buyids who came from a lightly Islamized region in the mountains of Daylam (present-day Gilan) near the Caspian Sea.
Even though most Buyids were nominally of ‘Alid or Shi‘i persuasion, as kings they derived their sovereignty from the iconic Sunni ‘Abbasid caliphs. They also turned to pre-Islamic forms of monarchy and celebrated their conquests with the ancient Iranian title “king of kings” (shahanshah). How did the Buyids reconcile their sovereign enactment in the manner of pre-Islamic Persian rulers with participation in the caliphal cosmology of Islam? According to Heribert Busse, their solution was to use the ‘Abbasid caliph as a ritual specialist or, more precisely, as the embodiment of a Zoroastrian priest, a mobedh, who in pre-Islamic times had anointed the monarch.Footnote 35 This act was part of a larger Buyid pattern of harvesting charisma from the pre-Islamic Iranian past: by visiting the monumental ruins of bygone Persian kings, having Zoroastrian priests read the Middle Persian inscriptions at these sites, and inscribing Buyid royal presence there in Arabic.Footnote 36 Toward the end of their century-long dominion, a Buyid ruler insisted that his pre-Islamic Persian title, shahanshah, be used in the Friday congregational prayer, a custom that was continued by the Buyids’ successors the Seljuqs, a Turkic group that conquered Iran in the eleventh century.
The Seljuqs belonged to what Nicola Di Cosmo terms the era of “dual administration” Inner Asian empires, which combined “a nomadic political culture with forms of direct control over the resources of sedentary peoples.”Footnote 37 Their support for the ‘Abbasid caliph was part of a ritual effort to adopt the established forms of sovereignty in Iran. As Seljuq power grew, congregational mosques across the eastern Islamic lands began to recite prayers jointly in the names of the Seljuq sultans and the ‘Abbasid caliphs, in many instances displacing the names of the Fatimid caliphs. But when Seljuq authority weakened then the practice switched in favor of the ‘Abbasids’ rivals, as for instance in 1075 when the public sermon in Mecca was recited in the name of the Fatimid caliph.Footnote 38
Such competition over caliphal icons can also be seen in the case of the Ghaznavids, who in their enunciation of sovereignty shared much with the Buyids and the Seljuqs. The Ghaznavids had also embraced the revival of pre-Islamic Persian kingship, and their raids into India to loot and destroy temples supplied them with the economic and symbolic means to establish their status as new Muslim sovereigns associated with Sunni Islam and the ‘Abbasid caliphate. But, as a powerful king, Mahmud of Ghazni had also been wooed by the Fatimid caliph. The sultan rejected the caliph's offer ritually. He spat in the middle of the Fatimid caliph's letter, tore it, and sent the desecrated objected to the ‘Abbasid caliph.Footnote 39 The ‘Abbasid caliph also burned a robe of the Fatimid caliph that had been given to one of Mahmud's generals.Footnote 40 Along the same lines, on one of his raids into India, Mahmud invaded the Isma‘ili kingdom in Multan associated with the Fatimids, and demolished the Isma‘ili congregational mosque.Footnote 41 Were not these acts—the destruction of the mosque and the desecration of the caliphal letter and the robe—akin to a Hindu king's ritual pillaging of a rival's temple?
Be that as it may, instances of mosque destruction by rival Muslim kings are rare in the historical record. More often mosques were simply taken over and a different set of sovereign names ritually pronounced.Footnote 42 From the pedestal of sacred kingship, the mosque and the temple were analogous, but only in part. The temple manifested the divine in concrete form and fused it with the body of the king. In the mosque, the Muslim sovereign's name was uttered along with that of the caliph—divinity's agent on earth—but the ruler performed no binding ritual act; there was nothing akin to the “great gift” (mahadana) ritual performed by the Indic king in the temple that brought together heaven and earth.Footnote 43 This did not mean that Muslim kings were forever constrained by the transcendentalism of Islam, that they were destined only to be “righteous” and not “sacred” or “divine”; rather, as we have seen in the Buyid, Seljuq, and Ghaznavid examples, it meant that they had to find sites that were complementary to the mosque, and embrace rituals that supplemented the congregational prayer for the ruler, in order to anchor themselves in the Muslim chain of being.
But looking at this moment in Islamic history it is difficult to discern a widely enacted parallel to the Indic mode of ritual building and consecration of holy sites associated with the body of the sovereign, or its converse of their ritual looting and desecration. This would change with the coming of Islam's “age of saints,” which made the enshrined holy man, as opposed to the person of the caliph, central to the performance of Muslim kingship in Central Asia, Iran, and India.
THE TOMB-SHRINE IN EARLY ISLAM
In the eighth and ninth centuries, well before the spread of organized Sufism, votive shrines existed mainly on the geographical and political margins of Islam.Footnote 44 Associated with frontier or sectarian figureheads, these sites were shunned by officialdom and clerisy alike. Nevertheless, as the cultural geography of Islam expanded, these sacred graves served a key role in bringing Islam to the masses. Muslims and non-Muslims, new converts and born believers, men and women, all began to visit these graves to benefit from what were believed to be the sacred remains of biblical prophets, the companions (sahaba) and traditionists (muhadithun) of Muhammad, and over time, the Prophet's descendants, the ‘Alids. Indeed, if there was anything akin to proto-sainthood in Islam's first two centuries, it was associated with the ‘Alids, and it is the popularity of early ‘Alid graves that presages the political competition and violence that came to surround Sufi saint shrines in later centuries.
Again, the early ‘Abbasid case is revealing. This caliphal dynasty had at various times both embraced and spurned the Prophet's descendants and their tombs. As mentioned before, the ‘Abbasids used ‘Alid messianism in their bid for power in the eighth century. But when they murdered their own chief propagandist, Abu Muslim (d. 755), he became a martyr in ‘Alid memory, which endured for centuries in Persian and Turkish epics and in shrines dotting Iran and Central Asia.Footnote 45 Even in their heyday, the ‘Abbasids could not ignore the popularity of such ‘Alid graves. The caliph Harun al-Rashid's (d. 809) wife and grandmother lavished funds on pilgrims to Karbala, the burial spot of Husayn, the martyred son of ‘Ali and grandson of the Prophet.Footnote 46 But a few decades later another powerful ‘Abbasid caliph, Mutawakkil (d. 861), destroyed and leveled Husayn's grave to punish rebellious ‘Alids and to stop the popular pilgrimage.Footnote 47 Such suppression proved ineffectual. The Karbala shrine and other ‘Alid votive sites grew in popular appeal and eventually gained official acceptance after the ‘Abbasids lost political power. Similarly, Husayn's tomb, unadorned in the ninth century, had a dome over it by 951, and was expanded by the Buyids, the aforementioned Iranian dynasty whose rulers, despite being pro-‘Abbasid, were devotees of ‘Ali.Footnote 48
The popularity of ‘Alid saint cults and shrines had contributed to the success of tenth-century ‘Alid missionary movements that brought to power the openly ‘Alid dynasties of the Buyids of Iran and the Fatimids of Egypt.Footnote 49 Although this era did not bring widespread “conversion” to doctrinal Shi‘ism, it did see broader religious and political trends shaped by many ‘Alid customs, a key one being shrine veneration. The ‘Abbasid caliphs who had been buried in simple palace graves in the eighth and ninth centuries were by the second half of the tenth century being interred beneath domed shrines like those of ‘Alid luminaries.Footnote 50 The grave of the famous Sunni jurist Abu Hanifa (d. 767), founder of the Hanafi school of jurisprudence, had already become a famous pilgrimage site by the time a madrasa or college was founded there in 1067.Footnote 51 By the eleventh century, most Islamic colleges and monastic retreats had entombed within them the remains of a pious scholar or revered holy man.Footnote 52 Just as the conversion to Islam was peaking in Iran and the Arab Middle East, and majoritarian religious practice was taking coherent shape, the tomb-shrine, almost imperceptibly, made its way from the edge to the very center of the new religion.Footnote 53
But the saint shrine was not yet central to the performance of Muslim kingship. The transformation that would give the shrine a pivotal role in royal rituals had to wait until after the thirteenth century, when most of the Middle East and Central Asia came under the sway of the Mongols, another Inner Asian group that changed the course of Eurasian history.Footnote 54
RULING FROM THE GRAVE: THE RISE OF THE IMPERIAL SHRINE
Though the Mongols were not Muslims, and ruled for nearly a century without embracing Islam, it was they who ushered in Islam's age of saints. By the time they converted to Islam, a well-organized and popular Sufism had reshaped the practice of religion and politics. Sufi orders absorbed local saint cults, Sufi leadership became hereditary, and Sufi shrines became centers of social life. Sufi leaders commanded considerable material resources and enjoyed high social status. The rise of Muslim sainthood was paralleled by a decline in the ritual value of the caliphate. Again, the non-Muslim Mongols are commonly seen as the key agents of change. Upon conquering Baghdad in 1258, they executed the ‘Abbasid caliph. Although in Mamluk Egypt an enervated caliphal line survived out of Mongol reach, in Iran and Central Asia descent and association with Chinggis Khan replaced the caliph as the dominant marker of sovereignty.
The post-Mongol Muslim empires also brought a broad shift away from Islamic law toward dynastic law.Footnote 55 In a bid to make their dynasties “sacred,” post-Mongol Muslim rulers turned to enshrined saints.Footnote 56 The most emblematic of them, Timur (r. 1370–1405), in his long reign did not build a single religious college (madrasa)—the conventional way of patronizing the Muslim clerisy since the eleventh-century Seljuqs—but instead erected colossal domed shrines for his patron saints and Sufi mentors; his own mausoleum in Samarkand mirrored that of a saint.Footnote 57 In short, there was a paradigm shift in the post-Mongol era in which the iconic caliph was replaced by the iconic saint, giving rise to what may be called the imperial shrine, a sacred site for the derivation and performance of sovereignty. This development is nowhere more apparent than in the history of what was to become the most important shrine of post-Mongol Iran: the tomb of Imam Reza in Mashhad.
The way Mashhad evolved over the centuries from an ‘Alid outpost on the frontier of Islam into a center of imperial patronage and a pilgrimage destination is emblematic of a larger historical process that led to a new style of Muslim sovereignty that was anchored in sainthood.Footnote 58 Early in the ninth century, when Imam Reza was buried in northeastern Iran in the region of Tus, no city existed there and the local population was mainly non-Arab and Zoroastrian. The site of the grave became known as the mashhad of the imam. By the tenth century, like other martyria and pilgrimage centers of growing import, this one had also acquired a domed structure and begun to attract regional rulers. For instance, the much later Arabic chronicles of Ibn Athir (d. 1223) reported that the Ghaznavid ruler Sebuktagin (d. 997) had the tomb demolished.Footnote 59 The historian gave no reason for the destruction, but if it was an act of ritual violence, it was undone by Sebuktagin's son Mahmud of Ghazni, who had the shrine rebuilt. Ibn Athir wrote that Mahmud's change of heart occurred because ‘Ali himself appeared to the sultan in a dream.Footnote 60 It would seem that Mashhad had grown in local importance to the point that it began to draw the attention, both violent and beneficent, of local rulers like the Ghaznavids. Yet Ibn Athir's account comes two centuries too late and, like the extant architectural evidence, is too bare in its details to support the notion that Imam Reza's shrine served as a cornerstone of Ghaznavid sovereignty.
The ‘Alid shrine's importance continued to grow under the Seljuqs, who replaced the Ghaznavids and like them enjoyed a staunchly Sunni and anti-‘Alid reputation.Footnote 61 Ibn Athir recounted how the powerful Seljuq sultan Sanjar once prayed at the shrine with his vizier on his way to battle. However, the shrine, while important enough for the occasional visit, was not yet central for Seljuq kings. More significant is the historian's description of a major riot (fitna ‘azima) at Mashhad in 1116, triggered by a scholarly debate between Shi‘as and Sunnis at the shrine. The people of Tus took advantage of the unrestFootnote 62 and surrounded the shrine, killed everyone inside, and looted a great quantity of wealth. Again, Ibn Athir said little about the attackers’ motive, but he wrote that after the riot ended the residents of the shrine boycotted the Friday sermon. Their refusal to pray congregationally for the Seljuq sovereign elicited a significant response from his local vassal, who raised a defensive wall around the shrine.
By the thirteenth century, Imam Reza's tomb—domed, fortified, and full of treasure—had become a center of the local political economy. When the Ghurids, another regional power, took over in 1201, they confiscated the stores of grain kept there under the saint's protection.Footnote 63 In religious terms, too, it was now much more than a sectarian votive site. Sufi mystics regularly visited, and students of the prophetic tradition (hadith) came there to learn and share their knowledge across sectarian boundaries.Footnote 64 The ninth-century frontier grave of an ‘Alid saint had ripened into a major center of Islam.Footnote 65
It is no small irony that the man responsible for driving a Muslim sultan into the arms of the enshrined saint of Mashhad was a reigning ‘Abbasid caliph, Nasir (r. 1180–1225).Footnote 66 Nasir broke from tradition by refusing to play the purely iconic role that the ‘Abbasid caliphs had been performing for generations. Instead, the caliph took advantage of the crumbling Seljuq state to pursue an active political agenda. A few of his predecessors had made such attempts in the early twelfth century, but none had been as successful as Nasir in establishing actual political authority in Baghdad and its environs.Footnote 67 The caliph's political ambitions drew howls of protest from the sultans; in the Seljuq leadership's opinion: “If the Caliph is the imam, then his constant occupation must be the performance of prayer, as prayer is the foundation of the faith … the interference of the Caliph in the affairs of temporal rule is senseless; they must be entrusted to the sultans.”Footnote 68
The sultan wanted the caliph to perform his ritual duties and sustain the community's link with the divine—to be still and sacred like an icon in a temple. Nasir's ambitions were considerably more down to earth, and he plotted successfully with the rising Turkic dynasty of the Khwarazmshahs to defeat and kill the last Seljuq ruler. But afterward he refused to include Khwarazmshah's name in the Friday sermon in Baghdad, and instead the caliph allied himself with the Ghurids.
In a harsh response, Muhammad Khwarazmshah declared the ‘Abbasid caliph defunct, installed a descendant of ‘Ali as his rival, and in 1217 marched on Baghdad to enforce this change.Footnote 69 This would have been the end of the ‘Abbasid caliphate except that bad weather and the Mongols intervened. Khwarazmshah's army, having lost a considerable number of men and livestock in the snow, was forced to turn back. Although Khwarazmshah could not lay his hands on Nasir, he had him proclaimed dead during Friday prayers. Subsequently Nasir's name was dropped from Friday sermons in mosques across eastern Iran and Central Asia. It would seem that Khwarazmshah had the ‘Abbasid “idol,” a former object of his devotion that had anchored his Muslim chain of being, ritually broken.
Supporters of the caliph also saw the matter in cosmic terms. The chronicler Ibn Athir, who lived through these times, wrote that the sultan's evil ambitions had come to naught because of the auspiciousness of the noble house of the ‘Abbasids.Footnote 70 Such rejoinders raise the question: If Khwarazmshah rejected the cosmic patronage of the ‘Abbasids, with what did he replace it? The Fatimids, once rival caliphs to the ‘Abbasids, were now defunct. Merely installing an unknown ‘Alid as caliph was not enough to challenge the ‘Abbasids’ iconic status, and so Khwarazmshah had to turn to the most important son of ‘Ali in their realm, the enshrined Imam Reza.
As May Farhat has shown, the Mashhad shrine received massive patronage from Muhammad Khwarazmshah, datable to after 1215, at the height of the conflict with the ‘Abbasid caliph. The Turkic ruler commissioned a series of high quality renovations to the shrine's tomb chamber, the most prominent being a monumental inscription over the main doorway: “This is the tomb—sacred, precious, highly esteemed and cleansed—of our leader the Imam, the Sinless, the Preserved, the Persecuted Martyr al-Reza … the son of ‘Ali b. Abu Talib. ‘Ali the Prince of the Faithful, the champion of the upright, the vice-regent of the Apostle who is the spiritual guide of the two worlds, that is Muhammad … the mercy and the blessing of God be upon him and his descendants, the excellent (al-tayyibin) and the Cleansed (al-tahirin) and upon his noble companions (ashab).”Footnote 71
This inscription simultaneously held two contradictory stances: it proclaimed the core Shi‘i tenet of the right of ‘Alid sovereignty, but at the same time broke a major Shi‘i taboo by praying for all the companions (ashab) of the Prophet, many of whom had been rivals of ‘Ali. In other words, it was a strongly enunciated but non-partisan declaration of ‘Alid sovereignty.Footnote 72 Through it, Khwarazmshah challenged the ‘Abbasid caliphate without affirming Shi‘i doctrinal allegiance. The sultan's patronage of the imam's mausoleum must be understood as a ritual effort by a Muslim king, enacted at the most important saint shrine of his realm, to fashion a new chain of being.
Khwarazmshah's move away from the ‘Abbasid caliphate toward enshrined ‘Alid sainthood was soon forgotten when the Mongols swept through Central Asia and Iran in 1219. But, as mentioned earlier, the shift toward saint-derived sovereignty accelerated under the Mongols. That Sufism was becoming a dominant model for religio-political life by the thirteenth century can also be seen in how the ‘Abbasid caliph Nasir himself used its modes and methods to cement his authority, even employing the services of an eminent Sufi, ‘Umar al-Suhrawardi, who served as his ambassador to other Muslim rulers.Footnote 73 The caliph also embraced the organizational techniques of Sufism and allied orders of “chivalry” (futuwwa) in Baghdad, and encouraged his courtiers and officers to submit ritually to him as initiates joining a chivalrous order.Footnote 74 A key feature of chivalry was the exaltation of ‘Ali as the ideal youth (fata) and model of ethical manliness. In all, the caliph tried to buttress his cosmic status by embracing popular ‘Alid loyalty and saint cults, the same phenomena that drew his rival, Khwarazmshah, to Imam Reza's shrine.
THE MONGOLS AND THE SAINT SHRINE
Although the ‘Abbasids and Khwarazmshahs were both thwarted by the Mongols, during the thirteenth century the shrines and institutions of Sufi sainthood spread from Iran and Central Asia westward into Anatolia and eastward into India on the back of Muslim emigration. Even in Iran and Central Asia, the Mongol attitude toward Islam proved to be a fillip for saint shrines. To understand why, it helps to apply the analytical vocabulary of Bellah and Strathern. Mongol religion as practiced by Chinggis Khan did not possess the universal claims and “transcendental intransigence” of soteriological traditions. Despite the violence they wrought in Iran and Central Asia, the Mongols neither wanted to eliminate Islam nor were they particularly hesitant to adopt it, just as they willingly embraced Buddhism and Christianity elsewhere. Mongol kings saw organized religion as an instrument of this world, as a means to cosmic power, not salvation. They bent religious customs to their will, producing in the case of Islam new forms of sacred art such as the previously “forbidden” portraits of the prophet Muhammad.Footnote 75 When seen from Inner Asian eyes, the thirteenth century appears as one of not civilizational regress and destruction but rather institutional progress and cultural creativity. As scholars like Nicola Di Cosmo and Victor Lieberman have argued, beginning in the first millennium BCE and accelerating rapidly in 900 CE, successive Inner Asian peoples built upon the experience of their nomadic predecessors and increased their ability to conquer and extract resources from the settled agricultural lands of Asia.Footnote 76 The Seljuqs, Khwarazmshahs, and to a degree the Ghaznavids can be seen as instantiations of this process in Iran and Central Asia. Similarly, as the Chinggisid Mongols (called Ilkhanids in Iran) conquered these Muslim regions, they not only learned to harness Persianate court culture and administrative techniques to establish their rule but also Muslim religious institutions. The most significant of these institutions was the saintly shrine.
In the beginning, the Mongols routinely sacked and desecrated Muslim shrines in Iran and elsewhere.Footnote 77 The Mashhad shrine was damaged in the initial wave of conquest in the 1220s.Footnote 78 We know it recovered because there is a report that it was looted a few decades later, when Mongol raiders made off with four silver globes placed over the imam's tomb.Footnote 79 The presence of expensive shrine ornaments along with the survival of pre-Mongol-era inscriptions indicate that the shrine had continued to thrive even under the adversities of Mongol rule. Sufi shrines prospered under the Inner Asians because even as non-Muslims they were drawn to the patronage of charismatic saints and associated brotherhoods, a process that only accelerated after the Mongol elite embraced Islam.Footnote 80 When in 1295 a Sufi scholar with a renowned spiritual pedigree, Shaykh Sadr al-Din Ibrahim, ceremoniously converted the young Mongol king Ghazan Khan, an eyewitness account recorded the ruler wearing a Sufi wool garment (suf) and talisman (haikal); another report suggests that, by raising “black banners,” the occasion was celebrated as the enactment of an ‘Alid messianic prophecy.Footnote 81 As a Muslim king, Ghazan Khan also went on a tour of the major saint shrines of Iran, where lay the “real” sovereigns of his realm.Footnote 82
In the era of Muslim Mongol rule, there were two major developments associated with Muslim shrines in Iran and Central Asia, and both were reflected in the changes to the Mashhad mausoleum. One development was architectural. Shrines grew in size and became centers of cultural innovation. For instance, Ghazan Khan, in a bid to associate ‘Alid charisma with his Chinggisid lineage, built in the major cities of his realm a new type of hospice for the benefit of the ‘Alids called “Hall of the Prophet's Kin” (dar al-siyada), which later became a feature of ‘Alid and Sufi shrines.Footnote 83 This was part of a larger process that spread across Muslim Asia in which saint shrines began to anchor extensive religious complexes containing madrasas, hospices, hospitals, and most significantly, mosques—an early notable instance being the Sultaniyya complex in Ilkhanid Tabriz.Footnote 84 These shrine complexes were built with prominent exterior profiles and eye-catching decoration in an effort to draw in pilgrims. With increased pilgrimage, entire towns complete with agricultural hinterlands sprang up around major saint shrines, and Mashhad also grew significantly in this era.Footnote 85 Moreover, people began to build large mosques as attachments to shrines—Timur's daughter-in-law Gawhar Shad built a magnificent one near Imam Reza's mausoleum.Footnote 86 This phenomenon is often seen as a revival of Sunni Islam in post-Mongol times, but it also reveals an inversion of sacred priority, as the shrine came to eclipse the mosque. While previously a Muslim body had been ennobled by being buried adjacent to a mosque, now a mosque was erected besides an enshrined saint. What had been unthinkable in the ninth century became the order of things in the fifteenth.
The second major development was the ritual derivation of royal sovereignty from saint shrines. Beginning with the first Muslim Mongol ruler Ghazan Khan and continuing with the fourteenth- and fifteenth-century successors of the Mongols, we see the custom of kings regularly visiting and patronizing the important shrines of their realm. The greatest of the post-Mongol conquerors, Timur, made this a rite of conquest. Later Timurid chronicles, which were written to reduce his “Mongolness” and exalt his “Muslimness,” inflated the number of visits Timur made to saint shrines during his thirty-five years of nearly continuous war campaigns.Footnote 87 As it became widely accepted that enshrined saints were the “real” sovereigns who granted authorities to earthly rulers, Muslim rulers began to link their power to one or two enshrined patron saints in their dominions. There developed extensive ritual measures for this purpose: coronations at shrines, anointments via dreams and visions of saints, burials of ancestors with saints, exclusive royal patronage of annual saint festivals, construction of mosques adjacent to shrines, votive gifts to shrines upon success in battle, perpetual endowments of tax revenue for shrine upkeep and welfare of pilgrims, and so on. Conversely, saintly lineages that benefited from such attentions assumed the rituals and expressions of kingship.Footnote 88
It should not surprise us, then, that by the fifteenth century Imam Reza began to rule from his shrine with a new title: the Sultan of Khurasan.Footnote 89
THE DESECRATION OF ENSHRINED RIVALS
The thirteenth-century geographer Qazwini left us an intriguing anecdote about the Mashhad shrine.Footnote 90 The ‘Abbasids, he reported, had tricked the supporters of ‘Ali by switching the bodies of Imam Reza and the caliph Harun al-Rashid so that the prayers offered for the former would instead favor the latter. Qazwini's report becomes especially relevant when read alongside the observation of the North African traveler Ibn Battuta, who visited Imam Reza's shrine in the fourteenth century. Ibn Battuta wrote that Shi‘i pilgrims to Mashhad prayed at the tomb of the imam but kicked the grave of the ‘Abbasid caliph next to it.Footnote 91 If they had indeed switched the two bodies, the ‘Abbasids had correctly anticipated the desecration of their caliph's shrine by their ‘Alid rivals. Such rivalry over enshrined sovereigns intensified in the post-Mongol era and violence committed against the patron saints of enemies increased sharply. With the shift from the caliph to the shrine as a fount of sovereignty, the mode of ritual desecration took on the characteristics of the temple-centered sovereignty of Indic kingship.
The political history of Iran and its environs in the century after the Mongol eruption, which was followed within a century by the destructive conquests of Timur, is extremely fragmented and the religious history is even more opaque.Footnote 92 And yet amid the confusion of the time there surfaced a dominant trend: the rise of charismatic saints and messiahs.Footnote 93 In Central Asia, Iran, and even India there occurred movements led by Sufis who claimed to be the saint of the age, the messiah, and often the embodiment of ‘Ali himself.Footnote 94 Nurbakhshi, Shaykhi, Musha‘sha‘i, Hurufi, Ahl-i Haqq, and Safavi are some of the better-studied charismatic movements in which Sufi saints sought direct political authority.Footnote 95 It is no wonder that this era has been called the “messianic age” of Islam. Even elite Sufi literature responded by formalizing theories of transmigration (tanasukh) and projection (buruz) of the soul to explain how a past sovereign or saint could experience rebirth in the present or future.Footnote 96 It was as if the enshrined saints had become reincarnated as worldly saviors. While this effusion of chiliastic movements may seem unsettling, it was a product of the historical trajectory traced above, which brought a rise in the number and significance of saint shrines. Sufis who derived their status from sainthood now had the socioeconomic potential and the institutional wherewithal to make successful bids for power. In the absence of the caliph, Sufis and sultans had to come to terms with each other as both competitors and collaborators.Footnote 97 Accordingly, this milieu saw an increase in shrine desecration.
Take, for instance, the case of the millenarian Musha‘sha‘i movement in Iraq whose leader Muhammad ibn Falah (d. 1451) called himself the messiah (mahdi) of Islam.Footnote 98 His son, Mawla ‘Ali (d. 1457), went even further and claimed that he was the ‘Ali reborn and, hence, divine. The Musha‘sha‘i had managed to rally large numbers of Arab tribesmen behind them against Turkic rulers and were able to take control of a number of cities in Iraq. Upon conquering Najaf, the most striking thing Mawla ‘Ali did was to sack the shrine of ‘Ali.Footnote 99 First he looted it, extracting as much treasure he could in the form of ceremonial swords and gold and silver chandeliers that had accumulated there over the centuries. Then he had the wooden covering on the grave split open and set on fire. His explanation reportedly was, “[‘Ali] is God, and God does not die.”Footnote 100 Since Mawla ‘Ali was himself an embodiment of ‘Ali and God, he apparently did not need the Najaf shrine as a source of sovereignty. Perhaps the Musha‘sha‘i messiah even saw ‘Ali's grave as a competing icon of sovereignty.
Fifty years later the Safavids rose to power in the same region. Although now known as the dynasty that converted the Sunni-Sufi population of Iran to Shi‘ism, the Safavids were originally an ‘Alid Sufi group that had advocated, much in line with the Musha‘sha‘i ethos, a millennial worldview complete with the recycling of saintly souls. Their leader, Shah Isma‘il, was considered to be the embodiment of ‘Ali and venerated as the Perfect Guide (murshid-i kamil) by his soldier devotees, the Qizilbash.Footnote 101 Within ten years of their first major victory in 1501, the Safavids had conquered Iran and parts of Central Asia. In the process they attacked rival Sufi networks. As one sixteenth-century observer recorded, the Safavids “crushed all the silsilas (Sufi orders); the graves of their ancestors were destroyed, not to mention what befell their successors.”Footnote 102 Some were the graves of iconic Sunnis, such as that of the jurist Abu Hanifa in Baghdad,Footnote 103 but this was not always the case. For instance, the Nurbakhshi Sufis’ allegiance with Shi‘ism should have shielded them from Safavid wrath but did not because their leader had built a fortress and was seen to be acting like a king. Any rival with an independent—that is, saintly and enshrined—source of sovereignty could be a threat. Thus, when their own Qizilbash soldiers rebelled, the Safavids destroyed the shrine of the legendary ‘Alid hero Abu Muslim, whom the Qizilbash revered as a patron saint.Footnote 104 Conversely, when Sufi orders like the Ni‘matullahi submitted peacefully to Safavid suzerainty, their shrines and sacred sites were left untouched and allowed to flourish. In sum, as they conquered Iran, the Safavids destroyed Sufi shrines as a rite of conquest.
A contextual sense of such rites of war can be gleaned from the Safavid conquest of Herat. The city had long been a center of Timurid power, but when the Timurid king Husayn Bayqara died in 1506, the Uzbeks, ruled by Shaybani Khan, took over. Then, in 1510, the Safavids defeated the Uzbek conqueror, and ritually desecrated the king's body. The Qizilbash reportedly devoured his flesh on the battlefield, and the Safavid Sufi king ordered a drinking cup made from his skull.Footnote 105 But the physical and ritual elimination of Shaybani Khan had little relevance for the people of Herat, who had no loyalty to the Uzbek sovereign. So how were they to be reminded that they had been conquered? The answer, it seems, was to desecrate the shrine of Herat's renowned saint, the Sufi poet ‘Abd al-Rahman Jami (d. 1492), who, along with his influential Naqshbandi brotherhood, had been allied with the Timurid ruling establishment and apparently harbored anti-Shi‘i sentiments.Footnote 106
According to an eyewitness account, when the Safavid forces entered Herat they gathered the city's notables in the courtyard of the congregational mosque to recite the victory proclamation (fathnama).Footnote 107 Commoners crowded on the roof to watch. The scion of an ‘Alid family volunteered to perform the ceremony. When the reciter realized that the ceremony required him to publicly curse the enemies of ‘Ali, the other companions of the Prophet, he hesitated and skipped over the worst language. This incensed the Qizilbash. The Safavid royal agent exclaimed, “How dare he shortchange (khayanat) the glory of the Shah!” and had the man hacked to death at the foot of the pulpit. Then the Qizilbash marched those present in a large procession to the Naqshbandi saint's shrine. Wood was piled up on the grave to the height of the vault and set ablaze, and Herat's citizens were made to watch on pain of death. The traumatic memory of the destruction of Jami's grave was kept alive in Herat into the twentieth century. These modern accounts repeat a story that also spread at the time that the Naqshbandi shrine attendants had anticipated the Safavid attack and so moved the saint's body to a secret location to prevent its desecration, a “resistance” trope comparable to those in stories from India of images rescued from temple destruction there.Footnote 108
The Safavids were not just the perpetrators of such ritual violence, and their own shrines were desecrated as well. As they transformed themselves from a Sufi lineage into an imperial dynasty, they enacted their sovereignty at two major shrines of their realm, initially at their own family shrine in Ardabil in northwestern Iran and, only a generation later, at the mausoleum of Imam Reza at Mashhad, the “Sultan of Khurasan.”Footnote 109 Under the Safavids, the shrine of Imam Reza was brought for the first time under direct control of the state, which appointed its officials, oversaw its endowments and accounts, and reserved its patronage for the monarch. The second Safavid ruler, Shah Tahmasb, had the dome gilded and topped with a gold spike. He donated gold and silver chandeliers and created a hall for the descendants of the Prophet (dar al-siyada). Most significant was the Safavid use of the Mashhad shrine for holding imperial court and for burying emperors. Such acts merged Safavid sovereignty with that of the enshrined saint. This also meant that when the Uzbeks took Mashhad, the shrine of Imam Reza, as an icon of Safavid sovereignty, became a prime target.
The Uzbeks were Sunnis who also venerated the imam and dared not desecrate his grave. But they could loot and undo the Safavid patronage of the shrine. In sixteenth-century Uzbek attacks on Mashhad, the gold spike was taken from the dome, the gold and silver chandeliers were stolen, and jewels from the treasury were used to adorn the Uzbek throne. Further, buildings the Safavids had endowed around the shrine were destroyed, the bodies of Safavid rulers dug up, and the Safavid-appointed shrine caretakers slaughtered.Footnote 110 The Uzbeks might have been Hindu rajas attacking the temple of a rival king.Footnote 111
SULTANS AND RAJAS: SHRINES AND TEMPLES
Shrine-centered sovereignty in Islam and temple-centered sovereignty in Hinduism developed independently in distinct historical eras, but they eventually intersected. At one level, this interaction occurred as saint shrines began to spread across northern India. The process started haltingly in the eleventh century with the coming of Turkic and Afghan armies and grew steadier from the early 1200s.Footnote 112 Yet in 1229, the Turkic ruler of Delhi, Iltutmish, asked for a robe of investiture from the ‘Abbasid caliph.Footnote 113 Even though caliphal sovereignty was on the wane in Iran at this time, in India there was no alternative ritual model for Muslim rulers. There were no long-standing ‘Alid or other saint cults akin to the Mashhad shrine. For good reason, Iltutmish and his successors saw Sufis, who often claimed that their saints were the true sovereigns of the land, more as a nuisance than an aid to rule. Thus till the end of the thirteenth century the Delhi sultans tried to keep Sufi institutions away from their core dominions.Footnote 114
The ‘Abbasid caliph remained an icon of sovereignty in India until the beginning of the fourteenth century, and for fifty years after the Mongols executed the caliph in 1258 coins continued to be struck in his name.Footnote 115 Eventually, in 1320, the last Khalji sultan declared himself the caliph. If this seems presumptuous, consider the report that his predecessor had wanted to declare himself a new prophet, but was dissuaded by his court savant, who pointed out to the Khalji ruler that even the mighty Chinggis Khan had been unable to replace Islam with his religion in Iran and Central Asia.Footnote 116 Time certainly seemed ripe for a new way to establish the Muslim chain of being.
By this time, the post-Mongol waves of Iranian and Central Asian emigrants and refugees into India had aided the growth of saint shrines.Footnote 117 Key rites of shrine-centered sainthood, such as the annual “wedding” (‘urs) or death anniversary celebration of the saint, had also migrated from eastern Iran into India.Footnote 118 Over the fourteenth century, the notion that the enshrined saint was an icon of sovereignty began to be accepted by Indian Muslim rulers. We can see this process at work in the history of the Chishti shrine at Ajmer, which became by the end of the sixteenth century the central imperial shrine of the Mughal Empire, a position comparable to that held by the Mashhad shrine in Safavid Iran.
The Ajmer shrine was little more than a mud grave when the Chishti saint, a migrant from Chisht in Khurasan, was buried there in the thirteenth century.Footnote 119 There were few Muslims and the ruler was Hindu. The “frontier” circumstances of the Chishti grave were similar to those surrounding the grave of Imam Reza in the eighth century. In its first two hundred years, the Ajmer grave acquired a stone cenotaph and little else. But it began to draw the interest of kings. Around 1332, the powerful Tughluq ruler paid it a visit.Footnote 120 In 1395, a Muslim governor of Gujarat prayed there before declaring himself a sovereign.Footnote 121 Fifty years later, another Muslim ruler, the sultan of Malwa, fought a battle with a Hindu raja and upon victory sponsored the building of a dome over the buried saint. By the early sixteenth century, the domed shrine was receiving large numbers of pilgrims, both Hindu and Muslim, who filled its coffers with votive offerings.Footnote 122 The Chishti saint was now a regional icon of sovereignty over which Hindu and Muslim rulers competed.
A late sixteenth-century Chishti compilation of the lives of Sufi saints, the Akhbar al-Akhyar, recorded that in 1526 when the raja of Jodhpur conquered Ajmer he converted the shrine into a temple and placed Hindu images over the saint's tomb.Footnote 123 The Muslim shrine thus served for a time as a royal temple, but when Bahadur Shah, a prince of Gujarat who had rebelled against his father, came to Ajmer and found it in Hindu hands, he made the following vow: “If God Almighty makes me a king (padishah), I will do my utmost to wreak vengeance for this from the unbelievers.” According to the Chishti text, Bahadur Shah upon becoming the sultan of Gujarat returned and made good on his promise to the enshrined saint. In sum, the Chishti shrine in Ajmer had evolved, much as Imam Reza's mausoleum had, from a marginal “frontier” grave into a major locus of sovereignty. In the process both shrines had become objects of both political patronage and agonistic pillage.
In 1526, another ambitious warrior, the Timurid Mirza Babur (d. 1530), rode in from Central Asia to carve out a kingdom for himself in North India, which later became the Mughal Empire. Babur's case highlights an important new perspective on how the shrine-centered sovereignty of Iran and Central Asia and the temple-centered sovereignty of India came to overlap. In his long princely career outside India, he had learned how to use saints and shrines as central props in his sovereign performance. We must ask how this experience prepared Babur to engage with Indian temples and images during the last four years of his life.
In his first two decades as a Timurid prince, Babur made three attempts to capture Samarkand.Footnote 124 During this time his archenemies had been fellow Sunnis, the aforementioned Uzbeks. When the Safavids rose to power, Babur was quick to side with the new conquerors against the Uzbeks. He accepted Safavid suzerainty and took Samarkand with their military support. It is notable that Babur subordinated himself, both religiously and politically, to a power that not only espoused an extremely “heretical” form of Islam—the Safavid shah was, after all, ‘Ali incarnate—but had also desecrated the Sufi shrines of Naqshbandi saints. But if he felt the ignominy of Safavid dominance, Babur did not write about it in his memoir, the Baburnama. The years in which Babur became a Safavid vassal are missing from his annalistic account, which does not record the events from 1508 to 1519. Neither does he mention the widespread Safavid desecration of Sufi shrines. However, when Babur acted the part of the victor, he followed a script of conquest similar to the Safavid one.
A few years after Babur's alliance with the Safavids ended, the Uzbeks pushed him out of Samarkand and Transoxania. He moved south, gained control of Kabul, and began his forays into India. In one such raid, on the border of what are today Afghanistan and Pakistan, he ordered the saint shrine of his Muslim Afghan enemies destroyed.Footnote 125 Babur described the events of the campaign with pride: the siege and storm of a mountain fort, the massacre of three thousand of its defenders, the execution of the local ruler, the enslavement of women and children, the building of a tower of skulls, and the confiscation of the villagers’ grain supplies. Later, Babur released the prisoners and reunited them with their families. He also took as wife the daughter of a local Afghan chief, but sent off his father-in-law with a robe of honor and written “instructions and threats.” Babur's final act of triumphal satisfaction was the demolition of a saint shrine venerated by two of the Afghan tribes of the region.
This shrine destruction took place a few weeks after the initial battle, as Babur continued his raids against the Afghans. He justified the act by stating that the mendicant Sufi (qalandar) buried there had been an apostate (mulhid) who had led the Afghans astray from Islam, and that the shrine, which was perched on a low mountain overlooking a beautiful valley, marred the landscape. However, Babur's concern for the doctrines of Islam rings hollow when we consider the fact, left unmentioned in his memoir, that he had recently sold his soul to the Safavid heresiarch, Shah Isma‘il. Babur's actions also betray a less pious purpose. After he had the shrine reduced to rubble, Babur sat there for a while and, with the remark, “since it was an agreeable place and the air was so good,” chewed ma‘jun, an intoxicant.Footnote 126 His conduct resembled less that of a Muslim seeking salvation and more that of a king enacting his sovereignty.
On Indian soil, Babur adapted his rites of conquest to local idioms of patronage and pillage. Let us begin with patronage. Early on, when faced with a massive army of the Hindu ruler Rana Sangha, his Turkish soldiers began to lose heart and his Indian allies to desert, and his Iranian astrologer forecast defeat.Footnote 127 Babur refused to give up, and instead acted ritually to reset the cosmic balance in his favor. First, he did something very Islamic when he and around three hundred of his generals and officers—all hard-drinking Muslims—publicly took an oath of temperance and the forbidden drink was poured onto the ground. The king and his men also swore on the Quran to keep on fighting, but now the un-Islamic wine had polluted the Indic earth. To atone, Babur ordered a step-well (vayin) dug where the wine had been spilled and made it a site of charity and pilgrimage.
The step-well (vayin, vav, bayin, or baoli) was a particularly Indic religious edifice, widely associated with Hindu temples and shrines and nowhere to be found in Iran or Central Asia.Footnote 128 In effect, Babur had improvised in the midst of war and combined Islamic acts of atonement with Indic acts of expiation. Many in his army were “Hindustanis,” men of Indian birth who, whether Hindu or Muslim, would have responded to Indic sacred customs. In an effort to summon cosmic powers to his side—to assuage the planets and establish his chain of being—Babur had erected a manner of saint-less shrine that transcended religious difference.
Now let us turn to pillage. With Rana Sangha defeated and his conquests more secure, Babur set off to survey his new dominions and offer terms to his new Hindu and Muslim vassals. He was in Gwalior when negotiations began with Rana's younger son via intermediaries. There, in the valley of Urwahi, Babur saw giant statues carved into the hillside. As if on a whim—he remarked in his diary that the place's “one drawback was its idols”—he had them destroyed.Footnote 129
Babur's desecration of Hindu “idols” is comparable to his destruction of the Afghan Sufi shrine. Such acts wrought for the king a signature of power and a seal of victory, which, by raising his status, also enabled him to carry out schemes of accommodation and alliance. This interpretation is borne out by how well Babur behaved over the next few days. News came that the Rana's son had accepted Timurid terms and offered fealty. Out on a tour of Gwalior's many temples, Babur described the varied architecture of the Hindu buildings, even noting in one instance that in the “lower chambers are stone idols.” He did not loot or desecrate but took in the sights and enjoyed himself. The Muslim conqueror seemed positively at ease in the Hindu edifices. Overcome by a sense of familiarity, Babur confessed that the temples felt to him “like madrasas, with porches and large, tall domes and chambers like those of a madrasa.”Footnote 130 In a good mood, he ended the day with a feast in a nearby garden. The Timurid ruler could have been in Herat, Kabul, or Samarkand on a tour of saintly shrines, which were always surrounded by madrasas, mosques, and cloisters, and were never far from a country garden. It was all in a day's work for a Muslim king in the age of saints.
CONCLUSION
In his excellent study The Lives of Indian Images, Richard Davis describes how religious icons came to be of central significance in medieval India.Footnote 131 In the seventh and eighth centuries, new Sanskrit texts, the agamas, were composed, which provided standardized instructions on how to worship images in temples. The theory of the transmigration of the soul was used to explain the mechanism by which a temple image came alive. A new devotional culture developed and along with it a new period eye—a way of perceiving images as alive and divine. Thus began the age of “image sovereignty,” when the real sovereign was presumed to be the divinity resident in a temple image. From then on, a Hindu sovereign was required to merge his “self” with the divine image in the temple's inner sanctum. As a result temple images began to be looted. Temple icons became trophies of war to be broken or carried away by rival kings, but such embarrassing acts were rarely mentioned or, to take the sting out of defeat, stories were told that the real “living” image had been replaced with a fake one. However, Muslim chroniclers in India, Davis complains, viewed Indian images with a single interpretive lens, reducing “the complex world of Indian images, with their varied iconographic forms and complicated mythological backgrounds, to a single interpretive criterion. All were classified under the general rubric of ‘idols.’”Footnote 132 Davis is right that Muslim chronicles did take such a stance, but did Muslim kings?
What I have suggested is that premodern Muslim attitudes toward Hindu images were shaped not just by Islam's doctrines of transcendence and salvation, but also by an evolving pattern of immanentist Muslim kingship that took shape alongside the cults of Muslim saints. Thus the phenomenon of saint sovereignty in Islam shared many parallels with image sovereignty in Hinduism. Temples and images were destroyed by Muslim rulers in India, but so were domed sepulchers of saints in Iran and Central Asia. At issue, in both cases, was not salvation but cosmic power. Surely Hindu kings would have agreed.
