An eventful politics of difference and its afterlife: Chittagong frontier,Bengal,c. 1657–1757

Introduction

Published in 1889, ‘Abdul ‘Ali Islamabadi bin Minnat ‘Ali’s account of the city of Chittagong in southeastern Bengal stands as one of the earliest ‘native’ histories dedicated to the past of his hometown. Drafted during his tenure as a professor at the Hooghly College, Islamabadi’s text is exemplary of the Indo-Persian ta’rīḵẖ chronicle. It provides an account of the climate of the city, a history of its grandees and a record of its rulers through to the colonial era. Significantly, his narrative also has a specific point of departure. Neither an early discussion of hostilities between Mughals and Afghans over control of Chittagong, nor his references to the presence of the Portuguese in this key port city of the Bay of Bengal (ḵẖalīj-i Bangāla), fails to obscure that for the author, the historical foundations of the city are to be sought at a distinct moment. ‘In the year one thousand and seventy six of the Hijri [1666]’, offers Islamabadi, ‘at the command of His Majesty Sultan ‘Alamgir, the good administration of Nawab Shayista Khan brought the world-seizing key to the grasped lock of the Arakani Maghs and conquered [Chittagong] for the House of Islam (dār al-Islām)’. For Islamabadi, the Mughal annexation of the Burmese city signalled its origins. Thenceforth, Chittagong, or Mughal Islamabad, became the ‘extreme…eastern frontier’ of India. ¹

At first blush, Islamabadi’s conceptualisation of an Indian territorial ‘frontier’ (sarḥad) appears decisively ‘modern’. If his was a history of events from Aurangzeb’s reign (1658–1707), we might take that in effect, his reflections were cast under the aegis of colonial rule. Chittagong had been an administrative unit of the British imperial state since 1761, when with Burdwan and Midnapur, it made up the Company Bahadur’s three territorial acquisitions in the ‘bridgehead’ of Bengal. Ever since, it had been conceived as a frontier—political, natural and cultural—in eastern India. Flanked by the rolling Lushai Hills and Chittagong Hill Tracts, it was set apart from Burma by the Naf estuary. It was the site at which the agrarian cultures of Bengal began to cede ground to the ‘tribal’ societies of the Reangs, Chakmas and Kukis. And, it was where the Company’s military command, revenue collection regime and ability to acquire state intelligence reached their collective limits. ² Yet the modernity of Islamabadi’s claim is belied by the fact that he does not seek meaning for the frontier in his knowledge of British rule in Bengal. He does not look to the terrain, which with its seagoing links made Chittagong ‘so detached from the other Bengal Provinces’. ³ Nor is his conception of the frontier quite laced with anxieties that stalked the colonial state, administrative distresses that led it to imagine Chittagong as the ‘Mercia of Bengal’. ⁴

Numerous scholarly interventions have posited that the limits of modern India—and by extension, South Asia—were fixed by colonialism. ⁵ By contrast, the claim advanced by Islamabadi rests on a precolonial genealogy. His ‘frontier’ is grounded on a distinction between Mughal India and Arakan, northern Burma—the unspoken ‘House of War’ (dār al-ḥarb). It is a conception that bares a politics of difference with a specific past, and by the same token, its historical trajectory begins with the Mughal conquest of the city. In truth, Islamabadi was not alone in these assumptions. The frontier in Chittagong, and the normative values it underwrote, cast a long shadow. Colonial histories thus reckoned that before Aurangzeb’s general and governor Shayista Khan annexed the city and renamed it Islamabad, there was an ‘absence of the Faithful from Chittagong’. ⁶ It was as though the city were wrested into an ‘Islamic’ fold overnight, whereas a century after its conquest, the early colonial state would go to some lengths to uphold the ‘ancient constitutional’ legacy of the old regime. ⁷ No matter the peculiarity of opinions regarding its latterly fate, modern Indian and colonial accounts agreed that the crystallisation of the Chittagong frontier, if not its ideological genesis, came in wake of the Mughal imperial conquest in 1666.

Categorical Frontier?

What kind of a history gave way to such a durable category of a frontier? How did the lines of difference it maintained gain such ‘universal’ purchase? Articulated in the nineteenth century, Islamabadi’s history points to an issue that has received scarce attention in the literature, namely the political ideology of frontiers in early modern India. ⁸ Conventionally, questions of expanding military, fiscal and ecological lines dominate this historiography. Studies of Mughal frontiers privilege the longue durée as chronological norm, emphasise the role of army and cavalry in subduing autochthonous war-bands and chieftains, and cite natural history to chart Mughal ‘zones of power’. ⁹ In this respect, a most significant frontier lay in the Deccan and the Maratha regions. As the culmination of a longer history of expansion inaugurated under Akbar, the conquest of the Deccan sultanates in 1686–89 was therefore read precisely through these terms. Thus, the ‘crisis’ of the post-1700 Mughal state was itself understood as ‘failure’ to cope with these overlapping issues at that frontier—a failure to allocate jāgīr assignments due to military overstretch, ¹⁰ failure to integrate local warrior groups as the Bedars, ¹¹ and indeed, a later inability of the Mughals to successfully spar against Maratha forces due to the natural constraints of crop destruction, poor monsoons and disease in 1702–04 in the western Deccan. ¹²

Nevertheless, such questions remain indispensable to the study of frontiers. Akin to the Deccan, but perhaps absent in some of its teleologies, histories of Bengal have rightly measured the eastern frontier with the yardsticks of the long perspective, ecological change and to be sure, administrative history. The creeping movements of the Ganges in this silt delta, the clearing of jungly lands by revenue-generating agrarian settlers and the state-military lines pushed first by the Bengal Sultanate and then more vigorously by the Mughal state gave Bengal —and especially, eastern Bengal (bhāṭi)—its quintessential significance as an expanding frontier. ¹³ For the Mughals, the move to east really came to its own in the decades after 1610, when Dhaka, renamed Jahangirnagar, was chosen as provincial capital. ¹⁴ But in light of all this, the category of the Mughal frontier in Chittagong in the lower delta still seems unusual. As an abstraction, it was not contingent on slow-moving history, natural or otherwise. Its emergence as a form of thought, as we saw, was understood to have been rather more sudden. Furthermore, in themselves administrative histories do not explain the ‘civilisational’ distinctions that the frontier sought to enforce between Mughal India and Arakanese Burma.

Against this background, in this essay I suggest that the birth, consolidation and abiding afterlife of the Chittagong frontier may be traced back to a remarkable set of events from the 1660s. They began with the war of succession fought between Emperor Shah Jahan’s four sons in 1658, but came to a head in 1659–60 through the actions of a Mughal prince, Muhammad Shah Shuja‘ (1616–60?), who after two unsuccessful military bids for the throne, fled from Bengal to the Mrauk-U Kingdom in Arakan through Chittagong. Exactly six years after his mysterious death in Arakan, the Mughal military followed to conquer the port city from Arakan. I argue that it is with some conviction that we can trace the political– cultural language of a ‘frontier’ against Burma to these years. Moreover, I illuminate how subsequent traditions in eighteenth-century Bengal shored up this notion as linked to the memory of Shah Shuja‘. In part, this essay takes its inspiration from recent examinations of frontiers in early modern South Asia, efforts that have placed close textual study against broader horizons, and have asked searching questions regarding memory, forgetfulness and their ability to forge the boundaries of identity. ¹⁵ By considering the ‘biography’ of the Chittagong frontier through the last years of Prince Shuja‘ and his later memory, the study thus adds to a growing discussion on the place of small narratives within macrohistorical change in premodern India.

At another level, my inquiry is framed by what William Sewell has called eventful history. While the evidence is uneven, they are suggestive of how the events of the 1660s—in particular, Shah Shuja‘’s misadventures in Bengal and Arakan—marked a turning point in Mughal political thought. As a result, the Chittagong ‘frontier’ took shape from a highly visible, politically transformative and emotionally charged series of events. ¹⁶ It was accordingly to those same incidents that later legacies of the ideology deferred. Anchored in the history of events, such an approach is also used here to reflect on the theme of ‘Persianisation’, or the interface between Mughal and vernacular cultures. ¹⁷ In the main, Indo-Persian traditions have been understood as cultures that disseminated from Delhi or Agra to inspire ‘cosmopolitan’ dispositions in far-flung provinces. The making of the Chittagong frontier, however, tells another story. It reveals how Mughal ideas often relied on provincial ideas to assume universal forms. Equally, it shows that Indo-Persian culture could produce not just commensurability, but divisive attitudes that allowed some to ‘look…down upon’ others. ¹⁸ Such attitudes, I assert, were not trans-historical. They turned on specific histories, and as with the case of the frontier that surfaced in Shuja‘’s steps, eventful encounters. But before elaborating on this any further, I provide below a short excursus on Chittagong before the frontier.

Mughals and Maghs before Difference

Chittagong was by all measures a cosmopolitan city prior to the Mughal conquest. At marked odds with the nineteenth-century evaluations, the maritime city was a port of call for Bengali, Tamil, Telugu, Malay, Arab and Persian merchants, not to mention Dutch and Portuguese traders. ¹⁹ In fact, the Arakanese or Rakhine state that controlled Chittagong was itself a paragon of the early modern cross-cultural polity. A nominally Theravada Buddhist state with its capital in the city of Mrauk-U (Myohaung) down the eastern littoral of the Bay of Bengal, the kingdom was consolidated in c. 1430. ²⁰ Yet it reached its apogee concurrently with Mughal inroads into Bengal, and from the second half of the sixteenth century the state turned to taproots of commercial revenue, an expanding military market and uneven alliances made with the local Portuguese to achieve a relatively high degree of centralisation. Central to this process was also Arakan’s direct administration of Chittagong, which began in earnest from as early as 1550. ²¹ From their coastal perch, slaving expeditions into Bengal’s hinterland and mangrove outgrowths by Arakanese seafaring forces—known in Bengal as the ‘Magh’ peoples—also proved crucial to its political strength. ²² For us, most intriguing were the many ways in which this kingdom engaged the political cultures of South Asian Islam.

Though a Buddhist polity, the Mrauk-U dynasty decked itself with the trappings of the typical Indo-Islamic court. And though Arakanese in culture, it patronised a rich corpus of Bengali literature partial to Persian precedents. A charter myth of Mrauk-U held that during his exile, the founder King Narameikhla had sought refuge in Sultanate Bengal. ²³ With the assistance of Gaur, Narameikhla established the Mrauk-U dynasty and styled himself with the Burmese and Persian titles, respectively, of Min Saw Mon and Sulaiman Shah. If this was once quaintly viewed as the moment when Arakan turned from the ‘fairyland of the Glass Palace’ to the ‘extravaganza of the Thousand and One Night’, the truth is that these complicated trends continued for several centuries. ²⁴ Arakanese currency is evidence that the state had acquired the Indo-Islamic custom of minting coins (sikka) to assert its political authority. Silver money from Raja-kri’s reign (1593–1612) proudly inscribed his Burmese titles on one side, whereas the reverse heralded him in both Perso-Arabic and Bengali as ‘Lord of the White Elephant Salim Shah’ (ṣāḥib [a]l-fīl [al-]baiẓ mālik al-‘ādil Salīm Shāh sult̤ān; dhabal gajēśbar śrī śrī Salim Śāh). ²⁵ Numerous other kings of Mrauk-U held Persian titles. They included Ilyas Shah Sultan (r. 1523–25), ‘Ali Shah (r. 1525–31), Sikandar Shah (r. 1571–93) and Hussain Shah (r. 1612–22). ²⁶ In Chittagong, their governors also struck coins with aniconic Persian motifs. ²⁷

Still, all this did not preclude moments of friction between Arakan and the chief bearers of Persian culture in the region, the Mughal Empire. But as they wavered between unfamiliarity and at times antipathy, it is important to note that Mughal perceptions of this polity had, till the 1660s, stopped far short of addressing Arakan as a hostile state at its imperial frontiers. For instance, in the earliest phases of Mughal rule in Bengal, Abul Fazl had made plain mention of Chittagong (Chāṭgānu) as falling under Arakanese (Araḵẖang) rule, yet in more ambiguous terms he added that it came under the revenue administration of Akbar’s domains. Moreover, in Abul Fazl’s eyes, the Arakanese stood apart in that they were neither Hindu nor Muslim (bar ḵẖilāf-i hindū wa musulmān nishān dahand). ²⁸ Actually, such assessments were characteristic. In a slightly later account, one Tahir Muhammad wrote of the Arakanese as through a perspective of ‘exoticism’. For him, they almost resembled Hindus. Yet, here too the Mughal author stayed clear of a derogatory outlook on the Buddhist kingdom. ²⁹ Texts more firmly rooted in the political experiences of the Mughals in Bengal offered different opinions. At times, they bristled with the tensions of violent encounter. What they lacked again were any meaningful vocabulary that turned a dim recognition of alterity into an explicit political argument on otherness as such.

Arakanese slaving raids in Bengal had even led, in the early seventeenth century, to angry exchanges of letters between the Mughals and Mrauk-U, escalating to the point of two unsuccessful Mughal invasions of Chittagong in 1617 and 1621. As late as the 1630s, the Bengal governor Islam Khan threateningly informed Arakan that the Mughal state could, if it wished, ‘bring the whole territory of Chittagong and Rakhang under the hoof and feet of our horses and elephants’. ³⁰ Nevertheless, sporadic violence, however intense, was constitutive of established patterns. As localised affairs within Bengal, they failed to incite any clear transformations in Mughal imperial ideologies with respect to Chittagong or Arakan. Representative of this are also the writings of Mirza Nathan, who served as Mughal mīr baḥr—loosely, ‘Chief of flotilla’—after a more active period of state building in eastern India. Patrolling Bengal’s intricate waterways during the reign of Jahangir (1605–27), Mirza Nathan referred to Arakanese raids on garrisons close to Dhaka as scarcely more than the activities of local ‘rebels’. ³¹ On skirmishes with armed Luso-Arakanese vessels in Bhalwa, for instance, he wrote as if they were merely the activities of regional ‘enemies’, who required subjugation through ‘conquest’. ³²

Taking general stock, then, we might grasp the connections between the Mughals and northern Burma till the 1660s by taking relevant cues from the South Asian past, even if scholars of Arakan have stressed its sui generis history. ³³ In essence, the mix of acculturation and antagonism in the eastern Bay did not differ from ‘Hindu–Muslim’ encounters on the subcontinent. Besides material cultures, courtly ideologies also had their counterparts. Like the story of Narameikhla in Bengal, the legend of the founding of the Vijayanagara Empire related how the brothers Harihara and Bukka were uprooted from their homeland and taken to the heart of the Delhi Sultanate. Later, they similarly made a heroic return to establish what was once understood as a representative ‘counterpolity’ to Muslim rule in India. As Phillip Wagoner demonstrated, however, this narrative—together with the flowering of ‘Islamicate’ ornamental traditions at the Vijayanagara court —primarily reflected how ‘local’ polities engaged the ‘universal’ cultures of South Asian Islam. ³⁴ Nevertheless, to view Mughal–Arakanese engagements through this lens, or to misappropriate the phrase from Vijayanagara and see the Mrauk-U royalty as ‘Sultans among Buddhist Kings’, would still leave unanswered the question of why the cultivation of Persian traditions ground to an apparent halt in Arakan after 1666. ³⁵ We lack insights into why an ‘Islamicate’ political language might have been used against a ‘Persianised’ polity. To repeat, we do not know under what circumstances ‘cosmopolitanism’ gave way to a politics of difference in South and Southeast Asia. Attempts at answers must of course begin with the historical contexts, or events, that gave meaning to such assertions of difference.

The Making of the Frontier

In terms of their reach and consequence, princely wars of succession were significant events in Mughal India. New work has revealed that if the absence of primogeniture at times spelt social instability, competition between Mughal princes also served the imperatives of state formation. Down to 1712, warring princes mobilised military and household alliances at different edges of South Asia. Inevitably, thus, their succession conflicts pulled the provinces to the core politics of the empire. ³⁷ In this regard, a succession struggles that broke out in 1658 stands out as remarkable. Fought between the four adult sons of Shah Jahan, it led to the accession of Aurangzeb, under whose dispensation the empire reached its most expansive limits. More evocatively, this ‘extremely dramatic’ war is often recalled in the literature for the rivalry between princes Aurangzeb and Dara Shikoh (1615–69), and for the unprecedented case of a forced abdication of an incumbent emperor, Shah Jahan (r. 1628–58). ³⁸ Nevertheless, the dynastic struggle has a much-neglected side to it, whose effects spilt over beyond Mughal India itself. To many at the time, it was Prince Shah Shuja‘’s military efforts that ultimately had large-scale implications for both Bengal and the Mughal Empire generally.

To summarise the narrative of Shuja‘’s participation in this war is to triangulate eastern India with the upper Gangetic plains and the Mrauk-U Kingdom. ³⁹ For, when Shah Jahan took ill in 1657, Shuja‘ was serving as the governor (ṣūbadār) of Bengal from his base in Rajmahal, Bihar, with a deputy stationed in Jahangirnagar. Shah Jahan’s decision to name Dara Shikoh as his successor meant that all the princes quickly threw themselves to the task of moving their armies to Delhi. Like his brothers, Shuja‘ crowned himself emperor, and marched from Bihar to the Doab plains. In the early months of 1658, Shuja‘ was defeated by Dara’s son in a battle near Benares. But Shuja‘ moved again. This time, he confronted Aurangzeb in Khajwa in Fatehpur, past Allahabad. An entire year after the first military expedition, Shuja‘ was once more routed in 1659. ⁴⁰ With Aurangzeb’s commander Mir Jumla in pursuit, the prince beat a final retreat to Bengal. In the eleventh hour, he found relief there from a Luso-Arakanese convoy of boats that had been sent for his sake. Moving from Jahangirnagar to Chittagong, Shuja‘ had apparently intended to set sail for exile in the Hijaz. In reality, the prince, his family and his entourage ended up in Mrauk-U, royal refugees at the court of King Candasudhammaraja (r. 1652–84).

What happened thereafter is a matter of controversy. Shuja‘’s death in Arakan following some courtly conflict has served as the source of not a few confusing accounts. In a few cases, it has even led to considerable distortions. Even modern histories have been given to portray this event as an example of one oriental despotism facing its mirror image or the tragic case of a prince meeting his end in an unruly frontier state. ⁴¹ Certainly, the incident attracted much contemporary notice, and part of the problem lies in that the sources have left contradictory reports. Many cited a story of violence. In numerous accounts, Candasudhammaraja had Shuja‘ murdered when the prince refused the king his daughter’s hand in marriage. Some argued that he was drowned, others still that he fled to obscurity in the nearby hills. ⁴² Dutch merchants in Arakan initially reported that the prince was found neither dead nor alive, only to later claim that he had indeed died. ⁴³ As the stories ramified, a few trained their sights exclusively to the drama of Shuja‘’s flight from eastern India, offering little of what happened after. In a seventeenth-century palm-leaf text, an Assamese account (burañjī) turns swiftly from Shuja‘’s betrayal in the battlefield at the hands of one of his own men to his getaway on an Arakanese vessel:

On another day, to go to war Cuja mounted an elephant, and there followed the terrible fighting of battle. The son of Racid Kha, who held the flywhisk, said, ‘Padsya, if a shot strikes you from afar, all will be lost. Get off the elephant and ride on the horse.’ As soon as Cuja got off the elephant and became a cavalier (cōrār; Pers. suwār), there was a battle. Cuja was defeated. When people heard of the change in Cuja [and his fate], his army joined the side of Mijamula [Mir Jumla]. Cuja got on a boat and escaped with the Mags (nārat uṭhi palāi magōlai gala). ⁴⁴

Indeed, it was an extremely similar narrative that was also adopted by the Mughal state. But the difference was that these events, which had entered quickly into circulation through very distinct narrative terrains, served for the Mughals a more particular political conceit.

In the pages of the ‘Ālamgīrnāma—a court chronicle written for Emperor Aurangzeb little over a decade following the events—much sarcasm is reserved for Prince Shuja‘, who is only referred to as nā-Shujā‘, that is, the ‘uncourageous’ or the ‘cowardly’. Yet what the author, Muhammad Kazim has to say remains valuable. This is not only because he gives us the only ‘official’ version of the events but primarily as he provides one of the first references to the new imperial idea of a Mughal state frontier in Chittagong. Muhammad Kazim writes that after suffering his second defeat in 1659, Shuja‘ and his weakened army returned to Bengal, only to contend with a local uprising sparked there by a confederate (muttafiq) of dissenting zamindars under the leadership of Munauwar Khan, a powerful landholder of Jangirnagar. At this desperate juncture, Shuja‘’s son and heir Zain-ud-Din reportedly sent gifts to the raja of Arakan, soliciting military assistance to help ‘suppress’ the revolt. At which:

The raja responded to the message for help and sent ships and corvettes of numerous groups of Arakanese (Raḵẖangiān) of demon-like features and frightful temperaments (dīv-sirat ba-hā’im-ḵẖo). That branch from the Tree of Corruption [Zain-ud-Din], with the help of the race of ignoble people, defeated Munauwar Khan. Afterwards, he hastily dispatched a significant sum of money in the hands of the departing group. He also sent a letter seeking confirmation that should his sedition-fostering father need to leave Jahangirnagar for Rakhang in his distress, he would require their assistance to move from this region and safely cross the frontier (sarḥad) into that abode of infidelity (kufr-ābād). The raja sent a message to his governor in Chittagong (Chāṭgām) declaring that the frontier of Rakhang ought be secured so that the cowardly nā-Shuja‘ may seek protection at the port without hindrance. When the vile and lowborn Arakanese arrived in these regions they caused great destruction to the areas around Jahangirnagar. This is because their traditions are irreligious and their customs deviant (shīmat-i ān bī-dīnān wa ‘ādat-i ān ẕalālat). ⁴⁵

Aside from providing a description of events that led to Shuja‘’s defeat, the account here carries the added burden of explaining the political and cultural (indeed, ‘Islamicate’) patterns of contrast between Mughal India and the Mrauk-U crown. In this studied explanation of a politics of difference against Arakan, the break from the earlier writings of Abul Fazl, Muhammad Tahir or Mirza Nathan is evident. At the same time, the notion of a frontier in Chittagong is also fortified by its narrative affiliation with the ‘seditious’ (fitna-parwarish) links that a Timurid prince sutured between the Mughal Empire and the Mrauk-U Kingdom.

As with the Assamese chronicle, the ‘Ālamgīrnāma does not dwell on Shuja‘’s death in Arakan. Also like the burañjī, the Mughal account specifies desertions from Shuja‘’s army as a key point of the second battle, saying that this left the prince ‘captive in fear and horror as the procession of the victory-adorning army with its clamour and salvo was heard to be nearing in [his] pursuit’. ⁴⁶ Yet what Muhammad Kazim leaves unequivocal is that in his escape from Bengal, a ‘miserable’ Shuja‘ had crossed the frontier into a very different polity, that is, to a ‘country (wilāyat) of exemplary infidelity’. ⁴⁷ In its departure from earlier Indo-Persian opinions on the Mrauk-U state, the new imperial narrative obviously found a fluent expression in the idiom of a ‘frontier’ at Chittagong. Notably, moreover, that idea was made meaningful with reference to Shuja‘ and his failed exploits in the war of succession. The very unexpected turn of events that led to his escape to Arakan thus not only pushed the new reality of a frontier into state sanction but as a pivotal incident, it also arrested the attention of many around India. Such celebrated European figures in early modern India as Bernier, Manucci and Bowrey therefore all reflected on the significance of this episode for Mughal politics. ⁴⁸

Made vivid by literary depictions, the events were also played up with some sound and fury in the Mughal imagination. ‘As the two armies roared at one another, gone from the heavens were sense or wisdom.’ Thus Muhammad Bihishti Shirazi, former servant to another defeated prince, Murad Bakhsh, on the military encounter between the Shah Shuja‘ and Aurangzeb. ⁴⁹ But through the ‘thundering crash of steel’ or the ‘noise of war drums’, what still came through in his mas̤nawī is a narrative that had become an established version of the events. Shirazi does not mention the involvement of Arakan in his poetic depiction of the war. However, like Kazim he does make known that a ‘hopeless’ Shuja‘ ultimately decided to cut back on the losses incurred in the battlefield, and after seeing soldiers desert from his side, fled from Bengal on a ship. ⁵⁰ Marked by a certain brevity, the many depictions of Shuja‘’s flight to Arakan fail in some senses to capture the extent to which it shaped the Mughal perspective towards Chittagong and indeed other frontiers of the empire. But the fact is that Shuja‘’s ghosts continued to agitate the perspective of the Mughal state. Later sightings of the prince were thus reported from one revolt in Bihar, from the unstable northwestern frontiers and even from the domains of the rival Safavid Empire in Iran. ⁵¹

The afterlife of Shuja‘ will be covered at some length later in this essay. For now, it is worth pointing out that like all frontiers, Chittagong had another side to it. Voices that flickered in Arakan offered an extremely dissimilar take on the maritime limits of the Mrauk-U state. The following, written in the 1670s, rings true:

Countless boats of many jatis, of many hues,

Beautiful kinds of vessels

The royal viewing window, the ruler’s staff, throne and flywhisk,

And an entire sea filled with ships. ⁵²

Composed by Alaol (c. 1607–80), a native of Bengal who had found work as a court poet in Arakan after some mishaps of travel in the Bay, the verse appears to turn the Mughal idea of an intractable frontier on its head and presents instead a serene vista of the Burmese coastline. Given this, and given that the above comes from one of Alaol’s highly original translation of a Persian poem, scholars have recently turned to his broader corpus of writings as reflective of the contradictory cultures of ‘Persianisation’ and ‘anti-Mughalism’ at the Mrauk-U court. ⁵³ His career and poems warrant revisiting here, however, inasmuch as they might offer a deeper appreciation of how the Arakanese crown construed its relationship with the Mughals, and how it too understood Shuja‘’s flight to Burma (or here, his arrival) as an eventful incident.

After paying his tribute to the Mrauk-U Kingdom, Alaol in his Sikāndarnāmā first considers his own rise as a distinguished poet. In what is one of the very few sources extant on his biography, he lets known that after coming to Arakan, widespread recognition of his literary talents led to a period of relative contentment. ‘Like so, much time passed in happiness’. Yet the happiness proved fleeting, and according to the poet was disturbed by no other incident than the appearance of Prince Shuja‘ in Arakan. His references of the Mughal’s arrival, which perhaps due to the ‘anti-Mughal’ patronage does not dwell on the prince’s murder, is almost oblique in its depiction. Even so, Alaol interprets the event as what brought forth considerable changes at the local court, to the point of upsetting his own life. Shuja‘’s appearance, he observes, was unforeseen, a ‘motion of fate’ with ‘disastrous’ (jañjāl) consequences. And though he blames the faulty counsel of Shuja‘’s advisors for the conflict that followed, he explains how he too became ensnared in the ensuing turmoil. Falling under suspicion for being one of many local Muslims who apparently sided with Shuja‘, Alaol writes of how he was also imprisoned for some time. ⁵⁴ In the end, things ended on a more buoyant note for the poet. Cleared of wrongdoing, he was also returned to royal favour. ⁵⁵ Leaving aside the difficult question of which party at the court Alaol was indeed sympathetic towards, how then to make sense of the evidence? Without overstating the point, it may be argued that the moment that brought the Mughal and Arakanese states together in a transformative encounter also touched off changes in linked cultural networks. Approached as a constitutive part of the past—rather than as simply reflective of it—Alaol’s writings show how Shuja‘’s journey from Bengal Arakan was grasped as a significant moment on both sides of the frontier. Moreover, they underscore how ‘Persianisation’ did not always necessitate cultural commensurability, and that if it were expressed, accounting for difference hinged, even for contemporaries, on referring to the impact of specific events.

It then remains suggestive, but not surprising, that when the Mughal military moved to annex Chittagong, it pointed to Shuja‘’s career in Bengal as what fuelled justifications for war against Arakan, though other reasons were elaborated besides. In the words of one scribe in Bengal, Shihab-ud-Din Talish, war was made necessary to curb Arakanese slaving within the province and to keep pirate attacks (ta‘arruẕ-i harmādān) offshore. ⁵⁶ The new governor, Shayista Khan, further submitted that the conquest of the port city would allow for the recovery of the Bengal fleet, or nawāra, which was in a state of disrepair and ‘worsening day by day’. That at this point the recent history of Shuja‘’s last years in Bengal was recalled to make sense of the ‘destitute’ condition of the fleet reflects how the prince and Arakan had become coupled in Mughal political thought. Talish confirms: ‘The reason for the decline of the nawāra may be summarised as follows. During the time of the governorship of Shahzada Shuja‘, a storm of chaos and confusion erupted from the furnace of [his] neglect.’ A period of rampant ‘extortion and violence’ by accountants responsible for allotting jagīr assignments led to ‘poverty and hunger’ for the shipwrights in their parganas. ⁵⁷ To undo this history of misrule under Shuja‘, the state therefore readied to invade Chittagong and Arakan.

Almost immediately, the provincial government began reconstituting the Bengal nawāra. In 1664, new officers were promoted to direct the fleet and presented with robes of honour (ba-mashrafī-yi nawāra sar-afrāz karda wa muḵẖalla‘ kardānīda). Meanwhile, new land assignments were accorded, and timber (chūb) and shipbuilders (kārigar) brought in from districts in and around Bengal. ‘It was ordered that at the ports of Hugli, Balasore, Murang, Chilmari, and Jasru (Jessore), all kinds of ships be built and sent’ to Jahangirnagar. ⁵⁸ The renewed fleet that was finally mobilised near Noakhali in the lower delta was joined by both renegade ‘Frankish pirates who had come from Chatgam’ and imperial servants. From the latter, a certain Ibn Husain alone commanded 288 vessels, ranging from larger frigates or ghurābs to smaller vessels like parandas. After a series of military confrontations with the Arakanese that began in around 31 January 1666 (25 Rajab 1076), and which unfolded on both land and water, Mughal forces managed to choke the Karnaphuli River, at whose mouth stood the Chittagong fortress. ⁵⁹ Finally, that following month, Shayista Khan’s son, ‘the Khan of the victorious banner’, breached the fortress gates. ⁶⁰ This victory at Chittagong apparently motivated the Mughals to continue. Wresting Arakenese arsenal and ships, another contingent ‘crossed dense jungles and choppy waters’ (jangalhāi-yi aṃbūh wa ābhā-yi dushwār) to seize Rambu, an Akarnese port 12 days’ journey from Chittagong. Their intention, it seems, was to push onto the heart of the Mrauk-U Kingdom. But with the monsoons (barash-kāl; Beng. barṣākāl) closing in and supplies running low (qulat-i āẕūqa), they returned to secure the beachhead in Chittagong. ⁶¹

The invasion of Chittagong, and the state consolidation of the frontier which resulted thence, cannot all be attributed to the events of Shah Shuja‘’s flight to Arakan. Other reasons surely impelled the Mughals to annex the Arakanese port city. In addition, broader transformations were already underfoot in Arakan and Burma. Alaol’s eulogies notwithstanding, historians have variously described this period as a ‘crisis’ or ‘watershed’ moment that cleared the way for the expansion of inland states in Burma. ⁶² Yet the extraordinary story behind the ‘land-based’ Mughal state waging war in the Bay of Bengal, and that against a state we have come to recognise as a ‘hybrid’ maritime–agrarian polity, illustrates several points that have been put forward thus far in this essay. ⁶³ It bears repeating that an ‘Islamic’ language reinforced the conquest of Chittagong and the early stages of Mughal rule at the city, even as the ‘cosmopolitan’ character of Arakan continued to resist this discourse. Apparently, the Mughals in Bengal were still under vague impressions that the Arakanese ‘do not generally possess religion, though their conduct inclines towards the Hindus’. ⁶⁴ Then again, in North India the state celebrated the conquest as what ‘made useless the banners of the powers of infidelity’. ⁶⁵ Most provocatively, Chittagong was then renamed Islamabad, the Abode of Islam.

Yet to the moment of the actual conquest, a theme that came up again and again was the last years in the life of a Mughal prince. Whether in his military defeats in the north, his disastrous visit to Arakan or in his callous rule as governor of Bengal, Shuja‘ had suddenly become the central figure in imagining a dividing frontier in the southeast. Approached from comparative perspectives, one might well notice distant parallels between this story and the flight of the last Ming pretender to Burma. Only a year later after Shuja‘, the Prince of Gui’s fugitive court was after all first welcomed and then massacred by the rulers of Ava. However, and as the following section will show, the Shuja‘ episode proved to have much more import for early modern India than Zhou Youlang ever did for Qing China. Not only as it cemented an enduring political idea, but also because its memory continued to resurface in astonishingly different contexts.

Remembering the Frontier

‘Don’t go, don’t go, Pari to the country of Roshang’, cries a ballad from Chittagong. Calling back on the tale of Shah Shuja‘’s daughter in Arakan, a princess here named Paribanu, the ballad is a forceful reminder of the freight that the events of the 1660s had on local memory. In its presence in the oral repertoires of village Bengal, and in the movement of these stories between Arakan and Chittagong, we also have evidence that should we account for their exact itineraries, the ‘universal’ dimensions of Indo-Persian culture seem much less quixotic. Rather, they appear as well as the product of Mughal immersion within ‘local’ milieus. ⁶⁷ The rural ballads lacked the polished precision of Indo-Persian writings to make sense of the frontier, but they unmistakably pointed to the perceived limits of Chittagong by commemorating a ‘great conflict’ that took place between Prince Shuja‘ (Sujā bādsājādā) and the king of Arakan (Rōsāṅgyār rājā). The conflict unfolded somewhere beyond the ‘limitless sea’, but so far as the minstrels were concerned, it still contoured a frontier whose limits could not be so easily tested:

Don’t go, don’t go Pari to the place of the hill-people,

They stew and cook and eat human flesh,

I tell you, don’t even set a foot there.

You drowned Pari at sea—

What the song here calls attention to as a ‘den of snakes’, home to a barbaric people (muraṅgyār ṭhāi), was actually part of a larger body of such references. Other songs similarly remembered a Mughal princess lost in Burma, and this theme would later frame the ‘high’ literary cultures of modern Bengali texts. ⁶⁹ In another song, an unnamed princess from the house of Timur berates her poor luck for being orphaned in Roshang. She weeps into her pillow; she affects haughty disdain for the foul foods and strange clothes the raja offers to her:

What fate was this?—

I was a celestial flower,

But I lost caste and kin to the hands of the Magh Raja.

Beneath the sea you made a grave for mother and father,

Who will ever get word of me in this country of hārmmāds

Mother, father, who will get word—

What fate was this? ⁷⁰

Eventful historical moments, as Sewell reminds us, are given meaning by the kind of amplified emotions expressed by these ballads. As well, their after-effects linger through rituals of reproduction across distinct sociocultural boundaries. The Chittagong frontier was thus invoked, as it was in village Bengal, by persistent reference to the memory of Prince Shuja‘. In the century or so that followed the coupled events of the war of succession and the Mughal annexation of the city, this stood true for ‘Muslim’, ‘Hindu’ and indeed, tribal and early colonial narratives. An exhaustive survey of this genealogy is beyond the scope of this essay. However, we might conveniently begin with a first look at how the frontier was conceived by the immediate ‘successor state’ to the Mughals in Bengal.

Arguably, Chittagong was really cemented as a political unit of provincial Bengal in the early decades of the eighteenth century, that is, under the rising star of Nawab Murshid Quli Khan (r. 1716–27). We are aware that the so-called ‘regionalisation’ that occurred under his satrapi did not merely occur from a state ‘crisis’ at the ‘centre’. To the contrary, imperial decentralisation was intensely spurred by efforts at the new provincial capital of Murshidabad to place Bengal under the regularity of a reformed ẓabt̤ tax system. ⁷¹ Given a firm administrative backbone during the reigns of the first quasi-independent Nawabs of Bengal, their fiscal policies brought about both medium- and long-term transformations in the region. Besides revealing changes in the state’s orientations towards Chittagong from an earlier period—when officials were sent on something like punishment posts to this frontier ⁷² —new forms of governance, which included dispatching military reinforcements and the construction of garrison forts (thāna) under Nawab Shuja‘-ud-Din Muhammad Khan (r. 1727–39), ⁷³ together with an assertive taxation regime, all pressed Chittagong more tightly into the purview of Bengal proper.

Revenue rolls from Bengal from 1133 AH/1720–21 CE indicate that when viewed through the prism of the tax demands (jama‘) made on the district, Murshid Quli Khan’s hold on Chittagong (Islamabad) penetrated tiers of both urban and rural settlements. Under his renewed chakla system—which maximised revenue collection by collapsing old district borders—the amounts assessed from the administrative unit included the jama‘ of towns and villages like Rahmatabad, ‘Abdullahpur, Sultanpur and Fatehabad, as well as newly cleared lands (no-ābād) whose existence came to light after fresh measurement surveys. ⁷⁴ Indeed, later incarnations of these assessments were described as being of ‘extreme intricacy and minuteness of the tenures’. ⁷⁵ But while this corner of Bengal certainly gravitated closer to a centralising provincial state, Chittagong’s standing as a frontierscape remained undiminished. An important ta’rīḵẖ on the Nawabs is thus unequivocal in locating Bengal’s boundaries (ḥudūd) on an east–west axis with Chittagong on one end and Teliagadhi in Bihar on the other. Its author, Ghulam Husain Salim suffered from few of the confusions of his predecessors in situating this maritime frontier (lab-i dariyā-yi shor) within Bengal. Nevertheless, he added that the relevance of this ‘ancient city’ (or its economic importance) had long since elapsed, and with the rise of Calcutta, ‘all the ports of Bengal have disintegrated (hama banādir-i Bangāla shikast shuda)’. ⁷⁶

This decline in Chittagong’s fortunes did not go unnoticed in other quarters. In 1718, some five decades after the events of Shuja‘’s flight, the Company servant Alexander Hamilton jotted down some first-hand observations on the port city. And he wrote, summoning the language of a frontier—‘Xatigam is a Town that borders on Bengal and Arackan’. Like Ghulam Husain, Hamilton suggested that at this point in its history, Chittagong was in a state of ‘poverty’, thereby rendering it a matter of ‘indifference’ as to whom it belonged. ⁷⁷ Further, he noted that though the Portuguese had in effect taken the city, a qāẓī still represented there the judicial writ of the Mughal state. But Hamilton then strikingly drew a straight line of comparison between this picture of political–economic marginalisation to the history of Shuja‘’s exile, stating that the disadvantaged Portuguese in Chittagong ‘followed the fortune of Sultan Sujah’, who was ‘forced to quit Bengal’. ⁷⁸ What reasoning goaded this remark, which glibly associated the frontier with the memory of Shuja‘? It is, perhaps, impossible to answer with certainty. Yet a likely source were the many allegories on Shuja that resonated across Bengal, with lore particularly legion on the route he had supposedly taken to Burma. ⁷⁹

Downriver from Dhaka, in Saadipur by Sonargaon, locals called mangoes from a particular tree ‘Shujā-pasand’, claiming that the prince had enjoyed a few on his way to Arakan. ⁸⁰ Southwards, in Chittagong itself, a mahalla was named Sujā-kāṭgar as he was said to have spent some time there. ⁸¹ To be sure, these stories did not bear an influence on all narratives and did not extend to many Indo-Persian texts of this period. But arguably, the reasons behind the silence of Nawabi chronicles lay elsewhere. As Faruqui explains, past political relations were conventionally reassembled following wars of succession in Mughal India, to the point of integrating back into the state those who sided with a defeated prince. ⁸² In the case of Shuja‘, this is borne out in the textual evidence. As when later Persian narratives from Bengal mention him, they remain quiet on the once-charged matter of Arakanese links, highlighting instead the history of his gubernatorial career in the province. ⁸³

That said, we might note that his memory does produce some fantastic narratives. An apocryphal travel account, written in either the eighteenth or nineteenth century by one Muhammad Haidar, consequently invents as its patron Shahzada Muhammad Sultan Shah Shuja‘, son of Shah Jahan, son of Jahangir. To satisfy the curiosities of the now-dead prince, the scribe crosses multiple frontiers to report on the marvels of countries (‘ajā’ibāt-i wilāyat) and examples of divine craftsmanship (ṣan‘at-i ilāhī). ⁸⁴ In the process, the Shujā‘-yi Ḥaidarī lays imaginary witness to lands near and distant from Bengal. Delhi-Shahjahanabad is lauded for the ‘sweet language’ of its denizens, Agra for its ‘cool and dry’ air and a landscape tinged ‘bloody red’. ⁸⁵ In the countries under the Padishah of Rum, ‘august felicity’ prevails. Ottoman Aleppo is home to a ‘tree whose delicate leaves emanate light’. Though it is bright enough to blister, Haidar tells readers that some people visit the tree every night to witness the spectacle (tamāshā). ⁸⁶ We also move to the West, or Farangistan. A land of ‘seven great cities’ for ‘seven different thrones’, it is blessed with miraculous waters that deliver from pain, creeks where Jesus bathed and people with ‘sheep eyes’ (mesh-chashm). ⁸⁷ Seemingly exceptional, such ‘mythic’ stories that centred on the prince in fact found expression in several other textual traditions.

Far up the hinterland of Chittagong, the Kingdom of Tripura thus also carved out its own place within this tradition. In sections of the Rājmālā, a genealogical poem compiled for the Manikya crown, one not only finds a reconfiguration of this tradition to serve a dynastic tale but also a recasting of its principal proponents. ⁸⁸ The originality of this verse tale is obvious enough, but its narrative sinews clearly draw from previous accounts of Shuja‘. The story begins in Tripura. Raja Govinda Manikya, we are told, has arrived at loggerheads with his half-brother, who is bent on usurping the throne. As it did for the Mughal prince, dynastic conflict here leads to a period of royal exile for Govinda. From Tripura, the king is forced to move downhill, and once more like Shuja‘, seek refuge across the frontier in Arakan. It is also at this point of the story that matters take a very interesting turn. We are now in Arakan, together with Govinda and his host, the Magh Raja (Candasudhammaraja?). Almost inevitably, Shah Shuja‘ himself then joins the courtly scene.

Seeing the Mughal prince, the hero of our poem Govinda commits what is certainly an act of deference. Reflective of the asymmetrical relations that once prevailed between the Mughals and the ‘little kingdom’ of Tripura, we see that Govinda stands to offer his own seat to the prince. But this gesture provokes confusion in the mind of the Magh king, who wonders out loud, why should Govinda present his kingly seat (siṃhāsan) to a Muslim mleccha? He is summarily disabused of his ignorance. After all, declares Govinda, this is the same figure universally known (bikhyāta bhuban) as lord among ‘kings like you and me’ (tumi āmi hēna rājā). Needless to add, the account ends on a variation of the known story, and Shuja‘ is ultimately murdered in Arakan. Nevertheless, the poem explains that upon regaining his throne in Tripura, Govinda Manikya carries back the memory of the Mughal almost as a talisman for the well-being of his future reign (or perhaps, as memento mori). He thus sells off what in different versions of the text was either a sword or a signet ring that Shuja‘ had given him and uses the money to commission the building of a mosque bearing the prince’s name in neighbouring Comilla. ⁸⁹

Now, it is not easy to account for the precise reasons for this story in the Rājmālā. However, the text’s implicit ‘vindication’ of the legacy of the prince warrants speculation if in the historical thought of the Kingdom of Tripura—which in any case had come first under Mughal and then Nawabi aggression over this period—there was an idiosyncratic understanding of political–cultural difference as what governed relations between Mughal Bengal and Arakan. Or indeed, did the story serve to underscore distinctions between the states of Tripura and Arakan? ⁹⁰ Whatever the actual intentions of the Rājmālā, it remains the case that other regional polities continued to place Shuja‘ within the folds of their tales, thereby preserving both the eventful memory of the 1660s and the normative meanings of the frontier it had promulgated. East of Tripura, the royal chronicle of the Meetei kings of Manipur hinted that the Mughal state had sent a few delegates to the court of Meetingu Khunchaopa (r. 1652–66) in order to inquire if Shah Shuja‘ had come to his court. ⁹¹ Evidently, the trail had gone cold. Had the envoys ‘Sangku’ and ‘Mankan’ gone to the hill tracts of Chittagong, they might have found some more leads in among the tales of the upland tribal groups.

Oral accounts of the Chakma people pointed to fascinating counterfactuals to the history of Mughal–Arakanese encounter. Again, history fortified a conceptualisation of the frontier and connected the present back to an eventful past. But the Chakma narratives of the 1660s also underwrote an origin tale for the tribe itself, and memory was the foil with which it made sense of its own place at the political–cultural margins of Bengal. One Chakma account collected in the nineteenth century insisted that they were previously Mughals. Tracing their lineage back to the conquest of Chittagong, the narrative suggested that the beginnings of the tribe were to be found at the moment when the imperial army came to the frontier and crossed from Bengal into Arakan. It was at the latter venue that the army had encountered and incidentally angered a holy man. A curse from this ascetic led to the defeat and enslavement of the soldiers. Thereafter, the history continued, a long period of miscegenation between the Mughals and the women of Arakan gave way to the Chakma tribe and its lines of descent. ⁹² In another oral report that surely rippled out from this master narrative, Chakmas argued they were once Kshatriyas who had invaded Arakan and intermarried there. ⁹³

Evaluated together, what could be taken from these stories from the frontier? Their diversity opposes a coherent answer. Nevertheless, what they share is a historical entanglement with the twin events of Shuja‘’s flight from and the Mughal annexation of Chittagong. Likewise, they all gave tacit or explicit voice to a politics of difference that governed relations with the people of Arakan. Indeed, this legacy persisted well after the fall of the Mrauk-U Kingdom to the Kon-baung dynasty of Burma. As late as 1798, the colonial official Francis Buchanan was in the region when he came across an Arakanese informant. As a rule, Buchanan noted, such men were ‘detested by the natives’ of Chittagong, be they Hindus or Muslims. Further conversations with the Arakanese also led him to suggest that they now face ‘severe oppression’ in Bengal. ⁹⁴ Perhaps some of this hostility towards Arakanese ‘migrants’ resulted from cross-border slave raids, which continued to the close of the eighteenth century. ⁹⁵ However, as this essay has shown, much of this language centred on the ideology of the frontier was neither contingent on the historical present nor did it fall outside history as such. Rather, they issued forth from an eventful moment and found a sustained afterlife at the crossroads of Mughal and vernacular cultures. We have, in this respect, arrived full circle to the home-grown historian of Chittagong, ‘Abdul ‘Ali Islamabadi, and to the arguments to which he gave voice in his nineteenth-century chronicle.

Conclusion

By historical definition, empires were societies of difference. Or, at any rate, scholars have come to the consensus that their management of ethnic, confessional and gender boundaries distinguished empires from modern nation-states. ⁹⁶ Likewise, imperial courts were sites of rich tension. As spaces, they often effaced and at times defined the seemingly incommensurable norms and beliefs of their subjects. ⁹⁷ Yet perhaps no politics of difference ever existed in timeless form, always present as atavisms in the thought-worlds of historical actors. More likely, differences emerged from historical events and specific conjunctures, or as Richards and Rao eloquently summed up in a similar context, past ‘convergences’. ⁹⁸ Viewed from this vantage, the many ideas and sites of imperial difference throw up the task of situating the particular circumstances through which states or its subjects defined themselves against others. While difference often resulted from longer and less conspicuous moments of encounter, in a distant corner of early modern Bengal it was the result of a dramatic moment. As much as it created the idea of the Chittagong frontier, the memory of that moment also proved sufficiently resilient to cut the grooves of modern political thought.