In accordance with Mughal practice, a deadly struggle between ‘Alamgir’s three surviving sons – A‘zam, Mu‘azzam and Kam Bakhsh – followed the emperor’s death. Prince A‘zam, at the time in the Deccan, at once declared his sovereignty and began marching north towards Agra. Prince Mu‘azzam, then posted near Peshawar on the north-west frontier, also declared his sovereignty and, adopting the title Bahadur Shah, rapidly marched south, beating A‘zam in the race to Agra. In June 1707 he defeated his younger brother, who died in battle near Agra. After spending nearly a year consolidating power in upper India, the new emperor marched to the Deccan to confront ‘Alamgir’s youngest son, Kam Bakhsh, who had also declared his sovereignty and established himself in Hyderabad. Badly outnumbered, however, he too was killed in battle, thereby completing another war of succession, albeit one far less bloody than the war that had raised ‘Alamgir to power.
The succession struggle following ‘Alamgir’s death differed vastly from that of 1657–9. Whereas the earlier war had been fought between powerful princes backed by substantial households and their allied nobles, the sons of ‘Alamgir possessed far smaller households and fewer resources. This owed in large part to their father’s policy of diverting resources from his sons to himself, for fear that one of them might overthrow him, as he had overthrown Shah Jahan. It was therefore the domineering ‘Alamgir, not his sons, who was the ultimate arbiter of the fate of their households.1 Two of his sons spent long periods in imprisonment, one was driven out of the country, and none had led any army in the Deccan after 1702. Some of the empire’s most powerful nobles simply sat out the struggle between them, defiantly refusing to support any prince. This reflected the quiet drift of real power from princely households to powerful nobles, some of whom felt sufficiently confident in their autonomy to bargain with the new emperor over administrative appointments. And, once in power, they were able to place limits on the emperor’s authority. For example, Zulfiqar Khan, the commander who had seized Raigarh from the Marathas and led the siege of Jinji, pressured Bahadur Shah into appointing him both governor of the Deccan and paymaster general (mir bakhshi). The latter position authorized him to approve all ranks and promotions, enabling him to thwart the emperor’s ability to enlist new nobles into the corps of imperial mansabdars.
By the time Bahadur Shah died in 1712, the institution of princely households had deteriorated even further. All four of his sons were adults at the time their father came to power and, since they all wished to be at the imperial court at the start of the next, inevitable succession struggle, they simply refused to serve in the provinces as governors. For the first time in Mughal history, all princes remained at court for the duration of their father’s reign. But doing so compromised their ability to establish independent power bases, build alliances, tap into provincial revenue streams or acquire the administrative experience that might enhance the likelihood of a vigorous and successful reign. Remaining at court for their entire princely lives meant they had much smaller households. Bahadur Shah’s eldest son and eventual successor, Jahandar, had only 300 cavalrymen, in contrast to the tens of thousands maintained by princes of the previous century.2
Consequently, when Jahandar succeeded to the throne in 1712, he did so mainly because of his lack of political, administrative or military experience or resources, which made him more pliable – and hence attractive – for Zulfiqar Khan, then the empire’s most powerful figure. It was he who engineered Jahandar’s accession by first forming a troika of Jahandar and two of the prince’s brothers against their more experienced and energetic sibling, ‘Azim al-Sham, promising that, after the latter was defeated, the empire would be partitioned among the trio. Then, after ‘Azim al-Sham had been defeated and killed, Zulfiqar Khan treacherously turned on Jahandar’s other two brothers, and when they were disposed of he escorted his royal patron to the Peacock Throne. Most of the followers of Jahandar’s three brothers were then imprisoned and their property confiscated, while several prominent nobles were publicly executed. It was the first time that supporters of the losers in a succession struggle were punished instead of being reintegrated into the ruling structure. From this point on, all Mughal emperors would be princes who had little or no experience outside the imperial harem. Clearly, the traditional Timurid mode of handling imperial successions had been corrupted beyond recognition. In the process, the emperor himself had become a virtual puppet of a powerful figure outside the ruling dynasty, in this case Zulfiqar Khan, now the imperial vazir.
The relative ease with which Jahandar had been placed on the throne tempted other power brokers to repeat the ploy by advancing the cause of some other prince. The cycle was soon set in motion by two brothers of a well-entrenched clan of Indian Muslims, the Barahas, who had their base in upper India between the Ganges and the Jamuna. These were the so-called Saiyid brothers – Saiyid ‘Abd Allah Khan, governor of Allahabad, and Saiyid Husain ‘Ali Khan, governor of Bihar. Both men owed their positions to the patronage of Prince ‘Azim al-Shan before the latter was killed in the coup that had brought his brother, Jahandar, to the throne. Within months, the ambitious brothers seized on a plan to lift themselves into supreme power by promoting the cause of Farrukh Siyar, the son of their slain patron. In late 1712 a rebel army led by the brothers marched from Bihar up the Ganges valley towards Agra, with Farrukh Siyar in tow. Up to this point Jahandar had been indolently celebrating his rise to power with his musician courtesan, Lal Kunwar, having left administrative affairs to Zulfiqar Khan. But when news arrived of Farrukh Siyar’s advancing army, an alarmed court moved frantically to mount a defence. Since the imperial army would not move without pay – and it had not been paid for nearly a year – the desperate court resorted to breaking its vessels of gold and silver, even stripping gold off the roofs of Delhi’s imperial palaces. In the confrontation that followed in early 1713, the rebel army routed the hastily organized and faction-ridden imperial army near Agra. Shortly after reaching Delhi, Farrukh Siyar was crowned and forthwith ordered the execution of both Jahandar and Zulfiqar Khan. ‘Abd Allah Khan was duly appointed vazir and Husain ‘Ali Khan paymaster general, the empire’s two highest offices.
Farrukh Siyar’s reign (1713–19) had hardly begun, however, before struggles broke out over whether the emperor or his highest ministers would have the final word in fundamental issues of state. Long-submerged tensions over the empire’s ethnic identity also rose to the surface. On one side were the Saiyid brothers, whose Baraha clan of Indian Muslims was as native to India as were Jats, Rajputs or Marathas. Towards all such communities the two brothers adopted conciliatory policies, abolishing discriminatory measures that ‘Alamgir had imposed, such as the jizya tax. On the other side was an equally powerful faction of nobles who claimed a Turkic or Iranian ancestry, had enjoyed Mughal patronage for nearly two centuries, and zealously promoted the dynasty’s original Timurid character. So poisonous had these factional struggles become that ‘Abd Allah Khan needed an escort of up to 4,000 cavalry when he moved through Delhi’s streets to attend daily court as vazir. When Ajit Singh, the leader of the Rathors of Jodhpur, refused to take up his appointed position as governor of Sind, Farrukh Siyar sent Husain ‘Ali Khan with an army to punish the Rajput mansabdar. But the emperor also communicated with Ajit Singh, secretly promising him rewards if he would kill Husain ‘Ali Khan. Instead, the two men became allies. Similarly, in the course of the Mughals’ long struggle with the Marathas, Farrukh Siyar sent secret communications to the Maratha leaders urging them to resist Husain ‘Ali Khan’s overtures to them. But, again, the two sides simply ignored the emperor, making mutually beneficial agreements that Farrukh Siyar refused to ratify, but which the Saiyid brothers and the Marathas honoured anyway.
Matters came to a head in late 1718 when Husain ‘Ali Khan arrived in Delhi from the Deccan with 25,000 cavalrymen, nearly half of them Marathas with whom he was allied. In February 1719, after a fiery verbal confrontation in Shahjahanabad’s audience hall between the emperor and the other Saiyid brother, the vazir ‘Abd Allah Khan, the latter seized control of the Red Fort while Farrukh Siyar, guarded by his armed female slaves, was holed up in the harem. In this political impasse, with armed supporters of both sides already fighting in the streets and a restive urban population on the verge of riot, the Saiyid brothers placed Rafi‘ al-Darjat, an unsuspecting grandson of Bahadur Shah, on the throne. They then sent their own men to the harem, where Farrukh Siyar was dragged out, blinded and later executed. When the tubercular Rafi‘ al-Darjat died several months later, the Saiyid brothers, now ruling the empire virtually by fiat, placed the deceased emperor’s brother, Rafi‘ al-Daulah, on the throne. Within weeks he, too, died of tuberculosis, at which point the brothers selected another of Bahadur Shah’s grandsons, Muhammad Shah, to be emperor.
Within the single year 1719, then, four emperors had occupied the Peacock Throne in rapid succession, a sure sign of structural instability. However, the dictatorial powers that the Saiyid brothers had assumed, which reduced the emperor to an impotent captive, were vigorously opposed by the empire’s older, Turko-Iranian nobility, the so-called khanazads. Their opposition focused not only on the outrage of executing a sitting emperor, Farrukh Siyar, but also on the gains that native Indian Muslims had made thanks to the Saiyid brothers’ patronage, which they resented. This old-guard opposition coalesced around the governor of Malwa, Chin Qilich Khan – entitled Nizam al-Mulk (d. 1748) – whom Muhammad Shah secretly implored to help release him from his captors, the Saiyid brothers. Help came in mid 1720, when one of the brothers, Husain ‘Ali Khan, was assassinated in a plot engineered by old-guard nobles bitterly opposed to the rise of the Baraha clan and to native Indian Muslims generally. The emperor now eagerly joined an effort to complete the coup by getting rid of the other Saiyid brother, ‘Abd Allah Khan, which was duly accomplished later that year.
The problems facing Muhammad Shah’s long reign (1719–48), however, were hardly resolved by the eclipse of the Saiyid brothers’ two-year incumbency. The first two decades of the eighteenth century saw a deepening of the jagirdari crisis that had begun late in ‘Alamgir’s reign. Successive vazirs irresponsibly doled out appointments and ranks to their political followers with no certainty that revenue-producing jagirs supporting those ranks could be found. Competition among nobles for such jagirs inevitably created tensions and factionalism within the nobility, furthering the divide between old-guard Iranian or Turkic nobles and newcomers such as Marathas and Indian Muslims. This posed a political problem. Insufficient land earmarked for jagirs also reduced mansabdars’ ability to generate the levels of revenue needed to pay their cavalry troops. And since soldiers would not take part in campaigns without regular pay, the mansabdars’ cavalry contingents fell below required levels. This posed a military problem. Finally, because an increasing number of mansabdars competed fiercely over a shrinking pool of revenue-producing jagirs, nobles with substantial land holdings resisted periodic transfers of their jagirs, a hallmark of the system since the days of Akbar. As a result, the jagirs of many such nobles became hereditary, undermining the entire idea of the Mughal nobility as a service class without permanent landed estates. This posed a systemic problem.
Further undermining Mughal administration, the regime gradually abandoned the zabt system of land taxation, whereby revenue assessments were based on measuring individual fields and periodically evaluating the quality of the land and the type of crops sown. In the early decades of the eighteenth century, such systematic management was increasingly overtaken by the practice of revenue farming, as the court simply sold the right of revenue collection to the highest bidders. After remitting the agreed amount of land tax to the government, many revenue farmers enriched themselves by squeezing further ‘taxes’ from cultivators. Not only did the government in this way lose control of its own revenue administration: such practices fuelled the rural discontent that was exploited by local zamindars, who stood between the Mughal state and revenue-producing village communities. These remnants of ancient royalty, long since pushed down in rural society by successive waves of conquerors, stubbornly clung to their traditional claims of local sovereignty, which they typically manifested by building small, fortified strongholds, maintaining armed retainers, adjudicating disputes and claiming a share of the land revenue. Ever since Akbar’s reign it had been Mughal policy to transform these stalwart local elites into state revenue collectors or functionaries, and, if such co-option proved impossible, to forcibly clear the landscape of ‘the weeds and rubbish of opposition’, as Akbar’s minister Abu’l-fazl contemptuously called recalcitrant zamindars.3 By the early eighteenth century, however, Mughal mansabdars had fewer resources to maintain troopers with which to confront unruly zamindars within their jagirs. Further tipping the balance of power in favour of the latter was the diffusion of muskets among the population at large, which considerably narrowed the military advantage that the state had previously enjoyed.4
The collapse of the Saiyid brothers’ rule in 1720 might have seemed a likely moment for the restoration of the old guard of Timurid khanazads, and possibly for a recovery of Mughal power and authority. In early 1722, in fact, the veteran noble Nizam al-Mulk – already the governor of the Deccan, Gujarat and Malwa – was summoned to Delhi and appointed imperial vazir. Having served ‘Alamgir with unswerving loyalty since 1688, this commander and administrator naturally evaluated Mughal rule of the 1720s in the light of his former master’s austere, disciplined administration. He accordingly dismissed Muhammad Shah as a negligent and weak-minded stooge completely under the influence of his favourite courtiers. Nizam al-Mulk was also aware of the extent to which long-established rules of governance had withered, and how deeply the entry of Marathas and Hindustanis into the nobility had undermined the prestige of his own class of Timurid khanazads. Accordingly, in 1723 he set before the emperor a set of sweeping reforms, which included ending the practice of revenue farming in khalisa lands, sacking all nobles ill-fit for the job, redistributing jagirs, abolishing khalisa lands that had been distributed as jagirs, and stamping out the menace of bribery. But Muhammad Shah’s favourites at court, fearing the loss of their influence should such proposals be enacted, poisoned the emperor’s mind against approving the reforms, which were quietly shelved.5
Despairing that the central administration could ever be restored to proper governance, and having been politically marginalized from the court’s inner circles since the death of ‘Alamgir, Nizam al-Mulk resolved to strike out to the Deccan, the part of India he knew best and where he had the deepest roots. Precipitating his departure was news that his own deputy governor of the Deccan, Mubariz Khan, had in his absence entrenched himself as a practically independent ruler.6 Departing Shahjahanabad in late 1723 ostensibly for a ‘change of air’, Nizam al-Mulk reached the Deccan and the next year defeated and killed in battle his rebellious deputy governor, after which Muhammad Shah bestowed on him the exalted title ‘Asaf Jah’. The huge block of territory that he carved out, essentially the whole of the Mughal Deccan, would become the state of Hyderabad, effectively launched in 1724 with Mubariz Khan’s defeat. After 1763, when its capital was moved from Aurangabad to Hyderabad, Nizam al-Mulk’s descendants would be known simply as the Nizams of Hyderabad, and their dynasty as ‘Asaf Jahi’ – both being titles of the state’s founder.
Although he had founded the Mughals’ first successor-state on imperial territory, Nizam al-Mulk never formally declared his independence from the Mughals, whom he continued to serve as viceroy of the Deccan. His own coins and those of his successors right up to 1858 continued to bear the emperor’s name. By retaining this fictive connection with the regime, the Deccan ‘province’ offered disaffected northerners the possibility of abandoning Delhi’s political chaos while technically continuing their service with the empire. The strategy appears to have worked. Viewing the Deccan as a robust imperial dependency, thousands of administrators, religious scholars, intellectuals, military men and master craftsmen migrated southwards during Nizam al-Mulk’s administration, as did, too, talented members of north India’s service castes, such as Kayastha accountants and Punjabi Khatri bankers. These northerners were then added to a political system that, under Nizam al-Mulk, had already absorbed a wide spectrum of the Deccan’s constituent communities. After 1728 his relations with the Maratha state stabilized to the extent that he was able to recruit Maratha chieftains, Brahmins and low-level revenue officials into his government.7 He also assimilated restive and potentially hostile Afghans as fort commanders and allowed more than a dozen Telugu kingdoms (samasthans) to subsist within his borders as tributary states, a policy reminiscent of Akbar’s accommodative policy towards Rajput states. Other indigenous groups – such as Tamils, Bhils and Gonds – were similarly brought into Nizam al-Mulk’s nascent Hyderabad state.8
Like Hyderabad, Bengal also became functionally independent in the 1720s, a process traceable to the career of another of ‘Alamgir’s officers. Kartalab Khan was a Brahmin slave who had been purchased in his youth, converted, brought up and tutored in the intricacies of clerical practices by ‘Alamgir’s imperial diwan, or chief financial officer. In 1698 Kartalab Khan entered Mughal service as the diwan of Hyderabad. Recognizing the young man’s exceptional talent for administration, in 1701 ‘Alamgir appointed him the diwan of Bengal, a rich but at the time poorly managed province. His mandate was very clear: enhance its revenue and remit its full share to the central government. Soon after reaching Bengal’s provincial capital of Dhaka, Kartalab Khan enacted sweeping reforms. First, he reclassified the delta’s jagir lands as khalisa ones and reassigned dispossessed jagirdars from Bengal to jagirs on undeveloped lands in neighbouring Orissa, which from 1703 was also placed under his management. As a result, unlike anywhere else in Mughal India, Bengal’s provincial government absorbed the whole of the delta’s agrarian wealth instead of having to share it with jagirdars. Second, he sent surveyors throughout the province to produce accurate figures of local yields. And third, the province’s many zamindars were made to pay the full amount of assessed revenue or face imprisonment or torture. Under such pressure, zamindars with smaller holdings were forced to sell their holdings to larger zamindars, with the result that by 1727 just fifteen enormous zamindari estates paid nearly half of the province’s land revenue. The diwan’s reforms also led to the emergence of wealthy banking houses that made enormous profits in interest on loans to zamindars unable to pay their stipulated revenue. In this way, by the 1730s the house of the Jagat Seths had become the largest banking firm in Bengal, if not in all India. In effect, these houses served as guarantors that the provincial government would receive its entire revenue demand.9
Kartalab Khan’s success in enhancing Bengal’s revenue proved a boon for the Mughals’ central government at a time when the imperial treasury was severely stressed by ‘Alamgir’s protracted Deccan campaigns. Beginning in 1702 and continuing for the next several decades, the diwan annually remitted ten million rupees from Bengal to the central government, which for some time comprised the only consistently reliable source of revenue for the whole empire. In 1703 he personally attended ‘Alamgir’s court in the Deccan, where the emperor awarded him the title Murshid Quli Khan in honour of the fiscal officer of the same title who had served Prince Aurangzeb when he governed the Deccan. ‘Alamgir also allowed the diwan to rename a city in northern Bengal after his new title – Murshidabad, which became the provincial capital in 1704. In 1717 he was made governor of Bengal in addition to diwan. Although this violated Mughal protocols prohibiting one person from occupying both offices, Murshid Quli Khan continued to remit Bengal’s large annual share of revenue to the central government until his death in 1727. Succeeding him as Bengal governor was his son-in-law, Shuja‘ al-Din Muhammad Khan (r. 1727–39), initiating a de facto regional dynasty. Yet Shuja‘, too, continued to send Bengal’s annual revenue to Delhi, in return for which he was allowed to do more or less what he pleased in his province.10 After his death in 1739, however, authorities in Bengal altogether ceased remitting its substantial revenues to the central government.11
In the 1720s and 1730s, still more Mughal provinces acquired de facto independence. In 1722 the emperor appointed Sa‘adat ‘Ali Khan governor of the central Gangetic province of Awadh, but within just three years he had usurped Delhi’s authority by removing imperial mansabdars from their allotted positions and reassigning jagirs within the province. The emperor was powerless to enforce his orders or punish the governor. Indeed, Sa‘adat ‘Ali Khan mobilized local communities around his own banner, entrenching his independent status. Because jagirs in Awadh were relatively small, jagirdars lacked substantial power bases, allowing the governor to exert greater authority over them. And, having no sons of his own, he named his nephew and son-in-law Safdar Jung to succeed him as governor of Awadh, thereby launching what amounted to another de facto regional dynasty. He also appointed as his financial officer a Punjabi Khatri who ran the region’s finances independently of Delhi’s supervision. He, too, passed that office to his son. Further west, the Punjab acquired de facto independence in 1726 when ‘Abd al-Samad Khan, who had been appointed governor in 1713, was succeeded by his son.12
The ebbing of Mughal power between 1707 and 1750 is seen in other ways, too. Already in Bahadur Shah’s reign (1707–12) serious gaps appeared between the assessed revenue and the amount actually collected. Powerful nobles resisted imperial orders to transfer their jagirs. Even the Mughal–Rajput alliance showed signs of fraying, as leading Rajput commanders dragged their feet on military campaigns, or abandoned such campaigns altogether. In the reigns of Jahandar and Farrukh Siyar, the practice of farming out revenue collection to the highest bidders, effectively bypassing the official revenue apparatus, became widespread. Jat uprisings in the area west of the Jamuna broke out with increasing frequency and intensity. Jat chieftains plundered trade caravans that moved along routes between Delhi and Agra, and from there east down the Ganges valley or south towards the Deccan. The most notorious and talented Jat leader, Churaman (d. 1721), briefly carved out his own independent state in the Agra region. The Mughals had to expend considerable resources attempting to suppress, or at least contain, such uprisings occurring so close to their imperial capitals.13 The state’s diminished profile is also reflected in the sharp drop in the treasury at Agra, from ninety million rupees in 1707 to just over ten million in 1720. For several more decades the empire continued to receive revenue from Bengal, the sole remaining prop to imperial finances. But, in 1740, it too ceased. By the time of Muhammad Shah’s death in 1748 the empire was totally bankrupt, while its former provinces were either practically independent or managed by outsiders from the south. These were the Marathas.