Introduction

This chapter addresses the complex history of slavery in South Asia during the early modern and colonial periods. This period witnessed a diverse range of practices that can be considered under this rubric, from household slavery and concubinage to military and agrarian slavery; it also witnessed both shifts and continuities in practice and discourse in the midst of profound political transformations. In fact, to talk about slavery in this broad geographical area, during this period of history, is to straddle divisions of polity, language, and era. To compare slavery in contexts as divergent as the North Indian Mughal Empire, Portuguese Goa, and British India within such a short essay entails a great deal of simplification. Nevertheless, this chapter can provide a useful starting point for thinking through the numerous forms of, and beliefs surrounding, slavery in this region during this period, as well as the kinds of shifts that occurred in the wake of the emergence of colonialism and abolitionism. What we find, as Richard Eaton notes, is not “a single story of slavery,” or a “tidy sequence of evolutionary ‘stages’,”1 but rather many different stories taking place in many different contexts, that may or may not relate one to the other. While it may not necessarily be productive to consider so many different practices under the single rubric of “South Asian slavery,”2 if we consider instead a range of South Asian slaveries, we can avoid presuming similarities or commonalities between disparate examples, and arrive at a more accurate understanding of our subject.

Preliminaries: On Terminology

The identification of slavery and enslaved individuals in early modern and colonial South Asia as well as the distinction between slavery and other forms of servitude are ongoing challenges for understanding slavery in South Asian history. Many scholars have argued against imposing a modern, post-enlightenment strict binary definition of slavery here, categorizing individuals as simply either enslaved or free.3 While some individuals were unequivocally identified as slaves who had been bought or gifted, many other social positionalities blurred such sharp binary distinctions. Aside from the existence of other highly coercive forms of labor, however, we encounter difficulties even on the fundamental question of identifying enslaved individuals. Although there are a range of words that unequivocally mean “slave” in early modern South Asian languages, including terms such as ghulām and banda (from Persian) and dāsa or dāsī (from Sanskrit), these same terms are also used in historical sources to simply denote loyal service and devotion. Thus, for instance, one sees references in the Mughal context to (free) nobles as banda-yi dargāh (slaves of the court).4

Furthermore, a range of terms blur the lines between slave and other categories denoting relations of servitude, kinship, and discipleship. In the north Indian context, the Hindi term chela can denote either slave or disciple;5 the Persian kanīz and parastār refer to either female servants or female slaves. Guha discusses a similar challenge with the Marathi terms batik (etymology unknown) and kunbini (derived from the term for “household woman”), used for female “servile dependents” in eighteenth- and nineteenth-century Maratha territories; he also notes that familial terminology is also employed, such that in much documentation the words batik and muli (daughter) are used interchangeably.6

These categories have also been redefined at several historical moments, due to attempts to (at least theoretically) curtail or ban enslavement and slavery. During the Mughal period, the Emperor Akbar (r. 1556–1605) took several actions against practices related to slavery. As described by court historian Abū al-Fazl, Akbar first issued an order forbidding the enslavement of the female relatives and children of enemy combatants in his seventh regnal year; other sources record that Akbar subsequently also moved to ban the sale of slaves. It is worth noting, with reference to the above discussion, that enslaved individuals were recategorized as disciples (chelas) as a corollary to these measures.7 During the reign of Akbar’s successor Jahāngīr (r. 1605–1627), an imperial order was issued banning the castration of slaves in particular, a practice that was highly profitable due to the monetary value of eunuch slaves.8 While scholars have noted the absence of reference to formal slave markets (which had existed in the prior period during the Delhi Sultanate) as a potential sign of the efficacy of such measures, it is clear that such government actions did not mean the end of slavery in practice, even within the service of the emperors themselves.9 In fact, there is continued evidence of enslavement and slave trafficking both during the reigns of Akbar and Jahāngīr, as well as thereafter, including those forms putatively banned. In short, as Eaton observes, these decrees ending slavery (at least in some forms) did not actually put an end to slavery, even if they did have an impact in terms of the nature and scope of the slave trade.10

In the colonial context, we encounter another, better-known moment in which slavery in British India is targeted for curtailment and eventually (again, theoretically) abolition under colonial rule. As is well known, while British merchants dominated the Atlantic slave trade in the seventeenth and eighteenth centuries, the emergence and success of the anti-slavery and abolitionist movement from the end of the eighteenth century led to the banning of the slave trade in 1807 and the passing of the Slavery Abolition Act in 1833. While the 1833 Act excluded British India, it was followed in 1843 by the Indian Slavery Act, which will be discussed at more length below. This Act, generally considered ineffective and toothless, may have been most influential in changing the terminology around practices of slavery rather than abolishing it. This phenomenon has been described by Indrani Chatterjee as “abolition by denial.”11 Despite the erasure of enslaved individuals in public documents such as Parliamentary Papers, examination of documents such as inventories or wills reveals their extensive presence in British colonial households. While the ownership of slaves was acknowledged in the case of indigenous elites (although even this form of slavery was downplayed as being benign, “mild” and familial) the presence of slaves was denied or else they were re-designated as “servants” in the case of British colonial households, despite ample evidence to the contrary.12 In South India, we see a similar phenomenon, with agrarian slaves referred to post-“abolition” as “agrestic laborers,” without any corresponding change in rights or labor conditions.13

Through both denying and camouflaging British involvement in slavery practices on the Subcontinent while refusing comparisons with Atlantic world slavery, any push to bring abolitionist campaigning to bear on British colonial holdings in India was for the most part effectively forestalled. In the process, slavery in British India was redefined contradictorily both as familial, and thus outside of the realm of appropriate official meddling, as well as a matter of choice, and thus falling under the ambit of Master and Servant laws.14 Further, slavery was sometimes described as having a “positive social function” in times of famine or other moments of desperation. In this context, it was argued, slavery could be seen as a “kind of welfare measure… an acceptable alternative to death by starvation.”15 Here it should be noted that the trajectory of abolitionism in British India is distinct from that in British-ruled Ceylon (modern Sri Lanka), where slavery began to be dismantled from 1816 onwards. In tracking this gradual process of amelioriation and abolition, Wickramasinghe emphasizes that in this matter Sri Lanka can be considered as much more similar to other crown colonies and entirely different from British India.16

As the above suggests, attention to the specificities of terminology and historical context is crucial to identifying enslaved peoples in South Asian history. These complicated questions of terminology and identification, alongside the diversity of forms of slavery in this period, underline the need to focus on “locally specific usages, relationships, terms, institutions, and processes as they shifted in time.” In other words, the elusiveness of modes for the easy identification of enslaved individuals, and the refusal of the terms employed to fit easily into the frames provided by comparative contexts, is not just a challenge to be overcome but perhaps reflective of the granular specificity of the practices and people being studied. In practice, acknowledging sometimes blurred boundaries between categories, however frustrating to the modern interest in pinning down precise free/unfree status, should be considered as fundamental to the project of understanding slavery in South Asia.17

Entry into Slavery

While slavery and the slave trade have a long history in South Asia, from the ancient and medieval periods onward, here we will focus on the early modern and colonial periods. Throughout the period under review, South Asia was both the location of the enslavement of individuals subsequently sold both domestically and abroad, as well as a market for those enslaved outside of the Subcontinent. In other words, the trade was multidirectional; it also occurred locally as well as across long distances. The long-distance slave market operated both overland through Central Asia as well as overseas across the Indian Ocean World.

While the historiography of these trade routes is difficult due to the fragmentation of the archive by language and polity, we can discern some basic facts. The overland trade in South Asian slaves in Central Asia during the medieval and early modern period was carried out by caravan merchants who “either purchased them outright or received them in exchange for other commodities in demand in India, especially horses.”18 This trade continued into the eighteenth century, although by this time the volume of enslaved South Asians sold in Central Asia seems to have significantly decreased, due to the shift to exchanging textiles (rather than slaves) for horses and/or shifting political dynamics in North India. Nevertheless, there remained small numbers of South Asians sold in Central Asia into the nineteenth century.19 While there is little scholarship on overland trade in slaves to South Asia, significant numbers of passing references to enslaved individuals identified for instance as Turkish, Qalmaq, and even Russian demonstrate that this trade went both ways.

When it comes to overseas trade in this period, the majority of scholarship has focused on enslaved Africans. The primary trade links between western India and East Africa can be traced either through connections between the Horn of Africa and the Swahili Coast and Gujarat, or connections via the Hijaz and Hadramaut with Deccan ports such as Chaul and Dabhol.20 For instance, there was a substantial flow of enslaved habashīs (northeast Africans) into the Deccan from the fifteenth to the seventeenth centuries, as part of a larger Indian Ocean slave trade. These individuals were enslaved due to a range of factors, including Arab raiding parties, internal war, or voluntary sale by poverty-stricken parents. Thereafter, many of these slaves would have been sold in slave markets in the Middle East; some among them would have reached South Asia via the western port cities mentioned above. It seems that at least in the early modern Deccan, the sale of enslaved Africans served in part to offset the demand for Indian cotton; it also served to supply foreign military slaves to local elites in the Deccan to manage the complex and indeed antagonistic political dynamics there.21 Such non-Indian slaves would have been considered luxury items, which perhaps explains in part why Mughal sources are so specific in listing the ethnic identifiers of slaves present in the royal household. However, despite the visibility given to such slaves in the sources, the majority would have been of South Asian origin.22

Within early modern South Asia, enslavement seems to have occurred most often in war or due to economic hardship. While accounts of military engagements often mention the taking of captives who thereafter live as slaves, we also see lower-level raids or kidnappings leading to enslavement as well.23 However, it is worth noting that actions to curtail and punish such crimes suggest other sources were a lower risk.24 Enslavement also took place in the wake of famine and other times of economic difficulty, when the desperate sold family members or even themselves. The market value of enslaved individuals could be impacted by a range of factors, as desired traits included physical beauty, professional skill or expertise (for instance, in textile work or engineering), and castrated status in the case of eunuchs.25

While most enslaved peoples in South Asia during this period were most likely enslaved within South Asia itself, enslaved individuals labeled as non-Indian, with identifiers such as northeast African (habashī) or Turkish (turkī) regularly find mention in the early modern sources. Thus scholars have tracked the life trajectories of individual slaves such as Malik ‘Ambar, who was born in modern-day Ethiopia in the Kambata region in 1548 before being enslaved either in war or due to poverty. After initially being sold in the markets of the Middle East he was later brought to the Deccan in the early 1570s where he constituted one among a thousand habashī slaves bought by the peshwā (prime minister) of the Nizām Shāhī dynasty, himself a formerly enslaved habashī.26 ‘Ambar himself is later described as leading an army including ten thousand habashī soldiers.27 The eighteenth-century case of Tahmās Khān, who was enslaved as a child in modern-day Turkey before, after multiple transfers and giftings, ending up being raised alongside a number of other Turkish-speaking enslaved boys in the household of Mughal nobleman Mu’īn al-Mulk, suggests significant numbers of enslaved Turkish boys in such elite households, even if only the most successful among them emerge fully in our archives.28

The evidence points to both slaves forming a part of tribute and gift exchange, as well as through regular purchase with cash. As was discussed above, we do not find mentions of public slave markets in this period (in contrast to the Delhi Sultanate), but historical records show the gifting of slaves among the elite. Thus in the Mughal context, we see the emperors Jahāngīr and Shāh Jahān receive habashī and qalmāq slaves as tribute from their elite servitors.29 We also see the emperor Jahāngīr, despite his professed banning of the trade in eunuchs, receive a tribute of dozens of eunuch slaves from the governor of Bengal in 1613, 1621, and 1622.30 In the mid-eighteenth-century Tahmās Khān also recounts being given to Mu’īn al-Mulk, governor of Punjab, shortly after being brought to South Asia by his former owners.31

Not all transfers of enslaved individuals were voluntary. In the context of early modern Rajasthan, the transfer of valuable, skilled enslaved performers could also become points of contestation and conflict as “the voluntary transfer of such performers signified a relationship of vassalage to an overlord; in contrast, their forced surrender signified subjugation and the loss of sovereignty.”32 We see a similar situation with the refusal in 1561 of the noble (and foster-brother to the emperor) Adham Khān refusing to hand over the female dancers of Bāz Bahādur taken into his custody following the defeat of Mandu to the emperor Akbar.33

While the courtly histories tend to provide more information about the gifting or forced transfer of enslaved individuals, not all slaves were gifted rather than sold. Despite the lack of evidence of public spaces for the buying and selling of slaves during the Mughal period, such transactions do find mention in sources. Elite texts do not generally give details on this, but the occasional (denigrating) reference to individual enslaved people as zar-kharīda (gold-bought) as well as sporadic records of sale such as those found in the collection of the National Archive in Delhi points to the regular sale of slaves for cash.34

Before moving on, it is important to acknowledge the role of caste and religious community to enslavement. In the Islamicate context, theoretically one could not enslave Muslims; however, the capacious definition of “non-Muslim” led, for instance, to the enslavement of Iranian Shi’is in the early modern period, alongside other groups such as Buddhist Qalmaqs.35 Here it should of course be remembered that the act of enslavement was not only guided from above, but rather also engaged in illegally and opportunistically, which suggests such proscriptions should not be assumed to be entirely effective.36 With respect to caste, while in some cases (especially agricultural slavery, on which more below), certain communities were particularly vulnerable to enslavement, we find enslaved individuals from a multiplicity of caste backgrounds. For example, we find high-caste slave-owners in eighteenth-century Maharashtra carefully making clear the high-caste status of their slaves.37 Caplan has argued that, at least in Hindu contexts during the colonial period, caste was significant in the kinds of labor or positions assigned, thus making a distinction between enslaved sāvarna (caste) Hindus more often assigned to the domestic sphere versus agrarian labor being assigned to those of lower-caste or Dalit background.38 With reference to the latter category, the case of South Indian colonial-era agrarian slavery is particularly interesting. While the relationship between caste and slavery may shift between the precolonial period and the better-attested colonial context, scholarship on the nineteenth century has underlined that low caste status was a prerequisite of—but not necessarily equivalent to—enslaved status among agrarian slaves.39 The link between caste and enslaved status was such that, at least in Tamil-speaking parts of the Madras Presidency, certain caste names were used interchangeably with the words for slave.40

European arrivals in South Asia from the turn of the sixteenth century onwards both participated in, and transformed, the slave trade and local forms of slavery they encountered there. In addition to the presence of enslaved people within colonial households in South Asia itself, along similar lines as indigenous elites, we see Europeans engaging in the slave trade as well. The above-described patterns of entry into slavery through war or financial distress continued into the early colonial period; there are also accounts of Europeans launching slave raids as well. For example, in relation to the slave trade out of the Horn of Africa which formed the primary source of enslaved Africans for Indo-Muslim states, both Europeans and a variety of Muslim traders were actors in this trade, “with shared interests as well as real competition.”41 More broadly, most European involvement in the slave trade was alongside Indian merchants, similarly profiting off of famine victims from at least as early as the 1620s. Most often slaves were obtained through local suppliers, and thus in many cases, Europeans were in some ways simply inserting themselves into ongoing practices and markets.42

But despite certain clear continuities, other evidence points to dramatic shifts in this period, in particular as the context of empire shaped the market in new ways. We can get a sense for the scope of change from a few examples. For example, the entry of European traders (in particular, the Dutch and the Portuguese) into the Indian Ocean also seems to have increased the trade in enslaved Bengalis in the seventeenth century.43 The disturbance this constituted can be seen in Mughal attempts to suppress Portuguese traffic in slaves in the Bay of Bengal—one of several examples of Europeans not working through local intermediaries.44 Although this particular trade began to lessen again in the second half of the seventeenth century, other arenas of the slave trade continued to thrive. Europeans regularly purchased enslaved South Asians in Western India in the seventeenth and eighteenth centuries for both local work and export.45 French and Dutch colonies such as Mauritius, Reunion, and the Cape in the south-west Indian Ocean were receiving large shipments of enslaved South Asians by the late eighteenth century in famine years.46 East India Company records also point to the ongoing trade in enslaved South Asians within the Bengal and Madras Presidencies into the early nineteenth century.47 Furthermore, the circulation of slaves within imperial spaces was not confined to the colonies but also included the bringing of slaves back to the metropole. Thus, we see enslaved South Asians in Britain throughout the colonial period, first only in elite households but subsequently also among middle-class families,48 as well as in Portugal, in capacities such as cabinet-makers or workers on river crafts in Lisbon.49

Among the significant differences when it comes to colonial attitudes towards and actions in relation to slavery, abolitionism is a particularly notable factor, albeit often in unexpected ways. While the Slavery Abolition Act (1833) legally abolished slavery in most of the British Empire, Indian slavery was not even delegalized until the Indian Slavery Act (1843). This delay was due to several factors including the widespread belief in the relatively “mild” forms taken by slavery in South Asia, as well as concerns about the political costs of antagonizing or alienating local elites. It is for this reason that the legislation produced was not only delayed but furthermore is best described as delegalization rather than abolition. As Major notes, “while masters’ right to legally enforce ownership of human property was removed, no slaves were actively freed.”50 As a result, scholars point to the continued existence of slavery in post-1843 British India, even if it was often practiced under another name. In another significant shift, the sudden dearth of cheap labor in many plantation economies in the aftermath of abolition was remedied through the development of a system of indentured servitude, which saw large numbers of South Asians transported to locations such as Guyana and Fiji to make up for this shortfall.

Experiences During Slavery

The experience of slavery varied radically depending on both the form of slavery experienced, as well as of course the specific circumstances of the slave. In the South Asian context, while domestic slavery seems to have been the most common form over the period surveyed here, other forms include military slavery, agrarian slavery, and eunuch slavery. The different forms slavery could take depended both on choices made by sellers and owners (to castrate an enslaved boy, to train an enslaved child in particular skills, etc.) as well as by the prior life experience and social identity of the enslaved individual (who may have received education or specialized training prior to enslavement, for instance). So, for example, in the eighteenth century, Tahmās Khān recounts receiving training from his owner Mu’īn al-Mulk over the course of his childhood, representing an investment in this particular slave. In contrast, seventeenth-century Bengal-born poet Ālāol was enslaved as an adult by the Portuguese, but his extensive education in Persian, Bengali, Hindavi, and Sanskrit led to the recognition of his value and a certain privileged standing in Arakan.51

The most common form of slavery throughout this period is domestic slavery and concubinage. This has led many scholars to emphasize the need to acknowledge and explore the deep intertwining of slavery with the history of the family in this context. Thus Indrani Chatterjee, in examining the significance of domestic slavery in the royal household of Murshidabad, demonstrates how “kinless” slaves were incorporated into the complex familial political dynamics of elite households. As kinship ties in this context were sources of tension and antagonism due to the open nature of succession, many members of the royal household relied upon adopted, kinless slaves, since their allegiance could be expected due to their dependence on their masters or mistresses alone.52 Even aside from the politics of royal households and questions of succession, as Finn notes in discussing the presence of domestic slavery in Anglo-Indian households, the prevalence of domestic slavery in South Asia (versus the plantations of the Atlantic world) necessitates that we consider the significance of the family as the site of slavery and as a consequence the entanglement of these two domains of kinship and servitude.53

The status and life trajectories of enslaved women would have often been impacted by their social status prior to enslavement, their birth, their education or training, as well as their appearance. For instance, we can consider the diverse forms of domestic service in elite Rajput households in the early modern period: at the lowest rung were the davris, unskilled female domestic slaves; above these were patars (skilled slave-performers, often dancers) and olaganis (singers); then come the vadarans, the senior female slaves. Above the vadarans stood the enslaved women who had become concubines, referred to as pasvan or khavasin; those given the right to veil themselves were known as pardayats. As this terminology suggests, these hierarchical distinctions were formal within such households.54

Colonial rule introduced both changes within practices of domestic slavery, as well as the continuation of the practice within Indian and colonial households alike. To the first point, the growing intervention of colonial officials within elite households over the course of the colonial period radically transformed not only the scope of political activity of elite women but also the roles of domestic slaves and concubines as well. For example, the East India Company's ongoing interference in the functioning of the royal household of Murshidabad eventually effected a dramatic reduction in both the power and status within it of both women and slaves.55 At the same time, domestic slaves were also an integral part of British colonial households. As demonstrated by personal wills and inventories alongside public reports submitted to Parliament, there was a profusion of domestic slaves in such households despite their erasure or obfuscation in the official documentation. We see in this context both the treatment of such individuals as property, to be sold or transferred at death, as well as humanized as the recipients of bequests and manumission.56

Within the category of “domestic slavery,” those enslaved women who were concubines, as a consequence of the sexual and reproductive labor performed, were often treated differently. In the Rajput context, this is clearly expressed by the domestic hierarchy recounted by Sreenivasan, discussed above. In the Mughal context, concubines who bore children (umm walad) would have most likely possessed certain rights, in particular, the right to not be sold or given away as well as to freedom upon the death of the owner.57 According to colonial-era wills, concubines were often left substantial legacies after the death of their masters, at times including not only property and money but also other slaves. While located hierarchically below the legal wives of British colonials, such concubines were often given more rights and resources than other enslaved individuals.58 Furthermore, enslaved domestics could also forge relationships with elite women as well to improve their status. Such bonds of dependency, even as they could work to ameliorate and improve the lives of such individuals in certain ways, were put in peril when the elite patrons or masters in question passed away.59 However, the majority of domestic slaves would not have achieved such relatively privileged statuses which were, it is important to note, also accompanied by sexual labor which we must understand as non-consensual. While the sources rarely discuss the violence of domestic slavery, stories such as that of Bharati, a domestic slave beaten to death, and other archival traces of slave murders in the colonial period going unpunished, demonstrate “the continued abuse and killing of slaves.”60

Professional female performers such as singers and dancers were also often in an enslaved status. This can be seen in the Mughal period, where enslaved female performers are referenced providing entertainment within the harem space, as well as in private assemblies. They also find mention as valuable gifts, for instance, as when Babur sent dancing women (pāturs) confiscated from his defeated opponent to his female relatives.61 We also see a high value placed on female performers in the early modern Rajput context, due to the skilled nature of their labor;62 later on, we similarly find the purchase and sale of trained dancing women in late colonial Tanjore, as “ornaments of palace life.”63

Another significant category of slavery in South Asian households and courts during this period was eunuch slavery. While eunuchs—enslaved, castrated men—could perform a variety of functions within the South Asian context, they are worth considering separately due to their unique positionality within elite households. Generally castrated prior to puberty, eunuchs were set apart by their unbroken voices, their inability to grow facial hair, and a host of other physiological consequences of castration which included the inability to have children. This status allowed them to occupy positions in close proximity to elite men and women alike. Thus, in elite households through the nineteenth century, we find eunuchs as harem guards and domestic attendants, albeit they were generally not allowed free access to female spaces. They also at times rose to positions of influence, wealth, and power, like other elite slaves, likely due to the relations of trust forged through such proximate labor.64

Another form of slavery that was especially common in the early modern period was military slavery, continuing practices that can be seen in the medieval period. Eaton has described the significance of military slavery in the Deccani Sultanates, while Walker underlines the participation of the Portuguese in similar practices, albeit with substantial differences. Military slaves were often viewed as valuable due to their outsider status, serving as buffers insulating central authority from the broader webs of kinship and social connection which might imperil that authority. Thus Eaton proposes, drawing on prior work by Wink, that it is the political instability and inadequate centralization of the Deccani Sultanates in this period which makes military slavery appear here, in contrast to the Mughal Empire to the north where, although we encounter individual military slaves, the practice is not ubiquitous in the way it is in the Deccan.65 While most commonly associated with Islamicate politics, military slavery also can be seen in Portuguese India, but there are crucial differences between these two contexts. In contrast to the practices discussed above, the Portuguese only ever allowed slaves to temporarily be used in military capacities for emergencies for “defense against external aggression.” This thus differs from the more regular usage of military slavery to provide internal stability. This may reflect, per Walker, both the relative internal stability of the colonial holdings in South Asia, as well as the fear of possible rebellion by enslaved Africans.66

Agrarian slavery was also a significant feature of the period under review, although the history of this practice looks different depending on regional context. In the context of North India in Bihar, Gyan Prakash has argued for instance that a group of agricultural laborers called the kamias, which had a relationship of dependency but not slavery to local maliks (lords) in the precolonial period, were transformed into laborers subject to a form of debt bondage under British colonial rule which essentially amounted to slavery.67 In contrast, enslaved agricultural laborers in South India predated the arrival of Europeans in the region, although the evidence of such figures in this period is sparse.68 Here coerced labor took multiple forms, with some individuals being viewed as individually transferable, while other landless laborers were tied to the land and viewed as “an integral part of their owner´s landed property.”69 Looking at the context of late eighteenth-century colonial Madras, Ravi Ahuja underlines that such different forms of labor relations should not be viewed as “separate” but rather as “divergent tendencies in the development of class relations between dominant peasants and agricultural labourers.”70

While the above suggests a relatively tidy division of forms of slavery into domestic, agrarian, military, and eunuch types, in practice there was a great deal of variation within these categories. To date, scholars have emphasized the wide variety of forms of labor enslaved individuals were called upon to complete. Thus, for instance, other assorted slave occupations for the colonial period include construction work, food cultivation, animal herding, work as sailors or fishermen, various artisanal occupations such as weaving or distilling, and dockwork.71 There was furthermore much movement between categories of labor. As Andrea Major puts it, discussing the colonial context, “domestic slaves often performed limited agricultural labor, while the women of agrestic slave castes might be subject to sexual exploitation.”72 Thus the blurring of lines between different “forms” of slavery should be expected in examining the South Asian context.

Exits from Slavery

Prior to the colonial period, there is very little information about subjects such as manumission or escape from slavery. For instance, while Tahmās Khān tells us of his manumission in his memoirs, sources yield little information about the manumission (or lack thereof) for other slaves in this period, regardless of their status and prominence. To give another example, the manumission of Malik ‘Ambar is only mentioned by a Dutch writer named Van den Broecke, even as Indian sources remain silent on this point. Eaton argues that such apparent disinterest on the part of Persian chroniclers can be attributed to different understandings of the distinction between enslaved and free.73 In a context in which an enslaved man could be wealthy, powerful, and a member of the ruling class, how should we understand the meaning of manumission?

Eaton’s argument on elite slavery—that it needs to be seen as a question of “starting points” rather than ongoing status—suggests that the focus on formal manumission may be beside the point for some high-status enslaved individuals. In this model, military or other “elite” slavery can be viewed as a “self-terminating process,” rather than an “enduring condition.”74 This could of course be countered by examining examples of elite slaves who become the subject of criticism or abuse for their very status as slaves, as we see in the Mughal context, at least with respect to eunuchs who may not have had the same ability to achieve social incorporation. For example, during the reign of Jahāngīr, the forcible seizure of the eunuch Hilāl Khān’s mansion by another Mughal noble, as well as the public criticism of eunuchs building mansions at all in front of this same eunuch, points to a social position within the Mughal elite that is ambivalent at best.75 In any case, arguments about elite slavery would not apply to the majority of lower-status enslaved individuals.

From the eighteenth century onwards we have more information on questions of manumission. Shamita Sarkar describes for instance a series of sale documents in Bengali that specify in one case that manumission can occur when the enslaved girl in question (eleven years old at the time of sale) reaches the age of seventy; it also provides the option of purchasing her freedom earlier at a set price.76 Guha mentions a few cases of enslaved people in eighteenth-century western India gaining their freedom through either outright cash payments, or else through providing replacement slaves for themselves. He also notes that occasionally children sold into slavery during famine were freed or resold for a small amount to their families once grain had become cheap again, as well as how in this context marriage could annul enslavement, and thus be a form of emancipation.77 Within British colonial households, as was mentioned earlier, manumission was often granted at the death of the master, although other outcomes included the transfer of ownership or outright sale.78 In addition, in the early colonial period conversion was at times a strategy for Indian slaves in British households to obtain manumission; we also see at least one example of a slave in the Deccan escaping to Portuguese territory similarly and converting in hopes of attaining freedom, although in this case these hopes of manumission are not realized and he remains enslaved.79

We occasionally see mentions of escaped slaves in the Mughal period, for instance, in Tahmās Khān’s memoirs or the Travels of Peter Mundy.80 In eighteenth-century Dutch colonial Cochin, there are examples of fugitive slaves sometimes successfully escaping to neighboring sovereignties, although we also see their prosecution alongside those who aided them.81 We also have accounts of enslaved people escaping Portuguese territories. While from the sixteenth century onwards, the Portuguese authorities had agreements with local rulers (both Muslim and Hindu) for fugitives to be either returned or compensated for, the expansion of British India and its delegalization of slavery in the nineteenth century meant that British authorities refused to return such individuals or to offer compensation.82 Vatuk also describes numerous cases of fugitive slaves in nineteenth-century records from the Presidencies (i.e. in British India), although the violent death of one such female slave suggests the potential costs of such attempts as well as the difficulty of resistance more generally.83 While theoretically, anti-slavery sentiment and increasing legislation to protect slaves/servants can be seen as providing a tool for enslaved individuals to seek redress for such abuses in British India, such laws “were frequently ignored or evaded in practice and the courts were not always sympathetic to those slaves who sought relief through resort to the legal system, especially if the perpetrators were people of high social standing.”84

Here, it is important to not interpret events such as running away or manumission in terms of a simplistic opposition between enslaved and free status. Guha emphasizes that flight cannot always be read as an attempt to escape slavery and achieve freedom, but rather as often a way to find better circumstances in which to live as a slave. For instance, a fugitive slave might seek a place within a larger, wealthier household for protection from the claims of prior owners. Furthermore, manumission may not have always been experienced as positive, as in some instances the manumission of older slaves can be seen as a cost-saving measure that made the former slave more vulnerable, rather than a longed-for transition to freedom.85

The status of the children of enslaved women varied dramatically depending on context. In Mughal households, the children of concubines were not viewed as “disenfranchised” and enslaved but rather as members of the royal family; thus the sons of enslaved women were viewed as legitimate heirs to the throne.86 In the case of military slavery, such as that occurred in the early modern Deccan, slave status does not seem to have been heritable, but was on the contrary “one-generational.”87 This was not true for the Rajput context, where “slave progeny” was understood to be similarly enslaved as well as relatively restricted in terms of their ability to raise their social statuses. The children of concubines were not generally bought and sold, although they could be exchanged in political negotiations. In fact, the decreasing opportunities for enslavement via war or raiding in the nineteenth and early twentieth century led to the need to “reproducing servility among themselves,” reshaping gifting practices in Rajput households to ensure enslaved women were provided with male mates. This of course also demonstrates the importance of the heritability of enslaved status.88 Guha’s work on the Maratha context suggests that the children of slaves were similarly born into the enslaved condition,89 as does the documentation from eighteenth-century Bengal.90

Conclusion

The period reviewed here offers a broad range of practices and vocabularies related to slavery in the early modern and colonial periods in South Asia. Depending on the context, the language used to describe enslaved status could intersect with the language used for service, kinship, or discipleship; it could also be intentionally vague and veiled. While this provides productive ground for beginning to think through the meaning (both real and metaphorical) of slavery in a variety of South Asian contexts, it also constitutes a challenge to identifying the enslaved themselves for modern researchers. Despite these complications, a significant body of work employing a wide range of evidence has described the many forms of slavery which existed in this era in South Asia, from military slavery to concubinage to agrarian slavery. Having entered slavery due to war, slaving raids, or economic hardship, enslaved individuals might remain close to their place of origin, or in contrast be traded through interregional, regional, and local markets. On this basis, it is clear that there was no single, prototypical form or experience of slavery either prior or subsequent to colonialism. Rather, as numerous South Asianists have noted, this topic needs to be approached and understood in a highly contextualized, localized manner. Even ostensibly similar categories of enslavement such as concubinage or military slavery could appear and be experienced in radically different ways depending on, for example, possession of education or specialized training, biological reproduction, and the specificities of particular locations and historical moments. Just as we find dramatic differences in how individuals entered into and experienced slavery, so too exits from slavery could vary quite widely, from effective and formal manumission to escape to conversion. Similarly, depending on context slave status could be inherited by children or quite the opposite. While a synthetic essay cannot do justice to or provide a simple schema for approaching this topic, this chapter has sought to provide a map to the rich and growing scholarship in this field.

Notes

  1. 1.

    Richard Eaton, “Introduction,” in Slavery and South Asian History, eds. Indrani Chatterjee and Richard Eaton (Bloomington: Indiana University Press, 2006), 1.

  2. 2.

    Nira Wickramasinghe, Slave in a Palanquin (New York: Columbia University Press, 2020), 7.

  3. 3.

    See for instance Gyan Prakash, Bonded Histories (New York: Cambridge University Press, 1990), 1–12.

  4. 4.

    John Richards, “Norms of Comportment Among Imperial Mughal Officers,” in Moral Conduct and Authority, ed. Barbara D. Metcalf (Berkeley: University of California Press, 1984), 264.

  5. 5.

    Jessica Hinchy, “Between Slave and Disciple in South Asia,” in Slavery and Bonded Labor in Asia, 1250–1900, ed. Richard B. Allen (Leiden: Brill, 2022), 49–76; Vijay Pinch, “Gosain Tawaif: Slaves, Sex, and Ascetics in Rasdhan, c. 1800–1857,” Modern Asian Studies 38, no. 3 (2004): 567–8.

  6. 6.

    Sumit Guha, “Slavery, Society, and the State in Western India, 1700–1800,” in Slavery and South Asian History, 164.

  7. 7.

    Irfan Habib, “Akbar and Social Inequities,” Proceedings of the Indian History Conference (1992): 300–2.

  8. 8.

    Gavin Hambly, “A Note on the Trade in Eunuchs in Mughal Bengal,” Journal of the American Oriental Society 94, no. 1 (1974): 125–30.

  9. 9.

    Shadab Bano, “Slave Markets in Medieval India,” Proceedings of the Indian History Conference 61 (2000): 368–9.

  10. 10.

    Eaton, “Introduction,” 11. See also Scott Levi, “Hindus Beyond the Hindu Kush: Indians in the Central Asian Slave Trade,” Journal of the Royal Asiatic Society 12, no. 3 (2002): 283–4.

  11. 11.

    Indrani Chatterjee, “Abolition by Denial: The South Asian Example,” in Abolition and Its Aftermath in Indian Ocean Africa and Asia, ed. Gwyn Campbell (London: Routledge, 2005), 137–53.

  12. 12.

    Margot Finn, “Slaves Out of Context,” Transactions of the Royal Historical Society 19 (2009): 181–203.

  13. 13.

    Rupa Vishwanath, The Pariah Problem: Caste, Religion, and the Social in Modern India (New York: Columbia University Press, 2014), 25.

  14. 14.

    Finn, “Slaves Out of Context,” 202.

  15. 15.

    Sylvia Vatuk, “Bharattee’s Death: Domestic Slave-Women in Nineteenth-Century Madras,” in Slavery and South Asian History, 222.

  16. 16.

    Wickramasinghe, Slave in a Palanquin, 5–8.

  17. 17.

    Indrani Chatterjee, “Renewed and Connected Histories: Slavery and the Historiography of South Asia,” in Slavery and South Asian History, 19.

  18. 18.

    Scott Levi, “Hindus Beyond the Hindu Kush,” 280.

  19. 19.

    Ibid., 285–6.

  20. 20.

    Sanjay Subrahmanyam, “Between Eastern Africa and Western India, 1500–1650,” Comparative Studies in Society and History 61, no. 4 (2019): 808.

  21. 21.

    Richard Eaton, “The Rise and Fall of Military Slavery in the Deccan,” in Slavery and South Asian History, 117–21.

  22. 22.

    Bano, “Slave Markets in Medieval India,” 366–7.

  23. 23.

    Guha, “Slavery, Society, and the State in Western India,” 165–8.

  24. 24.

    See for example Mustaʻidd Khān, Maʼāsir-i ‘Ālamgīrī (Calcutta: Asiatic Society of Bengal, 1871), 75.

  25. 25.

    Levi, “Hindus Beyond the Hindu Kush,” 278–81; Gavin Hambly, “A Note on the Trade in Eunuchs in Mughal Bengal,” 126–7.

  26. 26.

    Eaton, “The Rise and Fall of Military Slavery in the Deccan,” 115–6.

  27. 27.

    Ibid., 126.

  28. 28.

    Indrani Chatterjee, “A Slave’s Quest,” Indian Economic Social History Review 37, no. 1 (2000): 53–86.

  29. 29.

    ‘Abd al-Hamīd Lāhorī, Pādshāhnāma (Calcutta: Asiatic Society of Bengal, 1867–1868), 1:473 and 2:193; Nūr al-Dīn Muhammad Jahāngīr, Tūzuk-i Jahāngīrī (Lucknow: Nawal Kishore, 1863), 1:81.

  30. 30.

    Hambly, “A Note on the Trade in Eunuchs in Mughal Bengal,” 129.

  31. 31.

    Chatterjee, “A Slave’s Quest,” 79.

  32. 32.

    Ramya Sreenivasan, “Drudges, Dancing Girls, Concubines: Female Slaves in Rajput Polity, 1500–1850,” in Slavery and South Asian History, 141–3.

  33. 33.

    John Richards, The Mughal Empire (Cambridge: Cambridge University Press, 1993), 14–5; Shadab Bano, “Marriage and Concubinage in the Mughal Imperial Family,” Proceedings of the Indian History Congress 60 (1999): 354.

  34. 34.

    Shivangini Tandon, “The Liminals in Mughal Households,” Proceedings of the Indian History Conference 78 (2017): 382.

  35. 35.

    Jeff Eden, “Slavery in Islamic Central Asia,” Oxford Research Encyclopedia of Asian History (Oxford: Oxford University Press, 2019).

  36. 36.

    For example, we see illegal enslavement occurring within the Mughal context, for instance in Mustaʻidd Khān, Ma’āsir-i ʻĀlamgīrī, 75.

  37. 37.

    Eaton, “Introduction,” 8; Chatterjee, “Renewed and Connected Histories,” 33.

  38. 38.

    Lionel Caplan, “Power and Status in South Asian Slavery,” in Asian and African Systems of Slavery, ed. James Watson (Berkeley: University of California Press, 1980), 192.

  39. 39.

    Leslie Orr, “Slavery and Dependency in Southern India,” in Cambridge World History of Slavery, Volume 2, eds. Craig Perry, David Eltis, Stanley L. Engerman, and David Richardson (Cambridge: Cambridge University Press, 2021); Malarvizhi Jayanth, “Abolishing Agrarian Slavery in Southern Colonial India” (PhD Dissertation, University of Chicago, 2020), 44–7.

  40. 40.

    Vishwanath, The Pariah Problem, 23.

  41. 41.

    Subrahmanyam, “Between Eastern Africa and Western India,” 815–6.

  42. 42.

    Marina Carter, “Slavery and Unfree Labour in the Indian Ocean,” History Compass 4, no. 5 (2006): 800–4.

  43. 43.

    Carter, “Slavery and Unfree Labour,” 801.

  44. 44.

    Eaton, “Introduction,” 13.

  45. 45.

    Guha, “Slavery, Society, and the State in Western India,” 168–9.

  46. 46.

    Carter, “Slavery and Unfree Labour,” 802.

  47. 47.

    Andrea Major, “Unfree Labour in Colonial South Asia,” Oxford Research Encyclopedia of Asian History (Oxford: Oxford University Press, 2019).

  48. 48.

    Michael Fisher, “Bound for Britain: Changing Conditions of Servitude, 1600–1857,” in Slavery and South Asian History, 191.

  49. 49.

    Carter, “Slavery and Unfree Labour,” 804.

  50. 50.

    Major, “Unfree Labour in Colonial South Asia.”

  51. 51.

    Thibaut d’Hubert, In the Shade of the Golden Palace (New York: Oxford University Press, 2018), 6–7.

  52. 52.

    Indrani Chatterjee, Gender, Slavery and Law in Colonial India (New Delhi: Oxford University Press, 1999), 42.

  53. 53.

    Finn, “Slaves Out of Context,” 183.

  54. 54.

    Sreenivasan, “Drudges, Dancing Girls, Concubines,” 143–4.

  55. 55.

    Chatterjee, Gender, Slavery and Law in Colonial India, 34–77.

  56. 56.

    Finn, “Slaves Out of Context,” 186–9.

  57. 57.

    Shadab Bano, “Marriage and Concubinage in the Mughal Imperial Family,” Proceedings of the Indian History Congress 60 (1999): 353 and 356.

  58. 58.

    Finn, “Slaves Out of Context,” 190–2.

  59. 59.

    Sreenivasan, “Drudges, Dancing Girls, Concubines,” 148–52.

  60. 60.

    Vatuk, “Bharattee’s Death,” 221.

  61. 61.

    Katherine Schofield, “The Courtesan Tale: Female Musicians and Dancers in Mughal Historical Chronicles, c. 1556–1748,” Gender & History 24, no. 1 (2012): 156.

  62. 62.

    Sreenivasan, “Drudges, Dancing Girls, Concubines,” 140–3.

  63. 63.

    Davesh Soneji, Unfinished Gestures: Devadasis, Memory, and Modernity in South India (Chicago: University of Chicago Press, 2012), 36–37.

  64. 64.

    Hambly, “A Note on the Trade in Eunuchs in Mughal Bengal,” 125–30; Ruby Lal, “Harem and Eunuchs: Liminality and Networks of Mughal Authority,” in Celibate and Childless Men in Power, eds. Almut Höfert, Matthew Mesley, and Serena Tolino (London: Routledge, 2018), 92–108; Emma Kalb, “Slaves at the Center of Power: Eunuchs in the Service of the Mughal Elite, 1556–1707,” (PhD diss., University of Chicago, 2020), 1–46.

  65. 65.

    Eaton, “The Rise and Fall of Military Slavery in the Deccan,” 120–2; Andre Wink, Al-Hind: The Making of the Indo-Islamic World, vol. 2 (Leiden: E. J. Brill, 1997), 180–1.

  66. 66.

    Timothy Walker, “Slaves or Soldiers? African Conscripts in Portuguese India, 1857–1860,” in Slavery and South Asian History, 235–7.

  67. 67.

    Prakash, Bonded Histories, 140–83.

  68. 68.

    Orr, “Slavery and Dependency in Southern India,” 313–32; Malarvizhi Jayanth, “Abolishing Agrarian Slavery in Southern Colonial India” (PhD diss., University of Chicago, 2020), 11–40.

  69. 69.

    Major, “Unfree Labour in Colonial South Asia.”

  70. 70.

    Ravi Ahuja, “Labour Relations in an Early Colonial Context,” Modern Asian Studies 36, no. 4 (2002): 798–800.

  71. 71.

    Carter, “Slavery and Unfree Labour,” 805.

  72. 72.

    Major, “Unfree Labour in Colonial South Asia.”

  73. 73.

    Eaton, “The Rise and Fall of Military Slavery in the Deccan,” 130.

  74. 74.

    Ibid., 121 and 129.

  75. 75.

    See Kalb, Slaves at the Center of Power, 244–51.

  76. 76.

    Shamita Sarkar, “Slavery in Late Mughal Bengal,” Proceedings of the Indian History Conference 55 (1994): 603–4.

  77. 77.

    Guha, “Slavery, Society, and the State in Western India,” 175–82.

  78. 78.

    Finn, “Slaves Out of Context.”

  79. 79.

    Fisher, “Bound for Britain,” 190–1; Ananya Chakravarti, “Mapping ‘Gabriel’: Space, Identity, and Slavery in the Late Sixteenth-Century Indian Ocean,” Past and Present 243 (2019): 24–8.

  80. 80.

    Tahmas Khan, Tahmas Nama: The Autobiography of a Slave (Bombay: Popular Prakashan, 1967), 82; Shadab Bano, “Women Slaves in Medieval India,” Proceedings of the Indian History Congress 65 (2004): 318.

  81. 81.

    Matthias van Rossum, “On the Run: Runaway Slaves and Their Social Networks in Eighteenth-Century Cochin,” Journal of Social History 54.1 (2020): 66–87.

  82. 82.

    Walker, “Slaves or Soldiers?” 245.

  83. 83.

    Vatuk, “Bharattee’s Death,” 222.

  84. 84.

    Ibid., 226.

  85. 85.

    Guha, “Slavery, Society, and the State in Western India,” 180–1.

  86. 86.

    Tandon, “The Liminals in Mughal Households,” 383; Schofield, “The Courtesan Tale,” 156.

  87. 87.

    Eaton, “Introduction,” 8.

  88. 88.

    Sreenivasan, “Drudges, Dancing Girls, Concubines,” 152–5.

  89. 89.

    Guha, “Slavery, Society, and the State in Western India,” 167–8.

  90. 90.

    Sarkar, “Slavery in Late Mughal Bengal,” 604.