Theory from the South III: On Slaves and Freedmen in Precolonial South Asia - a Conversation with Indrani Chatterjee

Last year, our sister site CSSAAME published a thematic issue on the question of comparative slavery, in which a group of scholars push back on the tendency to deploy a US-Atlantic terminology of race to premodern/precolonial contexts of slavery across the Indian Ocean and Africa. Borderlines’ Anna Reumert sat down with Indrani Chatterjee, one of the contributors to this discussion, to pick up the thread on this discussion, asking: What might a comparative analysis of transregional slaveries look like? What can we learn from reading Indian Ocean histories of slavery beyond trans-Atlantic and colonial terminologies of race and difference?

Anna Reumert (AR): What defined the enslavable in pre-colonial South Asian societies? Through what processes and practices was the slave distinguished from the non-slave?

Indrani Chatterjee (IC): Rather than a blanket condition called ‘slavery’, Persian-language epigraphs distinguished slaves by their owners: there were ‘slaves of the shrines’ (bandah-i-dargah) and there were ‘slaves of the Sultans’ (bandah-i- khas).[1]  In the subcontinent, the Persian term dargah (literally a threshold of a court) was associated from the thirteenth century with a shrine built over the grave of a revered religious leader. By the fourteenth-fifteenth century, many such dargah were part of residential-cum-learning complexes made up of a school (madrasah), a free pantry (langar) as well as a lodge or hostel (khanqah). Both Sufi teachers and warrior Sultans presided over territorial estates (vilayat) and administrations in which slave and non-slaves worked together.

In the Indian subcontinent between the thirteenth and eighteenth century, free-born people valued descent from biological ancestors (nasab) alongside lines of spiritual descent or discipleship either to a Saiva or Vaisnava guru or to a Sufi teacher and place within the respective Saiva, Vaisnava or Sufi order (sampradaya, silsilah). By virtue of living in the same households, first-generation slaves (al abd, bandah, ghulam) and freedmen (atiq) alike were identified by qualifying names such as ‘al-Qutbi’, ‘al-Shamsi’, ‘al-Sultani’ that followed a personal name. [2] This qualifier identified the slave as belonging to and in the retinue of a particular master or mistress. Merchants, scholars and saints also had slaves.[3] Freedmen (atiq) too identified themselves in terms of their erstwhile masters, some of who were merchants.[4]   

AR: What meaning was given to honor and property in enslavement?

IC: Honor and dishonor are socially and historically specific ways of reckoning personal and collective standing relative to others. In the subcontinent where multiple juridical and monastic hierarchies had co-existed since antiquity, the dignity of a ‘caste’-identified ‘high rank’ could be retained despite being the ‘slave of a Sultan’ and the honor of a lowly caste-member could be elevated as ‘the slave of a deity’. So parallel systems of dignity and honor should be kept in mind for understanding the histories of honor and dishonor too in the subcontinent before European empires. This was equally true for the Sufis and Sultans of the period after the thirteenth century. Though under-investigated for the subcontinent, gifting slaves to a Sufi ‘saint’ or pir was also known in the fourteenth century. Ibn Batutta records that the Tughlaq Sultan, Muhammad, gave a preacher from Tirmidh “a hundred thousand silver dinars and two hundred slaves, some of whom he freed and some he took with him”.[5] What this meant for the persons involved was that the person or household thus committed could not be used/abused by a third party. Thus the pir, Maneri, advised ‘’As long as all his [slave’s] qualities of service are devoted to the Lord, he cannot become the slave of anyone or anything but the Lord.’ Effectively, commitment to a pir or his estate or household implied that the committed person or household or field was the owner’s alone. Precisely for this reason, the phrase ‘bandah-i-dargah’ was a claim to dignity and inviolability. In any case, male slaves sometimes identified themselves as both ‘al-Sultani’ and pious: hence the phrase ‘the lover of saints and pious persons’.[6]

Perhaps honor varied also with the kind of work the master or mistress asked of the slave. Male and female slaves alike in the political-military households were literate: where appropriate, they were further trained as performers, accountants, soldiers and guards. Slaves worked alongside freedmen and non-slaves at all kinds of crafts, regional militias and administrations. Junior male slaves could be found working as clerks (muharrir), superintendents of boats and stables, or as soldiers.[7] The most trusted slaves of the Sultan held the highest posts in the administration, and were also endowed with titles, lands and perquisites of office. Sometimes, as in one instance from a fourteenth-century port-city in western India, the Sultan’s favored slave (bandah-i-khass) was appointed to both territorial and moral guardianship of a populace when he became kotwal and Quran-khwan at the same time.[8] As delegates of the master or mistress, such slave-officers commanded both non-slave and non-Muslim groups, such as the Rajput war-bands and clans (in western India) and the numerate kayastha, khattri and brahman jatis [castes], the Jain financiers, the Saiva warlords. All the latter groups in turn acquired slaves of their own from local sources in times of war, other crises, and as gifts. By the fifteenth century, if not earlier, Hindu Vaisnava poets too included servility (dasattva) as one of the five moods of worship. Clearly, they had recognized – and considered worth emulating – a key aspect of political behavior as appropriate behavior towards a saint or deity as well.

AR: How did sexuality and gender factor into this system?

IC: Women slaves appear to have belonged either to the gods - in the various Hindu temples in southern India -, or to the human lords, many of who in turn claimed to be ‘servants of’ Saiva deities and their deputies in turn.[9] The key reason that female slaves were valued was that in addition to their skilled and unskilled work, their progeny too added to the human resources of a household, workshop, farm and kingdom. Of course, one should not assume that all children born of slave women were also counted as ‘slaves’. That was the case in all societies shaped by principles of Roman law, in which the child of a slave-mother was inevitably also a slave. But in the subcontinent, where multiple lineage-reckoning systems appear to have co-existed, practice and Islamic laws both favored the child born of a non-slave father and a slave mother as ‘free’, and gave the slave-mother in turn a legal protection against further sale. In Muslim households in particular, in sharp contrast to the practices of both Hindu households of earlier centuries and of Atlantic plantation-owners of the seventeenth-eighteenth century, women slaves who had given birth to children with their male masters attained the status of umm-i-walad (mother of a child) which carried with it the rights and responsibilities of freedwomen. Slave status was not automatically inherited by the child born of a slave mother with a non-slave Muslim father. In practice, there was even a term for such a child; khanahzad or ‘born in the house’. Such children were considered ‘clients’ with particular claims (of subsistence, legal protection) on the master’s house as well.

AR: To the extent that there was a distinct slave status, is it possible that this status became the basis of individual and collective self-definition among the enslaved?

IC: The goals of long-serving slaves of either shrine, court or merchant household were two-fold: to transform personal titles into inheritable goods, and to earn a ‘reputation’ from the larger community beyond their own households by doing what their epigraphs called public ‘welfare’ (khaīr). The first was often achieved because the better masters remained responsible for the marriages of their slaves.

In this regard, sometimes the ‘house-born’ (khanahzadeh) could successfully consolidate their dignity by means of further employment to their parents’ patrons. For instance, a record from western India identifies a Shaban, son of a Tuhfa Sultani, who had built a mosque (in 1452 CE) and endowed it with 6 ploughs’ worth of land in the village of Rakhiyal [close to modern Ahmedabad, Gujarat]. Since the builder was also a favorite of the ruling sultan, a royal decree declared the garden as well as its trees and wells to be the perpetual property of this man’s descendants — both in the sons’ and in the daughters’ lines — and forbade the sultan’s other officials from interfering with this estate. In sum, the second generation of a slave of a sultan (Tuhfa Sultani) did succeed in establishing a domain autonomous of the sultan’s own taxation regime.

This pattern made service at the courts attractive for many non-Muslim non-slaves as well. In her study of the Mughal ateliers, Yael Rice tells us that many (presumably non-slave) painters, poets, scribes embraced the position of ‘bandah-i-dargah’[10]. They simultaneously professed ‘devotion’ to the court and claimed a particular kind of special knowledge – whether about ‘nature’ or about human ‘physiognomy’ or mystic gnosis to which only they had access. This kind of claim made the paintings such artists produced ‘masterpieces’ for which they were rewarded with titles such as Nadir al-Zaman (Rarity of the Time) and Nadir al‘Asr (Rarity of the Age). Such official titles were critical acknowledgements of skill, but they say very little about the ‘racial’ or ethnic backgrounds of their holders. Furthermore, such titles and honors did not make them into ‘nobles’ on the pattern of medieval Europe. The sociological distinctiveness of the two were also germane here: for monogamous marriage, primogeniture of succession and inheritance characterized the nobility of western and northern Europe from the tenth century, while polygyny and a variety of inheritance practices remained hegemonic in the subcontinental ruling groups.

The inheritance of masters’ trust remained important to those who were either house-born (khanazad-i-qadim) or non-slave servants (khidmatgar-i-qadim) whether of a saint or of a civic or military official. From the perspective of the client, a long tenure of service was also expected to create an obligation for the kinsmen of masters towards the client. That is why Rajput (Hindu) men also claimed the status of khanazadeh in the seventeenth century.[11] Such groups continued to provide loyal service to Mughal armies. For instance, when a Shi’a rival’s capital (Bidar) was conquered by the Mughals, the emperor appointed one of his khanahzada as the fort-commander (qilahdar) to a fort there. 

The second path of social recognition was by means of construction ‘with their own money’. Both piety and probity were important to these slaves’ claims to honor. So Mufarrah Sultani, a slave of Firuz Shah Tughlaq who served as his royal inkstand-bearer, built a mosque ‘out of his own/personal money’ (az mal-i-khalis khud) in 1361 CE. [12] Similarly, a freedman (mu’tiq-i Badr al-din Aya Rumi) who built a well in a village near Delhi in the mid-13th century, insisted that the expenses for the construction had been met out of his “own money” (az khalis mal-i-khud)[13]. The umm-i-walad behaved identically when they inherited some of the wealth of their deceased master, and constructed works of public welfare, in the early sixteenth century.[14] By the seventeenth century, the sums accumulated and spent by some favored or long-serving slaves were considerable. An example is that of the 89-year-old male slave, who had served as the ‘stirrup-holder’ for the Mughal emperor Jahangir, who saved enough (290,000 tankas) to build a mosque, a resthouse with an attached well, and enclose a graveyard in Delhi: the sum, he said, had been acquired by ‘honest means of one’s own’ (halal wajah khud). [15] The same emphasis on public welfare can be found in the constructions of stepwells and gardens sponsored by favored female slaves (khavas, pasvans, pardayat) of Rajput rulers in the seventeenth-early nineteenth century.[16]   

AR: Anthropologists have long debated the relationship between genealogy and race as systems of designating kin from stranger through blood relations. What were the signifiers of difference in pre-modern South Asian slavery?

IC: Rather than a static, closed and pan-Indian ‘system’, historians now treat the caste and sub-caste (varna-jati) order as regionally specific and historically changing method of local governance. In the Vedic texts (1500 BCE-600 BCE) ‘varna’ referred to a four-fold classification of people in a hierarchy of moral-ritual ranks, with the brahman placed first in moral rank, followed by the ksatriya, vaisya and sudra. Between 600 BCE-200 CE, there were more groups (of artisans, doctors, manufacturers, farmers) than could be folded into an ancient four-fold varna order of the texts. So jatis emerged as the locally meaningful communities of reference for the regulation of conduct (sex, marriage, feasting) and for the maintenance of the occupational practices particular to each community. It was also the community of birth. Thus, charters of the 8th-10th century also identified the non-indigenous communities (usually labelled as the yavana) as jatis. As membership in guilds and collectives became valuable in an expansive economy, membership in each clan of craft-worker, farmer or financier also became the boundary monitored by each clan and occupation’s internal administration. This was the context in which elder, senior (male and female) members of the same community policed juniors. Alongside there were monastic teachers, favored by local warlords, who adjudicated jati-standing of individuals.

The Central Asian Sultanates incorporated the spokespersons of these non-Muslims jatis into their administrations from the beginning of the 13th century, and employed some of the men of priestly, scribal, mercantile groups in their own administrations. But the Portuguese who arrived at the end of the fifteenth century in southern India were fresh from having excluded the Jews (1391) and Muslims (1492) from Iberia as a whole. While the waves of West and Central Asian men living in the subcontinent referred to other groups around them in terms of their clothing distinctions (‘zunnardaran’ for Hindu brahmans), or geographical zones (‘habash’ for Abyssinia, ‘turk’ for Central Asia, ‘dakan’ for the plateau of the peninsula), the Portuguese used blood-based terms to discuss descent groups when they used the term "casta". Iberian terms such as mulatto, creolo, mestizo reveal this pattern of social classification. The Portuguese used the same reckoning when they described the numerous jati, khail and qaum and qibla in the Indian subcontinent as so many ‘castes’. This usage racialized ‘caste’ as blood-based and somatic inheritance through the subsequent centuries.

AR: You describe here how precolonial histories of jati, qaum and biradari were racialized by the late eighteenth century as ‘caste’. Is it possible to identify processes of difference-making in slavery as ‘racialization’ before colonial intervention introduced a new terminology of race and caste?

IC: Portuguese usage identified ‘black skin’ as a sign of bad blood. But in the history of South Asian art at the same time, ‘black skin’ was not representative of evil. Jinah Kim’s work on 4th century texts (such as the Natyasastra/ NS) reveals a different semiotics of color[17]. The NS said that the color blue (nila) was reserved solely for the deity, Visnu; and the color identified as blue-black (syama) was the color of other divine forms. These codes circulated through the craft lineages and ateliers of the Sultanate and Mughal courts as well so that when the Mughal emperor Akbar sponsored the translation of the Hindu epic Mahabharata into Persian and asked his atelier to illustrate the resulting text (called Razmnama or The Tale of War), painters like Daswant, Lal, Mukund, Kesav and others followed the color codes laid out in the NS. So the painters of the Razmnama painted the deity, the 4-armed Vishnu, blue and the incarnation of Visnu, the god Krishna, as blue-black.

Portuguese ideas of race and caste underscored southern European ideas of ‘nobility’ in the seventeenth century, when some Africans were employed on plantations in the Atlantic world over the next two centuries. Thus various shades of ‘blackness’ were associated with laboring bodies at the same time that all laborers were segregated from, and socially subordinated to, the free masters. As a result, European merchant-traders in seventeenth-century India simply never identified Afroasian ministers as ‘belonging to/in a saint’s entourage’, which is the correct meaning of ‘Sayyidi’ (written in English as ‘Siddi’).[18] Moreover, while the official Persian records of the period refer to significant leaders of the community either by geographical referents - qaum-i-habashi’ [lit. the community from Habash/ Abyssinia][19]- or by their titles, the French referred to them by the color of the skin and hair. Thus while a Mughal mansabdar refers to a ‘Khwaja Muhammad Darvesh’ as a diplomatic envoy, the French refer to a ‘Siddi Darvez’ and characterize him as a “man of the black African race and… the curly hair which is so common among them”[20].

After the independence of the colonies in 1776, British laws of inheritance and labor became more emphatic about blood-based descent. A single drop of African blood ensured that the bearer of such blood was available as a ‘laborer’. A two-color scheme emerged in the United States on the basis of such logic, while in the Caribbean and southern American plantations, a multi-color hierarchy continued to shape the meanings of ‘casta’. Thus by the late eighteenth century, British colonial officials dependent on the plantations in the northern Atlantic, refined their somatic and blood-based ideology to interpret Indian ‘caste’ as well.

Anglo-French warfare of the late eighteenth-early nineteenth century clarified European demographic disadvantages vis-à-vis the human resources of African and South Asian societies of the same period. British ministers offered special privileges to British planters in the Caribbean to let British navies use the colonists’ African male slaves. But the British colonial officers in India would not let such West Indian males come to the subcontinent because they feared that in India, these men would meet with “men of their own complexion, slaves originally like themselves, holding offices of the first trust and emolument. They would be liable to be seduced by these appearances and yield to the offers of the native powers…from the ignoble rank of a slave, or the reproachful name of a Coffree [kafir], they would be elevated at once to the title of Seedy, friend of the prophet”. [21]  

However, by the inter-war years of the twentieth century, British domination of the globe allowed British officials in India to forget things that even their nineteenth-century predecessors had known. Illustrative of the former was a Major (later Sir) T. W. Haig, an assistant secretary for the government of India in its Foreign Department. He read inscriptions in central India in ways that clarified his ignorance of Islam, his contempt for slaves in general and for those of African descent in particular. When he read a stone slab set up by a male ‘bandah-i-dargah yazdani’ (slave-of-the-divine-court) named ‘Parviz, son of a Qaranfal Sultani’, Haig identified ‘Qaranfal’ (lit. clove) to be a “habashi or at least half a habashi” in order to dismiss the eulogy of Parviz, the son, as a “bombastic and ridiculous” “Persian doggerel”.[22] Only in 1964 did an independent Indian epigraphist point out that Malik Parviz, a slave of the Sultan Ahmad Shah, had been deputed by his master to an important outpost, from where sixteen years later, he had fulfilled his deceased master’s wish to build a mosque in the tomb-complex of two revered Sufi saints.[23]

AR: What can we learn from this?

IC: First, that historians of slavery need to liberate ourselves from a method of study which naturalizes Eurocentric racial imaginations, histories and categories of race, caste and labor as uninterrogated and universal grounds for theorizing about all co-existent societies of the globe. Neither Hobbes nor Marx and Engels had access to the evidence that our generations of students have.  Evidence-based thinking requires us to reimagine a world before racialized capitalism was globalized. That is why we need to cultivate a deeper and wider-angle vision of that world.

Secondly, in the precolonial Indian subcontinent, male and female slaves eventually held and exercised moral and political authority on behalf of their masters. Till the nineteenth century, color was no bar to the exercise of such powers. Thus while Du Bois’ ‘black American’ came to be cursed and spit upon by his fellows, courtly Africans in precolonial South Asia were admitted into social and religious communities with Indian Ocean-wide connections. If anything, colonial and postcolonial governments in India profited from the aura around service to kings and gods. Historians of global slavery need to understand this hierarchical trajectory well if they are to understand themselves, or global societies in the present, at all. 

Indrani Chatterjee is a Professor in History at University of Texas Austin. She holds a PhD from School of Oriental and African Studies in London. Her works centers on colonial and precolonial history in South Asia, and intersects ideas of gender and political economy.

The images were provided by permission of Kenneth X Robbins.

[1] For the phrase banda-i-khas applied to a news-reporter, see inscription dated 761 A.H./1359-60 C.E, in Q.Ahmed, Corpus, p.42; for the phrase banda-i-dargah applied to a builder of a mosque, see inscription dated 859 A.A./ 1455, in Abdul Karim, (1992) Corpus of Arabic and Persian Inscriptions of Bengal, p.129 

[2] See Shu’aib, ‘Inscriptions from Palwal’, EIM 1911, pp 1-4; J. Horovitz, ‘The Inscriptions of Muhammad ibn Sam, Qutbuddin Aibeg and Iltutmish’, EIM 1911, pp 12-34 ; G. Yazdani, ‘The Inscriptions of the Turk Sultans of Delhi Muizzudin Bahram, Alauddin Masud, Nasiruddin Mahmud, Ghiyathuddin Balban and Muizuddin Kaiqubad’, EIM 1913, pp. 13-46. 

[3] For ‘Bahadur, bandah-i-Maulana Siraj al-din’, see inscription no 10, in W.H.Siddiqi and Z.A Desai, ‘Khalji and Tughluq Inscriptions from Uttar Pradesh’, EIA&PS, 1964, pp 1-20.

[4] For Misbah, the atiq of Zainuddin Ali, see Z.A.Desai, ‘Some Fourteenth-Century Epitaphs from Cambay in Gujarat’, EIA&PS, 1971 (published 1977), pp. 1-59.  

[5] The Travels of Ibn Battuta, AD 1325-1354, ed by C. Defremery and B.R. Sanguinett and trans by H.A.R.Gibb, CUP and Hakluyt Society, 1958, vol. 3, pp. 675-66.

[6] See G. Yazdani, EIM, 1913, no. 17, pp 33-34.

[7] For Nanak Sultani and his clerk, see inscription no. 6, pp. 11-13, in Z.A.Desai, ‘Khalji and Tughluq Inscriptions from Rajasthan’, EIA& PS, 1967, pp 1-24; for a freedman –soldier, see Inscription no. 8, in Z.A. Desai, ‘Some Fourteenth-Century Epitaphs’, op cit.

[8] Inscription no. 2 b, pp 6-7 in Z.A.Desai, ‘Khalji and Tughluq Inscriptions from Gujarat’, EIAPS, 1962 (published 1964), pp. 1-40.

[9] Tiziana Leucci, ‘South Indian Temple Dancers ‘Donated’ to the Diety and ‘Donors’ for the Diety’, pp 205-252 in A. Murugan ed. (2015) New Directions in Tamil Epigraphy, pp. 205-252; for rulers claiming the status of first servants of Siva (Eklingji), see Ulrike Teuscher, ‘Creating Ritual Structure for a Kingdom: The Case of Medieval Mewar’ in B.P.Sahu and Hermann Kulke eds. (2015) Interrogating Political Systems: Integrative Processes and States in Pre-Modern India, pp. 355-384.

[10] Yael Rice, "The Emperor's Eye and the Painter's Brush: The Rise of the Mughal Court Artist, c. 1546-1627," PhD Dissertation, University of Pennsylvania, 2012.

[11] Ibid, no. 18, pp 15-16.

[12] Inscription nos 4-24, Z.A.Desai, ‘Khalji and Tughluq Inscriptions from Gujarat, EIAPS, 1962 (published 1964), pp 1-40

[13] Inscription no. 2, pp 3-4 in Y.K. Bukhari, ‘Inscriptions from the Archaeological Museum, Red Fort’, in EIAPS 1959-60 (published 1962), pp 1-22.

[14] Syed Abdur Rahim, 2000, Arabic, Persian and Urdu Inscriptions of Central India: A Topographical List, New Delhi: Sundeep Prakashan, nos. 152-153, p. 37.

[15] Inscription no.9, pp 12-15 in Bukhari, op cit.

[16] For a list of concubine builders, see Vikramsinh Rathaur, Rajasthan ki Samskrti mein Nari (Marvar ke visesa sandarbha mein) pp 22-23.

[17] Jinah Kim, Garland of Visions: Color, Tantra, and a Material History of Indian painting. Oakland, CA: University of California Press, forthcoming

[18] Memoirs of Francois Martin (1670-94) trans. and annotated by Lotika Vardarajan, vol. 1, p. 330-331.

[19] For Mirza Raja Jai Singh’s use of the phrase, see J.N.Sarkar trans (1969), The Military Despatches of a Seventeenth-Century Indian General, p.119;  for another such use, A.A.Kadiri, ‘Nizam Shahi Inscriptions from Galna’, EIA&PS, 1967, p.49

[20] For a 'Sidi Darvez', the chief administrator of Golkonda during the 1670s beseeched by the French for help, see Varadarajan, Memoirs of Francois Martin, 1:328 – 42.

[21] Secret and Confidential Letter from Lt. Walker to Marquess Wellesley, 20 June 1805, Ms 13,653-1, Walker Collection, Edinburgh University, folios 6a-6b. I am grateful to Samira Sheikh for sharing the details of this correspondence

[22] Major T.W.Haig, ‘An Inscription in the Fort of Daulatabad’, pp 21-23, EIM, 1907.

[23] A.A. Kadiri, ‘Inscriptions of the Bahmanis of Deccan, EIA& PS, 1964, pp. 21-44, esp. pp. 38-39.