1

In the early decades of the eleventh century, when Muslim armies were marching into India, they took with them the famous Central Asian polymath Abu Rayhan al-Biruni on a number of campaigns. Although critical of Sultan Mahmud of Ghazna, the monarch who conceived of and led these campaigns, our brilliant polymath could not resist the lure of the new knowledge in India that Ghaznavid conquests had made possible. Travelling across the vast Gangetic plains and in beautifully mountainous Kashmir, al-Biruni learned Sanskrit as best as he could, conversed with pundits, read books, and then compiled the results of his research into a volume verbosely entitled after the fashion of the day Kitāb Taḥqīq mā li al-Hind min Ma‘qūla Maqbūla fī al-‘Aql aw Marḍūla, “The book of research of what pertains to India on matters whether acceptable by reason or rejected thereby” or commonly referred to in English by the mercifully short title of Indica. This substantial work by a prolific and perceptive scholar has been dubbed “the apex of Islamic scholarship on non-Muslim religious traditions”, and has received well-deserved attention for over a century now.Footnote ¹

Yet, while scholars have long analyzed this text against its possible sources and have pursued its various threads into Brahmanical religion, art, and sciences, no one to my knowledge has appreciated the significance of three brief passages in the text that are bizarre given the political context of the author's life and those of his royal patrons. These have to do with elephants and their symbolic significance in Hinduism. First, in a chapter called “The Beginnings of Idol Worship”, al-Biruni described an episode he had heard about in which the Vedic god Indra had appeared to a king named Ambarisha. According to this tradition, the apotheosis of Indra had occurred in the following manner: “Indra, the prince of the gods, [came] riding on an elephant”.Footnote ²

Indeed, al-Biruni had come to understand that there existed an association among Indra, kingship, and the elephant in eleventh-century Hindustan. Was this association current or was the author resurrecting an arcane practice from ancient books that was no longer relevant in his own time? Further down the same section al-Biruni quoted the Sanskrit guidelines for making the “idol” representing Indra. The god was to be depicted “rid[ing] on a white elephant”.Footnote ³ Elsewhere, in the section on “Astrology”, al-Biruni wrote the following about thunder, “Thunder is the roaring of Airavata, i.e. the riding-elephant of Indra the ruler”.Footnote ⁴ In short, thanks to al-Biruni's writings at least, a princely figure riding the elephant recalled the chief of the Hindu pantheon for members of the Ghaznavid court who now had some knowledge of but little religious sympathy for Hinduism.

But, how did al-Biruni and his audience reconcile this informed association with another image—that of the reigning Muslim monarch Sultan Mahmud, who appeared frequently atop the elephant to his subjects? Would not Mahmud, who presented himself as a relentless champion of Islam against Indian polytheism, recall in the minds of al-Biruni and others the Hindu god Indra when he paraded himself royally on Indra's animal? Would he not, in other words, come across as one of the highest embodiment of the very religion that he was supposedly at pains to extirpate?

By following this disjuncture in al-Biruni's writing, this article will pursue some significant and heretofore neglected aspects of sovereignty that seem to have arisen during the early “Islamic conquests” of India. I will argue that the Ghaznavids were actually constructing an imperial polity both in the subcontinent and West Asia. As such they had to promote their rule as legitimate to their non-Muslim subjects as well as to the Muslims. This involved, among other things, assuming a symbol of royalty that made sense to their Indian subordinates—the elephant being one of the most conspicuous. As we will see however, the animal had a very different symbolic significance among the Muslim population of West Asia: a negative one. A good deal of effort therefore went into offsetting the damaging associations that the elephant had brought upon Ghanznavid kings. The iconoclastic violence that is celebrated in Ghaznavid historiography should be partially understood in this context.

2

This argument goes against the grain of modern scholarship which has always tried to grapple with the scale of destruction described by the Persian and Arabic sources of the time. One school takes the sources at face value and sees these events as motivated by religious zeal. This line of scholarship goes all the way back to early British colonial figures, such the famous Henry Meiers Elliot who saw the invasions as the embodiment of Muslim encounters with pagan India—violent, fanatical, and always antagonistic.Footnote ⁵ While Elliot's position remains alive in popular political discourse of modern South Asia, it has been mostly abandoned by academics. Another strand of scholarship, exemplified by the works of V. Bartol‘d, Alexander Haig, and C. E. Bosworth dismisses the purported religious motivations as hypocritical and instead credits the lust for plunder for Ghaznavid military activity in India.Footnote ⁶ Yet another strand, represented by D. G. Tor, Jürgen Paul, and others has focused more on the Islamisation rhetoric of the Ghaznavids and analyzed it as part of a campaign to legitimise the Ghaznavids who were low-born upstarts in the Islamic world.Footnote ⁷ A fourth strand, perhaps best known by the contribution of John Keay, astutely draws numerous parallels between the imperialist activity of Muslim Ghaznavids and Hindu Cholas in South India in order to dismiss the communalist binary originating in Elliot's work. As it turns out, both the Ghaznavids and the Cholas behaved in similar ways in the Indian subcontinent and proudly boasted of their violent actions in the sources they left behind. It is surely modern nationalist ideologies that see these two polities as wholly apart: one alien and intrusive, the other native and unifying.

However, if we accept Keay's arguments, we need no longer simply view the Ghaznavid invasions as a series of devastating though profitable raids. Rather, if supported by evidence, we may conceive of them as part of a project of empire-building. This would of course violate the political chronology of medieval Indian history. Modern scholarship considers the time of the establishment of the Islamic Sultanates of South Asia to be in the late twelfth and early thirteenth centuries when Turkic slave generals put down roots in Delhi and elsewhere on the Gangetic plains.Footnote ⁸ Perhaps the best way to reconcile these chronologies is to consider both the Ghaznavid Empire and the Delhi Sultanate not as parts of a single continuous polity but as the expression of two separate states, the earlier being of much shorter duration and material significance. This position can help us to both build upon and also move beyond the most recent and probably most influential exploration of the topic—Finbarr B. Flood's 2009 Objects of Translations: Material Culture and Medieval “Hindu-Muslim” Encounter.

Flood has amassed an impressive array of textual and material evidence to argue that the Ghaznavids (among others) “reached accommodation with north Indian rulers” and that they engaged in “cross cultural encounters” with their non-Muslim counterparts.Footnote ⁹ Flood has thus completely reversed the communalist thesis of Elliot and others by showing knowledge, interdependence, and peaceful exchange between the Ghaznavids and their South Asian counterparts. However, Flood's analysis falls short in an important way. By divorcing especially the textual evidence from its context and sequence in which it occurs in the sources, Flood can primarily negate the problematic communalist view that had pitted Muslims and Hindus as eternal enemies. However, like them, Flood's argument cannot account for the contradictory evidence to his thesis. How do we reconcile both the descriptions of violence as well as episodes of peaceful coexistence that occur in the Ghaznavid sources? In other words, while Flood's evidence certainly dismantles the communalist explanation of Ghaznavid invasions, we still need to understand the violence that accompanied it.

I would argue that the entire range of behaviour (including violence) recorded in Ghaznavid sources are political acts. Here, I am using the term “political” in a technical way, following Carl Schmitt, to mean the agonistic competition among proponents of opposing visions for how the polity should be organised.Footnote ¹⁰ In other words, I would like to suggest that both the descriptions of Ghaznavid holy wars as well as their peaceful “encounters” with South Asian culture constitute elements of a single pattern of sovereignty which was not systematic. The resulting archive would necessarily be messy, reflecting the attempt to reach different audiences in different ways. However, it is precisely through trying to understand the whole structure with all its disjuncture that we can get a complete picture of the nature of the early Islamic military-political enterprise in South Asia. I will also argue that the figure of the elephant best exemplifies the rift in the overall structure of Ghaznavid sovereignty and serves as the entry point into its complex and contradictory world. The elephant is essentially an exceptional case. While Muslim authors generally advertised the Islamic credentials of Ghaznavid rule, they often had a difficult time with the un-Islamic or even anti-Islamic symbolism of the elephant. Yet the elephant continued to be not just as a military tool but also a symbolic one. In short it must have had importance in another value-system that was obviously not Islamic.

Of course, this is not the first book to analyze the role of elephants in Medieval Indian history. Specifically Simon Digby argued back in 1971 that the endurance of the Delhi sultanate depended on the success of Muslim rulers in controlling the supply of elephants and horses and in depriving their Mongol and Hindu rivals of them.Footnote ¹¹ Digby's pioneering work remains influential to the present day. However, Digby did not consider the symbolic significance of elephants for South Asian monarchy. Nor did he wonder how the incorporation of elephants for the first time by Muslim rulers might lead to tensions and difficulties. This was because at the time, Digby's conceptualisation of the place of the sultanate in South Asia history was not to different from Elliot's—meaning, it saw the sultanate as a forced imposition whose relationship specifically with the non-Muslim potentates it controlled was mainly militaristic. Ironically, Digby was actually trying to move away from “nationalist” explanations of the rise and endurance of the sultanate, opting instead for a “technological interpretation”.Footnote ¹² However, with over four decades of subsequent research at our disposal, we can once again revisit the topic from the perspective of culture without falling into older ideological camps. We can ask how Delhi's imperial project might have appealed to, engaged with, and incorporated the symbolic economy of rule as it was understood by subordinated rajas?

Since the publication of Digby's book, a rather vast literature on animal histories has burgeoned, focusing on the role of human animal interactions. Alan Mikhail's recent “unleashing the beast” provides a good bibliography for these studies.Footnote ¹³ With regards to the elephant, perhaps the best-known history is Mark Elvin's The Retreat of the Elephants. Here the author shows how deforestation and irrigation projects for agriculture destroyed the elephant's habitat and brought its widespread demise in pre-modern China. More closely related to the present article's concerns is W. G. Clarence-Smith's “Elephants, horses, and the coming of Islam to Northern Sumatra”, where the author argues that the decline of elephants in Sumarta was in part caused by cultural factors associated with the spread of Islam—namely, the preference for the horse by Muslims from Arabia and Iran in all sorts of contexts involving sports, diet, trade, transport, and so on. Finally, Thomas R. Trautmann in his Elephants and Kings: An Environmental History (2015) draws on Digby and Elmore to argue that a significant reason for the survival of the elephant in India but not in China was due to royal interest in South Asia. The capture, training, and maintenance of the war elephant took place in antiquity and coincided with the institution of kingship. However, like Digby, Trautmann also distances himself from the religious and symbolic meanings of the animal. When Trautmann writes for example that the religious priorities of his sources “may not coincide with the royal perspective I hope to capture”,Footnote ¹⁴ he downplays the fact that kings had to take into account a variety of constituencies among the population, and that the values and viewpoints of some of these people were partially defined by religion.

Thus, the present work will not pursue the main thrusts of the approaches listed above. Instead it will build on the contributions of Digby and Trautmann and fill the chronological gap between them while looking to another strand of animal studies that focuses more on the political significances that are assigned to beasts in human society. The inspirations for this study are the works of cultural historians Robert Darnton in his classic The Great Cat Massacre, Michel Pastoureau in his The Bear: History of a Fallen King, and the philosopher Jacques Derrida's The Beast and the Sovereign.Footnote ¹⁵ Darnton's books is useful as it draws on animal folktales or jokes as well as reports of bizarre behavior such as the massacre of cats in pre-revolutionary France in order to reconstruct the symbolic world of ordinary people. Darnton is less interested about the animals themselves. Rather he investigates how humans used actual or fabulous bestiaries to make sense of their world. Darnton's method is particularly important as it shows how by focusing precisely on the strange or superficially inexplicable episodes in our sources we can open up vistas into a world that has been hidden from the historian's purview.

In addition to Darnton, Michel Pastoureau's The Bear has also served as an exemplar for the present book. Pastoureau's argument is useful as he delves into the world of pagan northern Europe and shows how the bear was revered as a semi-divine king of the forest by ancient Slavs and Germans. However, during the process of Christianisation, medieval Church fathers launched an ideological campaign against the animal and branded it with all manners of vice and defects. The formation of two mutually exclusive and antagonistic systems of sovereignty around the bear is particularly useful for the study of the elephant in medieval India. A similar dichotomy of meaning existed in South Asia since the animal potentially stood for opposite symbolic connotations of sovereignty in Islam and Hinduism. Needless to say, the position of neither cultural tradition can be merely assumed or generalised and the range of available meanings for each in the eleventh century will be analyzed below.

Finally, the present book draws inspiration from Derrida's lectures published under the title of The Beast and the Sovereign. Here Derrida focuses on animalistic representation of the sovereign in European literature, especially where the sovereign is represented as a wolf, since both the beast and the sovereign occupied a parallel relationship to law. The sovereign, by virtue of being “above the law”, was like the wolf who was an outlaw and hence “outside of the law”. Derrida continues to tease out further connotations to this enigmatic comparison. He points out that, unlike the wolf, the sovereign was not just outside of human society but also inside of it, and even more disturbingly, in charge of protecting it. As such, Derrida argues, the sovereign was both human and wolf and yet neither. The sovereign was a werewolf who existed on the border between human and animal and as such threatened the very human society he was meant to protect. The danger of the sovereign as human-wolf then further enhanced his position and power through fear.Footnote ¹⁶ This thought-provoking formulation has inspired the present study of early Indo-Islamic elephantine sovereignty as it has helped to highlight the powerful and dangerous liminality of early Ghaznavid sultans who straddled the two conflicting symbolic significances of the elephant in Hinduism and Islam—one divine, the other satanic. The sultan was both a champion and a peril for his subjects—be they Muslims or non-Muslims.

Up to this point in this introduction, I have used the word “sovereignty” without defining exactly what I mean by it. Instead, I have alluded to it in the works of Derrida and Schmitt cited above. I find Derrida and especially Schmitt useful in so far as they investigate the notions of sovereignty as the power that stands above the law or prior to it especially during exceptional cases of crises.Footnote ¹⁷ I do this because I would rather not impose a modern definition on the eleventh century, but rather work with modern notions that deal with issues relevant to the political theory of the eleventh century. For instance, Muslim political theorists of this period described the rise of the Ghaznavids very much as belonging to legally exceptional cases. The best example of this is the Baghdad jurist al-Mawardi (d. 1058) who took on the political challenges of his age in his Ordinances of Government (al-Aḥkām al-Sulṭāniyah). While detailing the legal nature of the leadership of the Islamic community, as led by the Abbasid Caliphs, Mawardi knew that for at least two hundred years the actual governance of outlying provinces had been conducted by local power groups who had essentially seized it by force. Earlier jurists and statecraft theorists could turn a blind eye to such situations and simply ratify them by asking the caliph and the conqueror to acknowledge each other in mutual symbolic recognition. However, in Mawardi's time the military seizure of power had actually extended to Iraq itself, the very seat of the caliphate. So, in a section entitled “imārat al-istīla”, [rule by seizure] Mawardi tried to clothe this practice in a legal garb, and finding himself unable to do so, had argued that the legal requirements of legitimacy could be disregarded in cases of emergency when there existed a danger to public welfare.Footnote ¹⁸ This is indeed a very similar position to the ones analyzed by Schmitt and others. We are dealing with the tradition of political theory during times of crisis.

But if legality could be sidestepped, then the “imārat al-istīla” was open to creative definition. And so, precisely in the same period we find the expression of the desire to reunite the sundered religious and political functions in the person of the amir. For example we find eleventh-century chroniclers such as al-‘Utbi and Ibn Miskawayh writing for both the newly established and rival Ghaznavid and Buyid states quoting the old Persian dictum that “religion and kingship are twins”.Footnote ¹⁹ Other authors patronised by the Ghaznavids, such as al-Tha‘alibi, credited the sultan for protecting the world as well as religion, and praised the prophets Adam, Joseph, David, Solomon as well as Alexander the Great and the Prophet Muhammad for combining kingship [mulk] with prophet-hood [nabuwwah].Footnote ²⁰ In short, we find the extra-legal warrior-king desiring legal and religious sanction as well as claiming some kind of divine status. The same desire could find fertile ground in the non-Muslim part of his empire as well.

Ghaznavid sovereignty in India also ran into obstacles that only left it room as an exceptional case. In the post-Gupta period, Indian politics had witnessed the rise of numerous dynasties founded by lineages of fairly obscure origins. This situation led to a number of ways of establishing legitimacy with cooperation from priestly groups or Brahmins. One of the most common practices of legitimation involved the invention of warrior-aristocratic (Kshatrya) genealogies in accordance with Brahmanical ideals of social stratifications.Footnote ²¹ Another common way of establishing legitimacy entailed the patronage of temples for local deities by upstart monarchs. Through such patronage, the king could be identified with the deity or at least act as its agent.Footnote ²² Needless to say, neither option was available to the Ghaznavids. Mahmud of Ghazna would not (or could not) claim kshatrya ancestry. Moreover, Mahmud's patronage of religious sites was bestowed on Islamic ones and not “Hindu” temples. In short, if not “extra-legal” the Ghaznavids were certainly outside Brahmanical and political norms of kingship in north India. However, we do know that Mahmud did indeed make himself known to his South Asian subjects as a “king” [nṛpatī] in Sanskrit terminology on coins issued in the devanagri script. His assumption of other royal prerogatives (such as the elephant) was consistent with this self-presentation. I will argue all this in detail below.

3

Why did the Ghaznavids begin using the elephant?Footnote ²³ Initially, they gained possession of elephant herds in large numbers through their victories in the subcontinent, and they certainly put them to use for military or logistical purposes. The animal fulfilled several functions for the Ghaznavids. For instance elephants could be used to significantly improve transportation. Mahmud's soldiers knew about the animal's ability to swim, and could use it to great effect for crossing rivers in their campaigns in South Asia.Footnote ²⁴ Moreover much of the wealth gained from Indian temples was packed in elephant boxes and carried back to Ghazni.Footnote ²⁵ Yet, simultaneously, this functional use of the beast was very soon sublimated into rituals of sovereignty when early Ghaznavid sultans began distributing gold as booty to their soldiers by having the elephants arrayed near the royal tent while carrying boxes out of which the monarch would make the handouts.Footnote ²⁶ The logistical value of the elephant was combined with ritualistic effect.

Naturally, the Ghaznavids also employed the elephant in battle. The particular military techniques were adopted from India. They learned for instance about the use of a large number of elephants in the center of an army formation early on.Footnote ²⁷ These and other elephants could be mounted by soldiers who would sit atop the animal and throw down daggers and short javelins (shill).Footnote ²⁸ We also have evidence of the animal put to work for breaking down the gates of a fortress. From the late eleventh century we hear of elephants used for a charge on the enemy.Footnote ²⁹ Other military functions can be deduced from what the animal was outfitted with—a suit of armour and possibly metal covering for the tusk. ‘Unsuri for instance compares a line of Mahmud's elephants in their golden armour to mountains covered in blooming saffron of the crocus flower.Footnote ³⁰ Elsewhere he likens them to mobile metal fortresses—basically a weapon with characteristics similar to the tank.Footnote ³¹ Other poets refer to the tusk of elephants completely covered in iron (andar āhan giriftah sartāsar).Footnote ³²

Such practical uses are not surprising at all. Not only did the Ghaznavids learn about them through observation, but they also employed Indian elephant-keepers and Indian soldiersFootnote ³³ who clearly brought a culture of the elephant with them into Ghaznavid service. We also possess a list containing the names of Sultan Mahmud's elephants in a panegyric by Farrukhi. Even though the Indian vernacular originals have come down to us in irreversibly corrupted forms, nevertheless one can still recognises traces of Indo-Aryan roots in them. These include sur “hero”, jall “shiny”, suli “iron spike”, lol “corpulent”, mal “wrestler”.Footnote ³⁴ On another occasion, Sultan Mahmud took his armoured elephants north to Balkh on the shores of the Oxus River in 1024 to confront his Central Asian opponent Ali Tekin. There, the sultan ordered the riders to sound at once their wind and percussion instruments many of which were prerogatives of commanders and kings in medieval South Asia.Footnote ³⁵ These included the percussive thali and bhori and the snail-shaped sankh.Footnote ³⁶ The resultant pandemonium intimidated the foe and prevented war. In short, the Ghaznavid elephant technology was entirely Indian, and the symbolic use of the elephant developed in part out of practical usage.

The combined practical and symbolic use of the animal continued in later years. For example, during Sultan Mas‘ud's invasion of Gurgan on March 21^st, 1035, two elephants were brought along the Ghaznavid army for an assault on a bridge in a mountainous area. One female, was mounted by the sultan in the centre, surrounded by infantry, and the other, male “with a good name [i.e. reputation] and more warlike” (nāmītar va jangītar) up front. When the battle broke out, the male elephant in the front was wounded by spear-thrusts and arrow-shots. He turned around at a gallop, trampling the soldiers, and headed straight for the sultan's mount. Fortunately for the sultan, his driver managed to lead the wounded creature into a meadow, otherwise, writes Mas‘ud in a letter dispatched from the battle, “whenever a male elephant turns back with such injuries, it will stop at nothing”.Footnote ³⁷ This example of course shows that the elephant was a difficult animal to control. Yet, we can also detect a pattern by which the elephant reinforces the sovereign's position. After all, Mas‘ud's mount is not intended for battle.

Indeed, we possess sporadic evidence that suggests that Ghaznavid monarchy had become tied to elephants. On the day of his arrival in Ghazni and coronation (Sunday October 5^th, 1030), Mas‘ud sat on an elephant's litter and rode out of city into the Shahbahar meadow where celebrations were held.Footnote ³⁸ This was an emphatic act since, as the historian Bayhaqi says, Mas‘ud “had become so grand that he always hunted on the elephant's back”. Bayhaqi saw Mas‘ud in India hunting lions by throwing javelins at the prey from atop the armoured elephant—a re-enactment of battle activity.Footnote ³⁹ More significantly, on another occasion Mas‘ud rode to the hunting grounds on elephant back, changed to a horse to kill the lion, and then climbed on the elephant to bring the carcass back.Footnote ⁴⁰

Mas‘ud's brother, rival, and predecessor Muhammad was also associated with elephants as symbols of sovereignty. Supposedly, their father Mahmud had sent Muhammad a throne and an elephant, signifying to all his choice of heir and successor.Footnote ⁴¹ Indeed, mounting the elephant was only allowed with royal permission. When Bayhaqi describes the rebellious behaviour of a commander, Amir Bektughdi, he lists among his offenses not only murders and looting but “sitting on an elephant without [royal] permission (farmān)”.Footnote ⁴² Soon the Ghaznavid monarch and the elephant became symbolically inseparable. The world was seen as an elephant and the king as the elephanteer. Recalling the death of Sultan Mahmud, Bayhaqi writes, “when Amir Mahmud passed away, the elephant lost his elephanteer”.Footnote ⁴³ Later authors extended the metaphor, combining the image of the conception of Ghaznavid kingship as marriage as well as elephanteering. Mas‘ud-i Sa‘d refers to prince Sayf al-Dawlah Mahmud's elephant thus: “Upon being joined to the king of the world, it joyfully decorated its body with gold like a bride”.Footnote ⁴⁴

In essence, the elephant was incorporated into the Ghaznavid army for practical reasons (transportation, hunting, war), and then progressively for symbolising Ghaznavid sovereignty too. The reasons for this development were perhaps natural. The sheer size of the animal, its power, its exotic quality all would lend themselves to their exploitation as the symbolism of rule. The Ghaznavids need not have “borrowed” these ideas from a particular Indic source, but they would certainly be aware of its relevance in South Asia. We can infer this awareness from the court culture of the post-Gupta period in South Asia. Texts produced after the seventh century, as well as famous classical Sanskrit texts from antiquity give us a sense of what north Indian courts should have known about the connection of the elephant and sovereignty.

For example, in the Law Code of Manu, we already see the symbolic attachment of the elephant to royalty, where for example kings were required to punish certain offence by having an elephant trample the offender to death.Footnote ⁴⁵ Also, the animal was securely part of the paraphernalia of royalty and its theft was equated with burglarising a temple or royal treasury and subject to capital punishment.Footnote ⁴⁶ In the Arthashastra, the symbolic importance of the elephant is even greater. Here the author Kauṭilya equated the process of being trampled to death by an intoxicated elephant to “participating in the bath at the conclusion of a horse sacrifice”.Footnote ⁴⁷ Also, the life of the beast was so precious that even superintendents of the forest were expected to mete out executions to elephant-killers.Footnote ⁴⁸ Perhaps the most telling signifier of all was the fact that Kauṭilya was already using the metaphor of the animal for monarchy, for instance stating that a violent king is like an intoxicated elephant that cannot be controlled and should therefore be fought off with the power of a “rival elephant”.Footnote ⁴⁹ Outside of statecraft or legal compositions, we find better developed symbolism of elephantine royalty. The epic Mahabharata for example contains some of the most memorable portrayals of this phenomenon when it refers to a king who “when he rides out on his elephant of state, Brahmins perambulate the sacred fire around his person”.Footnote ⁵⁰ Here, obviously, the king and his elephant have morphed into the embodiment of divine rule and are worshipped by the priestly class. This sentiment is reinforced with the more explicit verse “priests were worshipping the king, who was seated on an elephant”.Footnote ⁵¹

Texts of the post classical period continued, if not intensified, the ideals of elephantine sovereignty. The post-Gupta text Vishnudharmottara-Purana (circa 600 ce), which was known to al-Biruni, begins with the praise of Ganesha and compares the gods to “royal elephants”.Footnote ⁵² It describes a battle between the gods and the demons in which “very mighty Indra, the king of gods with thousand eyes sitting on an elephant attacked Anuhlada”.Footnote ⁵³ But just as in antiquity the best and most vivid description of the elephant, especially as associated with sovereign rites comes from the epic literature of the early medieval period such as the seventh century Harshacharita. Here, the author Baṇa masterfully describes the spectacle of elephant hordes before the royal palace thus:

We find in this useful passage several descriptions of what we later encounter among the Ghaznavids as well. The elephant serves as symbol of royal and military power. It is acquired through tribute or as present. It is decorated with fabrics as well as musical instruments, precisely the same percussion and wind instruments later used by the Ghaznavids.

Baṇa also provides evidence for the identification of the royal figure with the elephant. For example he refers to the king's favourite elephant Darpashata as “his external heart, his very self in another birth”.Footnote ⁵⁵ Darpashata's position vis-à-vis other elephants is likewise identified with the monarch providing land-grants to his subjected lords, certified through metal inscriptions. He writes, “[Darpashata] seemed to be busy, night and day, bestowing all the woods on the other elephants by royal inscriptions of donation, made vocal with the swarms of bees which hummed round them as if reading them, and scratched by the half-cut branches of red sugar-cane as by a graver, and which seemed by a mistake to stay on hand instead of being dispersed abroad”.Footnote ⁵⁶ Or again, Darpashata's tail was compared with the royal emblem of elephant sovereignty passed down from father to son: “His tail waved like some fly-flapper marked with the signs of elephant-empire and handed down through a long succession of hereditary rulers”.Footnote ⁵⁷ Here again the elephant clearly resembles the dynastic king and his pedigree. The Harshacharita has brought us up to the seventh century. For the subsequent periods, we have the evidence of the tenth-century Deccani Jain text the Yashastilaka by Somadeva. Here too we find emphasis laid on the use of elephants for royal coronation, and their decoration by, among other things, armoring their tusks with metal.Footnote ⁵⁸ Somadeva also often identifies the king as the elephant, as when he states that a king, like an elephant, cannot come under the influence of another.Footnote ⁵⁹

All in all, we can state with a fair degree of certainty that elephantine sovereignty was quite well-understood (if not prevalent) in many South Asian courts around the time of Mahmud's invasion. We also know that Ghaznavids incorporated the animal in both practical and symbolic ways from early on in their reign. Whether or not it was initially intended as such, there is no doubt that such practices would make perfect sense among their north Indian subjugated lords. It would of course follow that the relationship of the Ghaznavid court with Hindu India was not simply one of predatory violence, but imperial over-lordship.

4

There is much evidence that suggests that the Ghaznavids were building an imperial polity in India. This is a fact generally overlooked or downplayed by most modern scholars.Footnote ⁶⁰ However, one can detect an awareness of this phenomenon in a number of places. For instance, in his 1963 The Ghaznavids: Their Empire in Afghanistan and Eastern Iran, C. E. Bosworth had decided that India was not the place to look for Ghaznavid state-building. Yet, on three occasions in his book, Bosworth made statements that belied that argument. On one occasion, while discussing the administrative system of the Ghaznavids, Bosworth wrote that in India, the sultans had opted “to retain the existing provincial patterns and to change them as little as possible”.Footnote ⁶¹ He also pointed out to the fact that both Mahmud and his father Sebüktegin struck coins for Indian circulation, including bilingual Arabic/Sanskrit coins. While Bosworth admitted that “it is difficult to estimate what the fabric of the Ghaznavid empire in its full florescence owed to local foundations”, he nevertheless suspected that “if the highly-developed superstructure was Persian and Samanid in inspiration, there must nevertheless have been a certain super-stratum of local practices to build upon”.Footnote ⁶² In short, Bosworth clearly suspected that the evidence of striking coins in India for a Sanskrit-reading audience necessarily implied some form of state-building activity that was rooted in the local traditions of South Asia.

Later on in the book, while discussing the trouncing of the Ghaznavids by the Central Asian Seljuks in 1044, Bosworth wrote that the defeated Sultan Mas‘ud “resolved to leave for India, but gave his officials full permission to stay behind in Ghazna and make terms with the incomers, in whose service they could expect to find employment”.Footnote ⁶³ This short sentence might seem rather unimportant on first glance. However, a good deal of political significance is implied here. Bosworth had argued that the Ghaznavids essentially believed that the affairs of rulers and subjects were strictly separate, and subjects had no right of political decision-making when it came to obeying the ruling class—even if that ruling class had just occupied their city from the forces of their current sultan. All this is of course very much in line with the right of conquests elaborated in the notion of “imarāt al-istīla” discussed above. But the statement above is striking as it depicts Sultan Mas‘ud expecting the same relationship in India as well. In other words, Mas‘ud's statements show that the Ghaznavids believed their victories over non-Muslim South Asian potentates endowed them [the sultans] with the rights of sovereignty in India.

Bosworth's other comments in this regard, support this argument as they show actual administrative experimentation by the Ghaznavids in parts of the subcontinent. Bosworth explicitly stated that Sultan Mahmud had attempted to set up permanent administration at least in the Punjab. The military governorship of Lahore was given to the slave-general Ahmad Intaltegin while a civil administration was entrusted to the highly capable Qazi Abu al-Hasan Shirazi. Bosworth's translation of Sultan Mas‘ud's words during the confirmation of the Qazi as well as Bosworth's concluding remarks are worth quoting in detail. He wrote:

This passage shows clearly that Bosworth knew very well that the Ghaznavid Empire extended far beyond Iran and Afghanistan and well into the subcontinent. At least for the Punjab, where most of the population as well as the ruling elite (ṭhākur) were non-Muslims a regular civil-military administration was envisioned. The use of the word kharāj for tribute is also significant as it implies a change in status of the subject population which was obviously no longer thought to be in the dār al-ḥarb of “abode of war”—an area outside of the rule of Islam and thus subject to raids. However, as we saw above, the Ghaznavids had understood their Indian holdings to be more extensive. The areas further east of the Punjab were not directly administered but were held in tributary and, one might say, feudatory status that required symbolic acknowledgement of the sultan in a relationship of alliance and mutual protection. Here, the symbols of sovereignty held paramount importance.Footnote ⁶⁵

We know from the evidence that Flood has so perceptively identifiedFootnote ⁶⁶ that the Ghaznavids adopted and replicated the methods of rule from other Islamic dynasties such as the Samanids of Central Asia or Abbasid caliphs of Baghdad. For instance, just as the caliph had sent Mahmud a robe to symbolise his “incorporation” into the Islamic body politic, so too had Sultan Mahmud gifted a robe to the Raja of Kalinjar for the same purpose.Footnote ⁶⁷ Also, in following an old Indian tradition that symbolised subjugation, Mahmud had demanded a clipping from the thumb of the Raja as well.Footnote ⁶⁸ Both these symbolic acts show the attempt by the Ghaznavid to create long-term political relationships with South Asian monarchs, even those who were not directly incorporated into the administrative machine of Ghazna.

Similarly, Carl Ernst has earlier discussed the significance of a bilingual coin issued by Mahmud of Ghazna, with the Arabic statement of faith “There is no God but God, and Muhammad is his prophet” on one side, along with a Sanskrit “translation” of this phrase on the reverse. Here the prophet Muhamamd is identified as an “avatar” and Mahmud's position as “master of men” (nṛpatī).Footnote ⁶⁹ This too is another sign that the Ghaznavids viewed non-Muslim South Asians as subjects in one form or another who were meant to understand and recognise the Ghaznavid ruler's authority and self-perception in a language with which they were already familiar.

Indeed there are many more such fragmentary traces of an imperial project in South Asia. This imperial project would have been understood in terms already familiar to people both in the Ghaznavid homelands of Eastern Iran and Afghanistan as well north India. Through the recent vogue of translation and compositions of historical texts, the Ghaznavids would have learned about the Sasanian notion of the emperor as a “king of kings” (shāhanshāh) where the Sasanian “Emperor” stood at the top of a royal hierarchy that included lesser kings.Footnote ⁷⁰ The same pattern of course existed in the Indian imperial legacy of the Guptas (contemporary of the Sasanians) and their successor states in which the supreme ruler was the great king of other kings (mahārājādhirāja).Footnote ⁷¹

In other words, the Ghaznavid ruler would function as an emperor—i.e. a monarch in charge of numerous lesser monarchs who in turn entered into feudatory relationship with him. While there certainly existed knowledge of religious difference, the political culture was still similar enough, and Mahmud and his subjugated underlings, be they Muslim or not, would have understood their role within the political network into which they were incorporated. We can trace this model all over the area subject to Ghaznavid rule—whether in the Islamic world or in India.

So, for example, when Mahmud set out for one of his campaigns in South Asia, he wrote a letter requesting the local ruler of Gharchistan in Afghanistan to accompany him. When the latter refused, Mahmud ordered the chamberlain Altuntash, Arslan Jazib (the governor of Tus), and al-Mani‘i (the governor of Marw al-Rud) to attack him.Footnote ⁷² On the other hand, when in the year 1009, Mahmud set out for the conquest of Narayanpur, he asked for assistance from the Ziyarid ruler Dara in northern Iran who sent him two thousand of his soldiers.Footnote ⁷³ Both episodes occurred in the sources immediately after a description of how each prince had sworn allegiance to the sultan in Ghazna. Significantly, the chronicler who mentions these events uses the local titles of each subjugated lord. The ruler of Gharchistan is referred to with his old royal title of “shār” while Dara is called the “chief” (ra’īs).Footnote ⁷⁴

Similar practices can be observed among the subjugated non-Muslim rulers in India too. One of the earliest instances of this exercise is somewhat ambiguous but goes all the way back to the time of Mahmud's father Sebüktegin. We are told that during a battle with the Hindu Shahi king Jaipal (Jayapala Deva, 946–1002),Footnote ⁷⁵ Sebüktegin consented to making peace with his Indian adversary “on account of the mercy he felt towards those who were his allied lords [awlyā’ihī]”.Footnote ⁷⁶ It is not clear whether these “allied lords” are the lords in the Ghaznavid army, or the group amongst whom Jaipal is also counted. We have much clearer examples within a decade or so. A good case in point was the ruler of Narayanpur who entered into tributary submission and trade alliances with Mahmud.Footnote ⁷⁷

Another example is the Pratihara ruler of Kanauj in northwest India, Rajyapal (960-1018), who had been reappointed in his position after submitting to the Ghaznavids. When Rajyapal was attacked by another South Asian monarch named Nanda/Ganda (Vidyadhara), who killed the Ghaznavid ally and replaced him with his own candidate Trochipal, Mahmud predictably led his army there out of obligation in order to avenge the transgression. His army successfully routed Trochipal at Bari on the Ganges River, then invaded Nanda's strongholds in Gwalior and Kalinjar, and eventually negotiated a submission. In the ceremony conducted the Ghaznavid army issued a diploma [manshūr] for the governorship [imārat] of fifteen forts to Nanda, and also gave him a royal robe [khil‘at].Footnote ⁷⁸ Again, Flood is right to note that Mahmud was essentially replicating a similar ritual to that which was commonly performed in the Islamic world and was enacted for him by the Abbasid Caliph al-Qadir Bi Allah twice (in July of 1008 and in November of 1026).Footnote ⁷⁹ But this is not merely an act of cross-cultural exchange. The political context shows the intricate web of feudatory alliances in which all the actors above were involved.

We also possess evidence of such tributary monarchs fulfilling their part of the political bargain with the Ghaznavids. For instance, al-Biruni writes that the Hindu Shahi monarch Anandapala (1002-1010)Footnote ⁸⁰ once offered his military services to Mahmud during a northern invasion by Central Asian Turks, because Anandapala did not want the monarch who had conquered him to be defeated by another.Footnote ⁸¹ This is an interesting case because it shows the complicated set of political ideas at work that went far beyond issues of mere subjugation or coexistence. Al-Biruni specifically states that Anadapala had the greatest hatred toward his Muslim overlord. Yet, he operated in a shared relationship of power with the Ghaznavids that had to be maintained and sustained for their mutual interests. There were numerous Indian precedents for this understanding of sovereignty, as Sanskrit inscriptions dating to the period of the latter Guptas and their successor states had clearly expected defeated kings who subsequently had been reinstated by their victor to acknowledge their new status by paying tribute and to render military service to their new overlord.Footnote ⁸² For instance, a surviving letter from a Chaulukya emperor, styled mahārājadhirāja (“great king of kings”) to a lesser king (rānā), instructs the princely subordinate to “proceed to the royal army camp. . .along with your horses, foot soldiers, various kinds of armor, war elephants, and all other material”.Footnote ⁸³ The question then arises, how did the Ghaznavids try to appear as “emperors” to these subject kings?

Again, Flood has correctly noted that the exact terminology used for Nanda's agreed-upon tribute, jizya (tribute that significantly included elephants), suggests that the Ghaznavids were equating Hinduism too with a religion of the book under the protection of Islam.Footnote ⁸⁴ A supporting case occurs after the battle of Narayanpur when Anandapala asked for a ceasefire and promised a tribute, which again, Ghaznavid chroniclers designate as the jizya.Footnote ⁸⁵ However, there is more to this phenomenon than merely evidence for “negotiating sociopolitical relationships”.Footnote ⁸⁶ In fact this was part of a wider effort by the Ghaznavids to elevate Hindusim to the status of a monotheistic religion as a convenient and religiously correct way of incorporating Hindu rulers and polities into the Ghaznavid Empire. We find more evidence of it in the works al-Biruni.

The best example of this occurs in the second chapter of the Indica. Here al-Biruni argued that the religious beliefs of the common people in India were simply abominable. However, he continued, similar “errors” could be observed among common Muslims too. On the other hand, he further stated, educated Hindus essentially held monotheistic beliefs in an abstract god. Here, significantly, al-Biruni uses the word “Allah” to refer to the ultimate deity of Hinduism, whereas otherwise he employs the more generic term of ma‘būd or “object of worship”.Footnote ⁸⁷ In short, the use of Muslim terms such as jizya or Allah for comparable South Asian phenomenon, as well as their network of political alliances, show that the Ghaznavids had attempted to form an imperial state in India, geared in part towards non-Muslim kings, which involved creating a network of tributary relationships.

The adaptation of the elephant as the symbol of Ghaznavid monarchy would make sense along other apparatuses such as the robe (khil‘at) and the jizya. As the emperor of Hindu monarchs, Mahmud's intentional or accidental use of the elephant would reinforce his positions among that particular target audience. In the absence of extensive Ghaznavid Sanskrit records, we must now turn to the Perso-Arabic chronicles and poems, read them both along and against their grain, in order to test and trace the formation of Ghaznavid discourse of sovereignty and its relationship to elephants.

5

It is not difficult to see that the use of the elephant in the making of Ghaznavid imperial image failed to gain general approval. The Ghaznavids did not only rule over subordinated Indian princes. Rather, a significant portion of the people under their rule, as well as the individuals who staffed their administration, were Muslims who lived in present day Afghanistan, eastern Iran, and Central Asia. It was also the perception of a Muslim population they had to worry about as they began to show themselves as elephant kings. How did Ghaznavid sources chronicle the gradual assimilation and the attending problems of elephants? Here I would propose using a strictly historical methodology in reading the Islamic chronicles, with special sensitivity to chronological sequence and context. Thus, there emerges a consistent pattern involving the admission of these animals into the Ghaznavid army and subsequent assaults on temples.

The first mention of Ghaznavid elephants in Muslim sources occurs shortly after the first battle of Emir Sebüktegin (Mahmud's father) against the Hindu Shahi king Jaipal. We are told that in the 980’s, Sebüktegin began raiding the frontiers of India as a holy warrior (ghāzī and mujāhid), capturing forts and treasures. Soon thereafter, in 986/7, Jaipal retaliated by invading Sebüketgin's domain in present-day Afghanistan. Sebüketgin defeated Jaipal however, and in exchange for peace, demanded a million silver coins and fifty elephants. But when Jaipal returned home, he violated the peace treaty by withholding payment, and Sebüktegin attacked again. However, this time, a new element was added, the first recorded assault on temples: “he [Sebüktegin] destroyed”, our source ‘Utbi, states, “idol houses, and established in their place the symbols of Islam”.Footnote ⁸⁸ The destruction was perhaps intended as a more severe punishment violence escalated, but the added veneer of Islamisation raises a question: why did the description of the initial raids not involve attacks on religious buildings? One should note here that between the two raids, there are fifty elephants. This could be a mere coincidence, but further reading suggests otherwise.

The next Ghaznavid invasion was not until two decades later. Sebüktegin's son Sultan Mahmud had been waiting for a pretext to attack the territories of Sistan (in present day Iran and Afghanistan) on account of earlier disagreements between his father Sebüktegin and the Saffarid ruler of Sistan Emir Khalaf. The opportunity was provided when Khalaf stationed his son Tahir in Quhistan and Bushanj (eastern Iran and western Afghanistan), and Mahmud's uncle Bughrachuq in Herat claimed these towns as part of his domains. A battle broke out between them in which Tahir managed to unhorse and decapitate the sultan's uncle who, so claims ‘Utbi, had showed up to the fray roaring drunk. So in December of 999 when Mahmud set out to lead his first ghaza in India as king (he had gone there previously under his father's command), he besieged Khalaf in an impregnable fort and accepted the latter's apology and a hefty payment as retribution. According to Gardizi, Mahmud then set out for the subcontinent where he captured many forts. No mention is made of temples or elephants.Footnote ⁸⁹

The following November he was back. This time he challenged Jaipal again in Waihind (Udabhaṇḍapura).Footnote ⁹⁰ ‘Utbi lavishly describes the battle in which Mahmud had fifteen thousand men, but Jaipal had twelve thousand cavalry, thirty thousand infantry, and significantly, three hundred elephants.Footnote ⁹¹ Mahmud prevailed, received a huge sum as reparation, and this time, asked for fifty elephants to be handed to him from those present on the battlefield. From there Mahmud went further to Waihind itself to do ghaza and jihad and plundered it.Footnote ⁹² Again, nothing here is said of “temples”.

Two years later (September 1003), the Ghaznavid army invaded Sistan again, and this time they brought with them their new Indian war elephants. Here ‘Utbi's narrative soars to truly fabulous heights. He writes that, Mahmud released the head elephant who rammed the gate of the outer fortresses and killed the guards there. The Sistani soldiers fled to the inner citadel. Emir Khalaf came out of his keep to survey the situation and behold, “He saw the universe awash in demon soldiers surging like waves upon satanic steeds” (wa ra’à tamawwuj al-faḍā’ bi-‘afārīt al-anjād ‘alà shayāṭīn al-jayād).Footnote ⁹³ Why are Ghaznavid soldiers and their horses compared to demons by ‘Utbi? This is a small but crucial abnormality in the ostensible mood of the passage. We will return to it shortly.

But to continue with the narrative, right before the emir's eyes wide with horror, the chief elephant of the Ghaznavids lifted up a Sistani rider from his saddle, tossed him up in the air the height of two lances, and then tore him in half with his tusk as he fell down. The terrible beast then proceeded to knock down the gate of the inner citadel as well, trampling more guards along the way. Khalaf naturally sued for peace and was exiled to Juzjan where he later died a prisoner.

The invasion of Sistan serves as a watershed in the history of Ghaznavid ghazas. It is after this event that Mahmud goes on new Indian raids all specifically involving a veneer of Islamisation, though at this point still without any harm done to religious structures. In October 1004, Bhatia (Bhatinda) is attacked. The sultan battles Baji Rai (?), defeats him, and this time takes one hundred and twenty elephants. No temples are destroyed but he appoints people there to teach Islam.Footnote ⁹⁴ In Spring 1006, Mahmud goes to Multan to fight the “apostates” [Ismailis]. As a result, we are told, his reputation as a helper of religion crosses the ocean and reaches Fatimid Egypt.Footnote ⁹⁵ Bhimnagar was also conquered again in the name of Islam. Thirty elephants were gained and great riches were won but again no assaults on temples were made.Footnote ⁹⁶ This brings the number of elephants in the Ghaznavid army up to two hundred by the end of 1008/beginning of 1009.

How was the sultan perceived at this point in the Islamic world? After all, he claimed to have waged war in the name of Islam and had brought in enormous treasures and elephants from the subcontinent. At the same time, he had helped bring about the downfall of many a Muslim noble house in the east as well and had certainly persecuted members of other sects of Islam. As it turns out, while Mahmud had inspired awe in the hearts of his subjects in the eastern Islamic worlds, he also evoked a peculiar horror and revulsion that cast him precisely in the opposite manner in which he wished to be seen: not as a champion of Islam but as satanic figure, perhaps even the dajjāl—the antichrist of the apocalypse! In addition to the allusion by ‘Utbi, another evidence for this incredible reversal occurs in one of the many panegyrics that were composed for the sultan. I will use this poem as an entryway into discovering the various meanings of the elephant to the Muslim intellectuals of the early eleventh century.

6

The author of this poem was Badi‘ al-Zaman al-Hamadhani (968-1008), a famous littérateur and stylist of the Arabic language, a man of wit and irony.Footnote ⁹⁷ He compared the sultan, in a manner that would become quite common among other poets of the period, to ancient pre-Islamic heroes, be they legendary ancient Iranian kings or Islamic/Biblical prophets. The historian ‘Utbi, writing his book decades later, quotes the poem just after the time of the Ghaznavid invasion of Sistan, and this would suggest a composition date of around 1003. However, Hamadhani specifically mentions two hundred elephants in the army, and this would date the poem to time of the sultan's campaigns in Multan and Bhimanagar shortly before the author's death. The text is rich with historical allusions of a contradictory quality and with apocalyptic overtones—a fitting reaction perhaps to the first sighting of proboscidean power in full display. While on the surface the sultan is praised, almost every reference evokes an association of sultan with the devil. It runs as follows:

Modern scholars have long argued that this poem is about genealogy, that of the Turkic Ghaznavids (Khaqan) dominating over Iranian Samanids or Buyids (Bahram, Saman, Shah)—a strange choice really, even in Hamadhani's lifetime, since the Ghaznavids themselves were eager to downplay their “Turkic” origins.Footnote ⁹⁹ On the other hand, there is obviously much more going on in these lines and the elephant plays a significant role.

The first line sets the tone. Why does the author ask God to increase his faith? Why has it diminished? Afridun (Faridun) and Alexander were certainly worthy heroes to compare the sultan with, but why is this done in the interrogative mode? “Is this Afridun wearing his crown?” he asks. Who else could be? Did anyone else wear Afridun's crown? A look at the stories regarding this old Iranian mythical king clarifies the ambiguity. On the one hand, Afridun/Faridun was a great king who was supposedly the first monarch to sit on an elephant and use it in war.Footnote ¹⁰⁰ In this, Mahmud would resemble him easily. However, Faridun's kingship began on shaky grounds. According to the various versions of the legend available in the eastern lands of the Islamic world in this period, before Faridun's birth, the foreign dragon-king Zahhak usurped the throne of Jamshid (the rightful monarch) and unleashed a reign of terror and oppression in the earth. Faridun, who was a descendant of Jamshid had been spared certain death when his pregnant mother had escaped to a secret and remote region to give birth to him and raise him. When a rebellion finally arose during Zahhak's absence in India, Faridun returned and eventually killed Zahhak. “On the same day”, writes the tenth-century historian Bal‘ami, the leader of the rebels “placed the crown on [Faridun's] head and gave him the world”.Footnote ¹⁰¹ Tabari's adds the extra detail that Zahhak was the devil “al-khābil” and that the Magians believed Zahhak was not killed but was taken to Mount Damavand and tied up there where he continues to be alive while being guarded by the jinn.Footnote ¹⁰² Firdawsi also calls Zahhak a demon (dīv), or the devil (ahrīman). He presents Faridun as a descendant of not Jamshid but his father Tahmuras and says that Zahhak's campaign in India was meant to enforce black magic (jādū). Finally Firdawsi credits the angel Surush (Gabriel) for instructing Faridun to bind Zahhak in the mountains.Footnote ¹⁰³ The anonymous author of the Tarikh-i Sistan (eleventh- century) adds that Zahhak is the antichrist (dajjāl) and will break out of his bonds on doomsday.Footnote ¹⁰⁴ Al-Tha‘alibi (also eleventh-century) confirms the prevalence of this belief when he says, “according to the legends and nonsensical superstitions of the Magians, Zahhak is still alive in the fort of Damavand, and like Satan awaits the resurrection (maḥshar)”.Footnote ¹⁰⁵ Thus to return to Hamadhani's question, “Is this Afridun wearing his crown?” The answer could be yes, or it could be no, in which case it must be the tyrannical demonic usurper, the antichrist. But if so, is the author implying the end of the world has arrived?

This brings us to the second historical reference. The question is “is this a second Alexander?” Why a second? The comparison between Alexander and Mahmud was on the surface perfectly acceptable. Other men of the pen such as the historian Bayhaqi and the poet Farrukhi continued to compare the sultan to the Macedonian hero throughout the rest of the Ghaznavid period, albeit in order to show the superiority of their patron.Footnote ¹⁰⁶ But Hamadhani's comparison has a complication that is brought to surface further down the poem where he says, “The Gog and the Magog of the [Ghaznavid] army surge like waves”. This is an odd sentence and recalls ‘Utbi's line above wherein Ghaznavid soldiers were compared to surging waves of demons riding on demonic horses. The Gog and the Magog in the Koran are depicted as a seething monstrous mass of evildoers on earth. Alexander was the person who travelled to the far north and built a dam to contain their destructive deluge until the apocalypse. Here, in a perverse inversion, the sultan is portrayed as an Alexander who actually leads the surging Gog and Magog into battle. The interrogative mode gains an extra significance here. Is this a second Alexander? Then why is he leading the demonic savages indeed? The release of the Gog and the Magog is the second implication of an impending apocalypse, and the sultan, whether as Zahhak or a false Alexander is beginning to look ominous.Footnote ¹⁰⁷

The third reference is to Solomon, and here too a double meaning can be found. The old prophet, according to the Perso-Arabic literature of the period, was just, loved “ghaza” (or holy war, exactly how the Ghaznavids described their wars in India), and had control over demons (which is what the elephants are compared to in lines 6–7). But then, there were other stories according to which Satan actually impersonated Solomon and ruled in his place for forty days. The details are as follows: Once upon a time, Solomon invaded the land of idolatry, slew the king who reigned there and took the king's daughter to wife forcing her to convert to Islam. All of this would resonate well with the way Ghaznavids depicted their invasions of South Asia. However, the narrative has a twist. Solomon's pagan wife secretly continued to practice her old religion in the prophet's house for forty days until she was discovered by her husband who destroyed the idol and begged God for forgiveness. But God decided to punish his erring servant in order to teach him a lesson.

Solomon, so runs the narrative, owned a signet ring that he would leave with his wife (or son in other versions) whenever he went to the toilet. One day, one of the Satans over whom the prophet had mastery, assumed Solomon's figure while the prophet was in the loo and took the ring from his wife (or son). When Solomon came out, no one would believe that it was really him. So Satan ruled in the prophet's place for forty days. Al-Hamadhani plays on the ambiguity of this case of identity-theft when he asks “has the return brought Solomon back to us?” Again, the answer could be yes, but it could also be no, and in this case too Mahmud would indeed be not Solomon but Satan impersonating the prophet.Footnote ¹⁰⁸

The next reference is highly problematic, if not immediately apocalyptic. In the pre-Islamic Iranian stories, in two instances does a hero named Bahram confront an adversary named Khaqan—one Bahram V (Gur) the Sasanian king (d. 438) of whose descent the rival dynasty of the Ghaznavids, the Buyids, had boasted, and the other Bahram Chubin (d. 591), whom the Samanids, but also the last Saffarids of Sistan,Footnote ¹⁰⁹ had claimed as their ancestor. The first Bahram had met in battle and beaten his contemporary Khaqan (ruler of Turks). He was also a brave warrior who had gone to India and fought with an elephant.Footnote ¹¹⁰ The second Bahram was another hero of the Sasanian age. He was certainly killed by the Khaqan, though this was a tragic death of a champion. According to Bal‘ami, “Bahram took sanctuary with the Khaqan. The Khaqan accepted him and treated him well. He showed great feats of manliness in Turkestan. However Parviz (Bahram's enemy) devised many a ruse until they finally killed Bahram in Turkestan”.Footnote ¹¹¹ By comparing Mahmud to Khaqan (something no other Ghaznavid writer did) and proclaiming his victory over the house of Bahram (alluding to the Buyids, Samanids, or Saffarids), the poet actually creates an uncomfortable reversal of the Iranian material. To anyone familiar with the stories, the victory of Khaqan's son would imply a violation of the heroic ideals with which the original stories were infused, and no praise can be gained by Sultan Mahmud through the comparison.Footnote ¹¹²

Another reference is also quite cryptic. The poet mentions that the elephants were covered in coats of mail and were renowned for their colours. This may reflect the actual condition of Ghaznavid elephants as noted above. However the story evokes a historical episode regarding the rebellion of Babak (d. 838). Babak was a local king from northern Iran who revolted against the authority of Abbasid caliphs. Using the inaccessible mountain ranges of his homeland to keep out Muslim armies, Babak taught a new religion rooted in pagan and Zoroastrian traditions. It was finally during the reign of the caliph Mu‘tasim that Babak was defeated, captured, and brought to Baghdad where he was gruesomely executed. According to Tabari, the caliph had asked his advisors how to make an example of Babak, and they had advised him to have the defeated foe be taken to the execution grounds on a painted elephant. The historian Tabari quoted a poet who had described the scene in these words, “The elephant is painted in the manner that is customary for it. It is carrying the Satan of Khurasan. The elephant's body-parts are not painted, except on very special occasions”.Footnote ¹¹³ In short, again the image of the painted Ghaznavid elephants would evoke equally majestic as well as cryptically demonic associations.

There exists a final allusion in the poem that is also couched in mystery. Yet this allusion is perhaps the most important, because in many ways it provides the key to the whole order of apocalyptic references with double meaning. The poet says, “Whenever he rides his elephant into battle or on the parade ground, your eyes behold a sultan on the back of a Satan”. By the merit of his mounting the monstrous beast, Sultan Mahmud is seen in possession of the supernatural power of riding a Satan. What hero is being evoked here?

A king who rides a devil in the ancient Iranian epic is Tahmuras the demon-catcher “dīv-band”. The primordial Tahmuras from the mythical age was said to have thrown the saddle on a devil's back and ridden him around the world.Footnote ¹¹⁴ According to Abu Mansur al-Tha‘alibi, in his book of history entitled Ghurar al-Siyar (composed circa 1020 and dedicated to Emir Nasr—Mahmud's brother and the governor of Khurasan):Footnote ¹¹⁵ “The Persians represent [Tahmuras] in their books, palaces, and sculptures as riding the devil. And a certain poet has likened a king who used to ride elephants to him [i.e. Tahmuras]”.Footnote ¹¹⁶ Al-Tha‘alibi then quotes the following poem:

Like the anonymous poet in the Ghurar, al-Hamadhani's image of Mahmud mounting an elephant was also meant to evoke the image of Tahmuras who had saddled and ridden the devil.

Yet, the identification of Mahmud's heroic prototype raises a number of unexpected problems. If Mahmud is indeed being likened to Tahmuras why is the mythical king not named at all by al-Hamadhani? He certainly had no difficulty specifying the other heroes. In spite of his relative popularity (based on al-Tha‘alibi's report), Tahmuras is still one of the least-mentioned legendary figure in poetry or prose. Why choose him as a model over so many other better-known ancient heroes? Moreover, we know that almost every historical reference in the poem has a negative double meaning that depicts the sultan as a villain, even the Archenemy himself. Therefore the reference to Tahmuras too should have its Satanic double. Al-Hamadhani must have chosen this vague figure of the pre-Islamic heroic material and have avoided naming him explicitly in order to allow for the image of Mahmud on elephant/Satan to allude to another historical figure, this time an antihero. Who was this villain? The best-known villain who was known as the elephant-master in Islamic literature was Abraha, the Yemeni king.

The anonymous author of the eleventh-century Tarikh-i Sistan asserts that in the ninth century, the ancestor of Emir Khalaf (defeated and killed by Mahmud), the famous Ya‘qub al-Layth had been presented with the opportunity to incorporate elephants into his army long before the Ghaznavids.Footnote ¹¹⁹ The great hero however, so alleges the author, refused to employ the beasts because they were associated with the Koranic tale of Abraha's expedition against Mecca—the failed attempt by the ruler of Yemen to raze the Ka‘aba (House of God) to the ground just before the birth of the prophet Muhammad in 540 ce. Bosworth has argued that the story is probably fictional, but what matters is that this anecdote was circulating in Sistan roughly around the time al-Hamadhani composed his poem.Footnote ¹²⁰

One could find the original tale in the Koran as well as various popular books of the time such as the Arabic history of al-Tabari, the Persian translation attributed to Bal‘amī, and of course the Sira (biography of the prophet Muhammad). The basic narrative is as follows. In the year 570 the Christian king of Yemen Abraha marched to Mecca to destroy the Ka‘ba, taking with him droves of war elephants. However, upon reaching the shrine, God sent forth swarms of birds called “Ababil” who bombarded the hostile army with small pebbles that decimated the aggressors whose crushed bodies on the field resembled chewed grass. The event was so momentous that a chapter of the Koran was devoted to it, and that year was remembered as the year of the elephant.

The details of the story are much more elaborate in the later historical and exegetical literature. One such detail, which is particularly relevant here, is the name of the chief elephant that led the pack, for indeed it had a name. What was the elephant called? Sure enough, as Tabari writes, “The name of his elephant was Mahmud”!Footnote ¹²¹ Bal‘ami expands on this, “the name of that elephant was Mahmud, and they had never taken it to a battle in which it had not been victorious, and it had never turned its back on anything except after victory. It was a huge elephant”.Footnote ¹²²

In short, al-Hamadhani's allusion to the elephant king may have evoked Tahmuras in the eastern Islamic world, but it most certainly have reminded Muslim readers of the Yemeni king who was inimical to God and his house in Mecca. Moreover, the similarity between the name of Abraha's elephant and the Ghaznavid sultan would cement the damning connection. Based on the subtle allusions by Ghaznavid authors ‘Utbi and al-Hamadhani, as well as the explicit mention in the Tarikh-i Sistan, we can safely assume that other people in Khurasan and Sistan also understood the potential negative connotation of Ghaznavid sultans and their use of the elephant. This would indeed create something of a crisis of sovereignty for the Ghaznavids. The use of the animals, which was necessary for political and military reasons would recall in the minds of some of the Muslim population the image of the enemy of Islam, the arrogant Abraha who tried to destroy the House of God.

The tainted association never went away. One can detect how problematic it was to contemporary and later Indo-Persian writers. The Ghaznavid historian Gardizi (eleventh century) excluded the elephant's awkward name, Mahmud, but could not deny its presence during the attempted assault on the House of God and its subsequent destruction by rock-pelting birds. He wrote, “God Most High appointed the abābīl birds to them, and the birds struck them with mud rocks, that is sijjīl, and whatever animal was struck by that rock was killed until they were defeated”.Footnote ¹²³ A later historian Juzjani (thirteenth century) completely absolved the elephant from the damning action of the destruction of the Ka‘aba. He wrote, “[Abraha] had an elephant named Mahmud. He rode it at the head of the army until they reached the shrine. The elephants all stopped there and turned around. They did not at all trespass the sacred precincts of the shrine. Then God Most High sent from the sea the Ababil birds, and all of [Abraha's] army was annihilated so completely that no trace of them remained”.Footnote ¹²⁴ Juzjani's version is departure from the previous and subsequent narratives of the Koranic story as it exonerates the elephants, especially the elephant Mahmud. Finally Firishtah (early seventeenth century) makes the problem explicit by citing the following anecdote, which although probably fictitious, laid bare what others had tried to cover up. He wrote:

This story is obviously much later than the events it describes and probably a legend, but it expresses explicitly what is implicit in the poems of Mahmud's contemporaries such as al-Hamadhani. By possessing the elephant, Mahmud suffers a problem in his sovereignty. His power is a challenge to the supposed sovereign, the caliph, and to the political understanding of rule in general as it equates the possession of elephants with satanic pride. The Ghaznavids were faced with a dilemma because the temptation to employ elephants as often and as ostentatiously as possible was simply overwhelming. It provided them with superior military technology and helped bolster their image as the emperor of Hindu kings. But simultaneously it cast them in an unfavourable light among Muslims. For as we saw above the immediate association made in the minds of Muslims would not be between the elephant and sovereignty but the animal and anti-Islamic arrogance.

How did the Ghaznavids solve this dilemma? The Ghaznavid court engaged in a reversal of its own. The sultan was not Abraha the destroyer of the Ka‘ba, but the prophet Muhammad himself, the man who purged idols from the Ka‘ba. The self-presentation of the Ghaznavids as temple destroyers in part goes back to this. It was important to show the Islamic world that Mahmud was the breaker of pagan temples not Islamic ones. We can in fact trace this development in the narratives of the invasions of India. As stated above, the reports of temple desecration in India enter the various accounts only after the elephants appear on the scene.

7

It is thus after 1008, the probable year of Hamadhani's damning panegyric that equated Sultan Mahmud with the antichrist and Abraha, that sultan's ghazas take on a destructive nature. The subsequent attacks on Narayanpur (late 1009), Nardin (or Nandana, 1014), Thanesar (late in 1014), Mathura, and Kannauj (1018) all end in acts of vandalism.Footnote ¹²⁶ The purpose may have indeed been to present Mahmud as the protector of Sunni Caliphate as contrasted with the Shiite Buyids and Fatimids that had treated the Abbasids with such open disrespect. Yet, following the chronology of events strictly, the problem of the Ghaznavids and the role of elephants in their public self-presentation plays a crucial role. Each campaign brings in more elephants but each elephant requires a disassociation from Satanic power. This is accomplished by claims of temple destruction. The attack on the temple of Somnath in western India is probably the best example of this. It was the last major campaign of the sultan. Here the connection is made explicitly between Mahmud and old iconoclasts like the Prophets Abraham and Muhammad who cleansed the Ka‘ba of idols. The poet Farrukhi, who accompanied the army on this journey punned on the name Somnath and the pagan Arabian idol named Manat to emphasis this fiction. He writes:

The sultan as a second Muhammad indeed! The poets could not get enough of this. Later poets continued to draw on this association. For example, the panegyrist Manuchihri felt the need to equate the sultan atop the elephant with the prophet Muhammad—a fact belying precisely the problematic association of the Ghaznavid monarch with Abraha or the antichrist. He wrote about Mahmud's successor, his son Sultan Mas‘ud, “You came back to Khurasan on the back of the elephant just as the prophet Muhammad came to heaven on the back of the Buraq”.Footnote ¹²⁹ The allusion is to the nightly journey of the prophet Muhammad to heaven, hell, and purgatory on the back of a horse-like fantastical beast. Elsewhere, the poet stated: “The people raise up the cry of ṣalawāt, when your howdah appears in the distance”.Footnote ¹³⁰ The ṣalawāt is a statement of praise used for the prophet (and sometimes his family). The implication here of course is that the Ghaznavid was not a demonic or pagan figure but the prophet Muhammad, and that his ride was not the beast of the apocalypse or the destroyer of God's house but the prophet's mount.

There was of course a second solution to this problem. What could the later Ghaznavids do when their military powers waned and there were no more temples within reach to destroy? One could argue that the elephant was not demonic, or only superficially so. This solution was offered by Sufis. So for instance, the poet Mas‘ud-i Sa‘d Salman describes the elephant thus: “[The elephant] is fearless (jilf) and somewhat fierce. It drinks wine and listens to music [samā‘-niyūsh]”.Footnote ¹³¹ The association between the elephant and the Sufis is achieved by its state of madness (mast, literally intoxication) as well as its association with music that leads to ecstasy, which by the twelfth century had become a common mode of practice among Muslim mystics. The musicality of the elephant perhaps derives from the practice (mentioned above) in which the animal would be dressed in chain mail, and be made to carry on its back percussion and wind instruments that would be blown specifically in order to create disorientation among the enemy or guest.Footnote ¹³² The human observer in this psychological state is often described by our sources as being madhūsh (losing senses) or mutaḥayyir (dumbstruck).Footnote ¹³³ Thus, while visibly demonic, the beast is in fact connected to God in a mystical way. We are here in the environment of malāmatī Sufsm which was esotericist and encouraged public censure and reversal of superficial values.

Elsewhere the poet writes about Malik Arsalan Ghaznavid and his elephants: “They are all drunk and dance to the music and song of the musician [surūd va samā‘]. They are demons on the day of battle. Demons have been punished by their harm. Four hundred efrits are lined up. When they gallop they fly as winged birds. To what king belongs such an amazing feat? A world of demons obeys [our king]. He sits in fulfillment as Solomon”.Footnote ¹³⁴ Here the comparison with Solomon as the king who tamed demons returns, but now in a mystical way. The animal is only superficially frightening or repulsive, but through its intoxication by wine and music serves its divinely inspired master. In this way, a parallel could be drawn between elephant and Sufi dervishes. We know specifically of such men in the late Ghaznavid milieu who would dress in animal hides, act as if intoxicated, engaged in ritual listening to music in the pursuit of union with God.Footnote ¹³⁵ But in Sufi thought, the seeker and the sought are one. Was the elephant ever understood as divine? Eventually it was.

This brings us to Mas‘ud-i Sa‘d's contemporary Sana'i, the court panegyrist in Ghazna who turned Sufi and through his writings greatly influenced later Persian poets such as Farid al-Din ‘Attar and Jalal al-Din Rumi. In his massive Hadiqat al-Haqiqa, dedicated to Bahramshah Ghaznavi, Sana'i used the elephant as an allegorical metaphor for God. In this famous and oft-cited story, itself of Indian provenance and already occurring in early medieval Buddhist and Jain texts such as the Udana,Footnote ¹³⁶ Sana'i recounted the narrative of a great monarch who camped in the vicinity of Ghor near a town where all the inhabitants were blind. Desiring to know about the king's famous elephant, the townspeople sent representatives who arrived at the encampment and began feeling up the elephant with their hands. As each one grabbed a different part of the animal, he derived a totally incorrect impression, and upon his return, described it in completely different manner than his fellow envoys. Sana'i completes this brief anecdote with this line, “People do not understand divinity, wise men have no way into this matter”.Footnote ¹³⁷ The elephant then has come to stand for God at the Ghaznavid court by the twelfth century—a useful transformation that would help ease the problem of the monarch atop the animal that had tried to destroy the Ka‘ba. This was so because at the end of their rule, the option of presenting the sultan as the prophet Muhammad was not available to the descendants of Mahmud. Whereas their illustrious ancestor had actually led numerous destructive raids into India, the later Ghaznavids could boast of far fewer acts of idol-smashing, if any at all.

All in all, we can say with certainty that the Ghaznavid monarchy experienced a good deal of contradiction in creating the symbolic component to its rule. The fact that the Ghaznavid court persisted in trying to reconcile this contradiction shows the primary significance of the elephant. We can deduce that the positive association of the animal, which would have been confirmed in non-Muslim systems of thought, outweighed the negative. This would suggest that Ghaznavid rule in India was not simply the imposition of Islam on subjugated South Asian rulers. Quite the contrary, we can hypothesize that their rule was depended, if not primarily, then at least as much on the consensus of non-Muslim followers. As such it would be fitting to call the Ghaznavids in India a thoroughly Indian monarchy, which was self-consciously un-Islamic. However, in the Middle Eastern parts of their empire the Ghaznavids were equally an Islamic monarchy, and self-consciously so. While their attempts to appear so has been well recorded in the form of early Ghaznavid historiography, the other side of their rule has only left traces in these very sources. The present article has tried to extract and analyze these.