“When Slaves were nobles: The Shamsi bandagan in the early Delhi Sultanate”,

Studies in History http://sih.sagepub.com When Slaves were Nobles: The Shamsî Bandagân in the Early Delhi Sultanate Sunil Kumar Studies in History 1994; 10; 23 DOI: 10.1177/025764309401000102 The online version of this article can be found at: http://sih.sagepub.com Published by: http://www.sagepublications.com Additional services and information for Studies in History can be found at: Email Alerts: http://sih.sagepub.com/cgi/alerts Subscriptions: http://sih.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. When Slaves were Nobles: The Shamsî Bandagân in the Early Delhi Sultanate* Sunil Kumar Department of History University of Delhi I The ubiquitous presence of military slaves has been a distinguishing feature of the armies and the political systems of the Abbâsid Caliphs and the succeeding Sultanates in Islamdom. In the middle ages, the Seljuqid wazir Nizam al-Mulk Tusi and the scholar Ibn Khaldun had separately commented on the virtues of the system of ’recruiting’ slaves to consolidate the political strength of a monarch. Ibn K_haldun had astutely noticed that military slaves (ghilminlmamiliklbandagdn; singular: ghulamlmamluklbandah) were valuable subordinates because they were natally alienated and socially dead.’ The process of enslavement had carried slaves far from their natal homes, and their non-existent legal and social status in their new environment increased their dependence upon their masters. In the early thirteenth century the chronicler Fak_hr-i Mudabbir noted the value of such a deracinated military group to the ruler: [slaves] are taken from their hearth, their kin and their dwellings, har chand az khanah wa aqribâ wa wilâyat-i khlid di2rtar uftad, the more valued, precious and expensive they become, qadr wa qimat wa bahd’f-yi u ziyadat gardad ....2 the further 2 Scholars familiar with the phenomenon of agrestic slavery in the southern United States, find the usage of slaves as military notables (bandagin) highly unusual in Islamdom. In particular, the commodification of the slave, * An earlier, abbreviated version of this paper was read Congress at Delhi in 1992. at the fifty-second Indian History ’ Ibn K aldûn, The Muqaddimah, trans., F. Rosenthal, ed., N.J. Dawood, Princeton: h Bollingen series, Princeton University Press, 1974 reprint, pp. 103-5, 146-49; on the significance of natal alienation and social death in slavery, see Orlando Patterson, Slavery and Social Death, Cambridge, Harvard University Press, 1982. 2 , ed., E.D. Ross, London, -i Fak h r al-Dîn Mubârak Shâh h r-i Mudabbir, Ta’rîk h Fak Royal Asiatic Society, 1927, p. 36. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 24 possessed no social status, contrasted sharply with his as a military commander. Yet, in an important standing political apparent sense, it was the shared characteristic of all slaves, as objects and commodities, socially dead and natally alienated, that lent them value as dependents engaged in military functions in the Delhi-and other-Sultanates. Not unlike the agrestic slaves in the southern United States, it was the absence of any social privilege that allowed the slave to be considered as a potentially useful subordinate. At least to the masters concerned, the menial status of the military slave in Islamdom ’qualified’ the bandah for tasks of military command and governance, tasks that individuals of free status would have willingly performed but were often not trusted to perform. Unique to Islamdom, therefore, the tasks of slaves could extend to include where the bandah the power ’to command and forbid’ freemen.’ In the ninth century A.D. the chronicler al-Tabari recalled the ’AbbAsid Caliph al-Mahdi’s (158-69/ 775-85) sentiments: The mawali [freed slaves] deserve such a treatment, for only they combine in themselves the following qualities. When I sit in public audience, I may call a mawlâ and raise him and seat him by my side, so that his knee will rub my knee. As soon, however, as the audience is over, I may order him to groom my riding animal and he will be content with this and will not take offence. But if I demand the same thing from somebody else, he will say: ’I am the son of your supporter and intimate associate’ or ’I am a veteran in your [’AbbAsid] cause [da’wil’ ... and I shall not be able to move him from his [obstinate] stand.’ I have studied elsewhere the consequences of this ’social inversion’, where military slaves, at the behest of their master, often held command over freemen.5 Here I am more interested in understanding the logic whereby from a general class of the socially unfree, some military slaves could be distinguished with more sensitive responsibilities and trusted with command over others. The study is located in the early thirteenth century during the reign of the Delhi Sultan Shams al-Din Iltutmish (607-33/1210-36), whose slaves had a large impact upon the political system of the Sultanate ’ ° 3 Several scholars have attempted to conjecture the reasons why the phenomenon of military slavery came to be ’peculiar’ to Islamdom. In particular see Patricia Crone, Slaves on Horses: The Evolution of the Islamic Polity, Cambridge, Cambridge University Press, 1980; and Daniel Pipes, Slave Soldiers and Islam: The Genesis of a Military System, New Haven, Yale University Press, 1981; see also, Maurice I. Finley, ’The spectrum of slavery: A peculiar institution?’, The Times Literary Supplement 3, 877, 2 July 1976, pp. 819-21. 4 Cited by David Ayalon, ’Preliminary remarks on the mamluk military institution in Islam’, in V.J. Parry and M.E. Yapp, eds., War, Technology and Society in the Middle East, London, Oxford University Press, 1975, p. 49. See Sunil Kumar, The Emergence of the Delhi Sultanate 588-685/1192-1286, Duke University, Department of History, Ph.D. dissertation, 1992, pp. 171-81. 5 Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 25 for the better part of the century. I have introduced my analysis with a brief historiographical section to provide a clearer context for my arguments. n historiographical attention paid to the study of the bandagân proportion to their influential role in the politics of the Delhi Sultanate in the early thirteenth century. In the light of this lacuna, the recent writings of Peter Jackson and Irfan Habib on the 4maml(4k institution’ and the ’ruling elite’ of the thirteenth century are particularly welcome additions.6 Although Peter Jackson’s work appeared The little seems to be in inverse two years before the latter was Habib’s, from the absence of any citation it appears that surprisingly unaware of Jackson’s research.’ That the two The article of Gavin Hambly, ’Who were the chihilgânî , the forty slaves of Sultan Shams al-Dîn Iltutmish of Delhi’, Iran, 10, 1972, pp. 57-62, was an earlier attempt to study the Shamsî bandagân . The author also presented a useful summary of the historiography in the field which I have chosen not to repeat. Hambly’s article ignored the monograph of S.B.P. Nigam, Nobility under the Sultans of Delhi, A.D. 1206-1398, Delhi, Munshiram Manoharlal, 1967, where, as Jackson noted, the bandagân only received a fleeting mention. Jackson’s and Habib’s writings are the most recently published full-scale treatments of military slaves in the Delhi Sultanate. Peter Jackson, ’The mamlûk institution in early Muslim India’, Journal of the Royal Asiatic Society, 1990, pp. 340-58; and Irfan Habib, ’Formation of the Sultanate ruling class of the thirteenth century’, in Medieval India 1: Researches in the History of India, , Delhi, Oxford University Press, 1992, pp. 1-21; amongst unpublished material see 1200-1750 also Peter Jackson, The Mongols and India, 1221-1351, Cambridge University, Ph.D. dissertation, 1976; and Kumar, op. cit. 7 This was unfortunate for Jackson’s research would have saved Habib from making many mistakes, especially with regard to the identification of elites. For example, Habib argued that Qutlugh K ân was a freeman (’Formation of the Sultanate ruling class’, p. 11, and fn. 61), and h went on to repeat Nigam’s error (op. cit., pp. 198-203) in confusing the slave with Qilij K ân, h the son of the free notable ’Alâ’ al-Dîn Jânî. Jackson correctly clarified the distinction between the two notables ( ’Mamlûk institution’, p. 344, fn. 16) but failed to provide the references that had guided him to this conclusion. The confusion originated from the fact that Jûzjânî provided at least three variations of Qilij K ân’s name (further compounded by h scribal foibles, for example, Minhâj-i Sirâj Jûzjânî, T irî, 2 Vols., ed., Hayy s abaqât-i Nâ Habibi, Kabul, Anjuman-i Ta’rikh-i Afghanistan, 1963-64, Vol. 1, p. 476, where Khalaj may stand either for Qilij or Qutlugh, and Jânî became K ânî). Despite the many versions of the h name, Jûzjânî never altered the attendant nisbah of affiliation, clarifying that he was, in fact, always referring to the same individual. Thus the notable was either referred to as Qutlugh nî h &K acirc;n pisar (son of) Malik Jânî (ibid., Vol. 2, p. 35), or Jalâl al-Dîn Mas’ûd Shâh Malik Jâ (ibid., Vol. 1, p. 495), or Qilij K ân Mas’ûd Jânî (ibid., Vol. 2, pp. 77-78; here, after the first h reference, Jûzjânî dropped the nisbah ). By contrast all references in Jûzjânî to Qutlugh Khân, the slave, were devoid of a nisbah (but note the reference to Tâj al-Dîn pisar Qutlu gh K ân h ave of Shamsî, the son of Qutlugh K ân, whose nisbah, Shamsi, indicates that he was a sl h Iltutmish. Diyâ’al-Dîn Bar , ed., S.A. Khan, Calcutta, Bibliotheca nî, Ta’rîkh-i Firûz Shâhî a Indica, 1860-62, pp. 24, 83; ed., S.A. Rashid, Aligarh, Department of History, 1957, pp. 29, 98. Moreover, the confederates of the notables also differed. Amongst others, Qutlugh Khân ân who was also his son-in-law (Jûzjânî, , h was allied with ’Izz al-Dîn Kushlu K aba T a q t Vol. 2, h acirc;n. At the time when Qutlug h p. 64) and Qilij K ân was allied with Tâj al-Dîn Arsalan &K h 6 Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 26 authors reached their conclusions independently is useful to us, however, since we.can better follow the presuppositions with which both independently approached the question of military slavery in the early Delhi Sultanate. These early formulations also directed the type of questions that the authors ignored or posed to their sources. Of the two authors Irfan Habib’s work is less directly concerned with the bandagân. Although he recognized that they were an important part of the ’ruling class’ of the early Delhi Sultanate, much of his argument was directed towards showing the importance of the ’free-born’ notables in the political system of the thirteenth century.’ In so far as the Shamsi bandagan, the military slaves of Shams al-Din Iltutmish, were concerned, Habib satisfied himself with a description of their actions in the politics of the thirteenth century Sultanate. His narration, however, was based on several assumptions concerning the internal organization of the Shamsi slaves and the structure of the Delhi Sultanate wherein they were deployed, and it was this second aspect which was more self-consciously articulated in his article. Although Jackson’s work was directly concerned with the bandagân in the thirteenth century, and the author differed completely from Habib in assessing the value of the sources upon which he was dependent, the two authors shared a common vocabulary in describing the structure of the early Delhi Sultanate, the context where the bandagdn need to be located. Jackson provided many reasons to be dissatisfied with the sources available to historians of the early Delhi Sultanate, and to overcome this disability he opted for a comparative analysis of the Shamsi bandagân with the Mamluk Sultanate of Egypt (648-922-23/1250-1516-17). This was convenient because the mamluk institution ’entered its heyday in the seventh/thirteenth century, with the military coup of 648/1250 in Cairo,’ a time roughly contemporaneous with the Delhi Sultanate.’’ The attraction in following a comparative methodology, however, did not originate from the close proximity of the two Sultanates in time if not space; Jackson perceived the Egyptian mamluk institution as a paradigm of military slavery against which the north Indian system could be measured. The comparative method served to highlight the point where the Delhi Sultanate institution was similar or differed from its model. Since it was the same institution, the paucity and unreliable nature of information concerning the north Indian ân and ’Izz al-Dîn Kushlu K h K ân were rebelling in the area of Samana and Kaithal, prior to h their march upon Delhi, Tâj al-Dîn Arsalan K ân and Qilij K h ân were in the region of h Awadh and Kara. It was after Qutlugh K ân and Kushlu K h ân were driven off from Delhi h (and sought sanctuary with the Mongo .[on which see Jackson, Mongols and India, p. 97, fn. ls 171]), that Tâj al-Dîn and Qilij K ân were conciliated by Balban, the latter receiving Lakhh nauti as his command (Jûzjâni, , abaqât Vol. 1, pp. 492-95, Vol. 2, pp. 34-35, 39, 73-78). T 8 Habib’s argument concerning the free-born notables coincided with Jackson’s, an analysis that I do not share. See Kumar, op. cit., pp. 110-16. 9 Jackson, ’Mamlûk institution’, p. 341. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 27 variant could be overcome by the data from Egypt. After all Sultanate chroniclers were chronically unreliable, therefore any explanation of what they were ’actually’ trying to communicate had to originate from sources which did not share these problems. &dquo;’ But the exercise remained descriptive; a fleshing out of points of similarity, and highlighting of those areas where there were variations in the institution. Jackson’s analysis left no space for the local north Indian milieu, or the process of Sultanate history which might have left its own mark on the ’institution’ of military slavery. When Jackson did hazard an explanation for a ’curious’ practice regarding the bandagin in the Delhi Sultanate, there was always a sensible explanation available in the voluminous sources of Mamluk Egypt. Thus the chihilgtini or the ’forty slaves’ of Sultdn Iltutmish defied explanation, until Jackson provided a possible solution by comparing them with the amir tablkhdna, or commanders of forty horses in Mamlûk Egypt.&dquo; From a different perspective, Jackson’s interest in the emergence of the mawlazâdagân (sons of free slaves) in the reign of Sultdn G_hiyas al-Din Balban (664/1266-685/1286) only went so far as to note that their deployment remained departures from the ’model’ mamluk institution of Egypt.&dquo; In other than a study of details (especially concerning the prosopographical problems of identifying nomenclature and titulature-a valuable exercise in itself) Jackson’s comparative method curtailed his analysis by limiting the questions that he posed to his Sultanate sources. Working within the vocabulary of ’institution’, Jackson’s analysis was unable to study process and change within the cadre of Shamsi bandagan. This was particularly unfortunate since Jackson did point out that the Shamsi bandagin were not an undifferentiated monolithic mass, that hierarchy did distinguish slaves from each other. But the presence of this hierarchy was merely stated as a property of the ’institution’, and no explanation or rationale was offered by the author for its presence.&dquo; As a result when ’fratricidal’ conflict split the bandagdn cadre after Iltutmish’s death, and junior Shamsi slaves eventually rose to positions of political power, this was merely treated as a phenomenon peculiar to the ’institution’ of military slavery in the Delhi Sultanates. 14 Change came about in ’the 10 In support of his ideas Jackson cited the writings of Peter Hardy, Historians of Medieval India, London, Routledge and Kegan Paul, 1960; and Jackson, ’Mamlûk institution’, p. 344, fn. 15; see also Peter Hardy, ’The oratio recta of Barani’s Tarikh-i Firuz Shahi —Fact or fiction’, Bulletin of the School of Oriental and African Studies, 20, 1957, pp. 315-21 and his ’The Muslim historians of the Delhi Sultanates: Is what they say really what they mean?’ Journal of the Asiatic Society of Pakistan, 9, 1964, pp. 59-63. " Jackson, ’Mamlûk institution’, pp. 345-46. 12 ., pp. 353-54. Ibid 13 Ibid., pp. 347-48. 14 Jackson noted that, ’the sense of solidarity among the Shamsis—what the Egyptian Arabic sources term , ushdâshiyya the group feeling conventionally associated with the h k slaves of the same master—was remarkably absent ...’, ibid., pp. 351-52. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 28 institution’ only when the ’Turkish slaves’ were swept aside with the ’revolution’ of 689/1290.’S It needs to be remembered, however, that the emergence to power of younger and junior Shamsi slaves in the post-Shamsi period was of great significance, because these new entrants into the political arena were also relatively unimpressed by the existing traditions of the Delhi Sultanate. Conflict and change in the composition of the Shamsi bandagân did not imply a mere mechanical exchange of one group of power brokers for another; it also empowered individuals who did not share the unitary vision of the Sultanate emanating from Delhi. Changes in the composition and status of slaves within the bandagin cadre (and the larger ruling elite) materially affected the structure of the Delhi Sultanate. Because of these changes, when Sultans like Ghiyas al-Din Balban and ’AIA’ al-Din Khalaji (695/1296-715/1316) sought to reimpose the authority of Delhi, they were forced to intervene in areas and take political measures which took the Sultanate in directions very different from those envisaged by Iltutmish. Since the ’mamli2k institution’ was not a dynamic, evolving cadre of military slaves, Jackson could not study them as a historically changing phenome- non. Jackson’s work, however, was far from being unique in carrying a baggage of synchronous presuppositions. Although he does not use the term, for Habib it was the ’institution’ of the iqtd’, a transferable revenue assignment which was a determinant feature of the Delhi Sultanate political system.’6 The model for this ’institution’ was provided in the Siyâsat nimah of Nizam al-Mulk, whose normative statements were cited at length by Habib in an earlier work.&dquo; Together with the iqt,!’, Habib argued that the character of the ’ruling class’ was the other distinguishing feature of the Sultanate, and he traced the changes which had occurred amongst the elites from the ShansabAnid, to the Shamsid, the post-Shamsid, into the Khalaji period. For Habib, however, ’the substance of power and wealth was, of course, represented not by the titles but by the iqii2s or revenue and military charges. ’18 Yet, the author left unclear how changes within the ’ruling class’ affected the political system of the Sultanate; did they affect the nature of the iqta‘, or in a variation of Foucault’s argument, did the ’institution’ have an autonomous life of its own? As a matter of fact, Habib ., p. Ibid 15 355. Ifran Habib, ’Formation of the Sultanate ruling class’, p. 1. " Irfan Habib, : I &’ t acirc;’s q Distribution of revenue resources among the ruling class’, in Irfan Habib and Tapan Raychaudhuri, eds., The Cambridge Economic History of India, Vol. 1: c. , Cambridge, Cambridge University Press, 1982, p. 68. 1200-1750 18 Irfan Habib, ’Formation of the Sultanate ruling class’, p. 11. This idea was also shared by Peter Jackson who regarded the iq â’ as one of two ’institutions’ brought to north India from t the Islamic world (the other being military slaves). Hierarchy within cadre of slaves was also made apparent by the grant of an iq â’ which indicated ’prominence in the state appart atus ....’ Jackson, ’ Mamlûk institution’, pp. 340, 347. 16 Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 29 clearly identify the agents of change, although the Delhi Sultans frequently stated to be the ’makers of history’. He charted the history of the iqt,!’ elsewhere: did not were First of all, the Sultans from Iltutmish onwards enforced the practice of in spite of discovering great abuses, Balban transferring muqti‘s did not seek to abolish the assignments .... Major changes occurred Qhiyâsuddîn Tughluq had during the reign of ’Ald’u’ddin Khalji no radical changes to introduce in this system .... Under Muhammad Tughluq we find a further extension of the control of the sultan’s government .... Firuz Tughluq’s policy was to assign away lands in iqta‘s .... Under the L6dis, the system remained essentially similar ....19 ... .... .... For Habib, the institution of the iqj£ ’ and the rulers remained the agents that enforced ’centralization’, a measure which was from time to time resisted by the notables. The primacy of action, however, remained with the energetic Sultan, and in a variation of the well-worn theme, if an ambitious, capable monarch sought to impose his will upon his subordinates, he relied upon the iq{â’. Why, even Mu’izz al-Din Ghûrî, after he had received the appanage of Ghazni from his brother, learnt about the authoritarian traditions of the Ghaznavid SultAns, through osmosis as it were, and proceeded to assign iqta’s to his military slaves in north.India!z° In this analysis, for considerable periods in Sultanate history, members of the ’ruling class’, rather ironically, possessed no autonomy of independent action. Devoid of any historical agency they were ’transferred’ from one appointment to another. That this was manifestly untrue is brought out by the example of ’Izz al-Din Husain Kharmil who was not ’transferred to Gurzawdn in northern Afghanistan’ from Sialkot, as Habib would have it, but was actually ’one of the Maliks of Gurzawdn’, who together with his retainers, chose to leave the service of Mu’izz al-Din and retire to his home district.&dquo; If this is a convenient reading of the source, Habib’s own 19 Irfan Habib, , I &’ t acirc;’s’ q pp. 69-71, 73-74. Irfan Habib, ’Formation of the Sultanate ruling class’, pp. 6-8. Since he published his article on the ’Iq â’s’ in 1982, Habib’s opinion on the evolving character of the revenue t nids as a â’s under the Shansab t assignment has also changed. Earlier he had regarded the iq ’autonomous principalities’ (p. 69). By 1992 in the ’Formation of the Sultanate ruling class’, he referred to them ’as the classic device for centralization in Islamic states’, and implied that it was one of the institutions used to counter ’the Ghorian polity . . .based upon the clan and 20 family system’ (p. 6). For Irfan Habib’s version see ’Formation of the Sultanate ruling class’, p. 5 and contrast az k az Mulûk-i Gurzawân bûd abaqât Vol. 1, p. 402: ’ T idmat-i Sultân h Jûzjânî, , wa hashm Gurzawân lashkar batamâm birûn âmad a t baraf-i raft, [he] [Mu’izz al-Dîn] with all [his] soldiers and retinue [he] left the service of was from the Maliks of Gurzawân Sultân [Mu’izz al-Dîn] [and] departed towards Gurzawân.’ 21 with .... ... ... ... Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 30 narration of Bakhtiyâr Khalaji shows how free notables picked and chose the best possible career opportunities available. Bakhtiyâr Khalaji sought service in turn with Sipahsalar Hizabr al-Din Hasan Aznab of Budaun, Malik Husam al-Din Ughul-Beg of Awadh and eventually Qu~t~ al-Din Ai-Beg.11 Juzjani himself notes how his father possessed sufficient initiative to leave Lahore and return to Bamian without so much as informing and getting the permission of the Ghurid Sultan, bî’ijâzat-i Sultin Ghiyith alDin bahadrat-I Bâmiân raft.23 All the examples that I have cited refer to ’free notables’, the term ’free’ carrying with it the implication of ’freedom’ to choose areas of service. By default, if nothing else, one could argue that the unfree must have then lacked this social initiative, a conjecture apparently borne out by the history of deployment of the Mu’izzi and Shamsi bandagin during the reigns of their respective masters. With only one recorded exception, Iltutmish’s military slaves abided by the trust reposed in them, a loyalty which was in marked contrast to their intransigence during the reigns of their master’s successors. But the capacity for intervention in Sultanate politics was not gained overnight by the Shamsi bandagan. Already during their master’s reign, senior military slaves possessed political power together with the ability to exercise it without constant supervision, especially in distant areas like Lakhnauti, Multan and Uchch. The key to understanding how the slaves could be deployed as useful subordinates, therefore, lies in the contrary virtues ingrained in their unfree status. On the one hand, they lacked the freedom to choose their own areas of service, but on the other, the mere act of deployment implied that the Sultan trusted them with considerable local initiative. Not all Shamsi bandagdn possessed the Sultan’s confidence, and as part of Juzjani’s information, summarized in Table 1, suggests, many slaves were never trusted or given independent responsibilities. The different manner in which the Sultan treated his bandagân is noteworthy only if we recognize that Iltutmish himself perceived the capacity for autonomous decision making that military command offered to his subordinates, and thus his selective discrimination in ’honours’ dispensed to the unfree. If we recognize that military notables also possessed the agency of action (potential or otherwise), then questions relating to the Sultdn’s choice of one slave over another get to be relevant. These queries do not arise in the writing of Habib and others since the despotic monarch or the institutions of the Sultanate are the only historical actors. - 22 Irfan Habib abbreviates much of this, but see ’Formation of the Sultanate ruling class’, pp. 7-8, and Jûzjâni, , abaqât Vol. 1, pp. 422-24, 427. The details of the transfer of T allegiance from Husâm al-Dîn Ughul-Beg to Qutb al-Dîn are unclear. Jûzjânî suggests that veto offer allegiance and plunder to Qutb al-Dîn. Yet the Bakhtiyâr Khalajî took the initiati chronicler’s effusive account of the reception and the honour given to the K alajî commander h leads one to suspect that it may well have been Qutb al-Dîn Ai-Beg that sought to recruit ., Vol. 1, p. 423. ibid tiyâr h Bak alajî. See in particular the account in h K 23 Ibid., Vol. 1, p. 389. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 31 Neither Habib nor Jackson is entirely novel in presuming the overpowering presence of the despot and the state in pre-Mughal India. Scholars such as Nizami and Ashraf, writing on the religious and social history of the Delhi Sultanate worked with similar presuppositions.24 All of these scholars, for example, used the term ’noble’ to refer to the assorted groups that comprised the ’ruling class’. The manner in which these scholars understood the term ’noble’ was clarified by Athar Ali, who suggested that the term should be removed of its European feudal connotations, and be taken to mean the officers of the king, a superior class in the political order. 2:’ As he admitted, this had been the conventional translation of the term Umarâ’ in the past, and he perpetuated the tradition which has continued unchanged to the present. From Habib’s and Jackson’s usage, those individuals who were politically enfranchised by the Sultan and given iqtâ’s, were identified as ’nobles’. In Habib’s logic it followed that when new groups of people were recruited by the Sultan, there was a conflict between ’old’ and ’new nobles’; thus, some of the Khalaj, primordial residents of Afghanistan, and some Ghurians were the original Mu’izzi ’nobles’, and it was ’natural that the sudden rise of the Turkish slaves (in the Ghaznavid appanage) should draw the hostility of the old Gh6rian and Khal[a]j nobles 121 (emphasis 24 K.A. Nizami, Religion and Politics in India during the Thirteenth Century, Delhi, IdarahAdabiyat-i Delli, 1974 reprint, p. 24; Kanwar Muhammad Ashraf, Life and Conditions of the People of Hindustan, New Delhi, Munshiram Manoharlal, 1988 reprint, p. 84. 25 Athar Ali, The Mughal Nobility under Aurangzeb, New Delhi, Asia Publishing House, 1970 reprint, p. 2. 26 Irfan Habib, ’Formation of the Sultanate ruling class’, p. 8. Habib has an interesting theory on the K alaj. On the basis of fifteenth century Sati inscriptions in Devanagari where h the aK of a &K icirc;, lc Habib suggested that h alic K iicirc; &K lc and h h lajî Malwa are referred to as h alaj. He substantiated his argument by citing the sevenh might be the correct reading of K i teenth century chronicle of Lâhorî, Pâdshâhnâma , in which one of the areas near Bust, where the K h K a lic Habib, therefore, argued for an alaj had resided in the past was recorded as . h K aa lajî, laj/K K i & a icirc;/K lc rather than h Afghan origin, and an indigenous reading of h where ’ ch came to be [mis-]read as ’j’ by the Persian chroniclers. Habib’s argument could have been considered seriously if the recorded divergences in orthography were contemporaneous with the thirteenth century. The unacceptable implication of the author’s argument is that every chronicler writing in Persian in the thirteenth, and the next four centuries, quite remarkably made the same mistake in consistently writing ’j’ for ’c ’ in K h alajî. It is also not h true that ’in the thirteenth century no one spoke of them as Turks’ (ibid., p. 3) since a passage from Muhammad b. Najib Bakrân’s Jihânnâma , written circa 1200-1220, clearly recognized their Turkish background (cited in the anonymous, H udûd al-’âlam: The Regions of the World, A Persian Geography, trans. and explained by V. Minorsky, London, Luzac and Co., E.J.W. Gibb Memorial, new series, Vol. 11, 1937, p. 348). Moreover at the turn of the thirteenth century the Il K ân, Rashid al-Dîn Fadl Allah, counted h azan K h nid wazîr of G a h Turkish tribes (see Rashid al-Din Fadl Allâh, Jâmi the the a Oghûz twenty-four amongst K h laj ichte der Oguzen des Rasid ad-Dîn h , Vienna, Osterreia, ’ h l-Tawârik trans. K. Jahn, Die Gesc chische Akademie der Wissenschaften, Philogische-Historische Klasse, Denkschriften 100, alaj as Oghûz Turks is not proof in itself h 1969). The fact that these later sources refer to the K eK h of their antecedents, and as Bosworth and others have argued t alaj were probably early h migrants into the Trans-Oxania, Khurasan and Afghanistan regions, perhaps a part of the Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 36 added). Since the term ’noble’ refers to a politically superior class created by the arbitrary whim of the despotic Delhi Sultan, no further reflection concerning why specific groups of people were empowered was deemed necessary. As a result the history of the ’nobility’ does not progress beyond a mechanical review of a conflict between political competitors, either ’old and new nobles’ or the ’crown and nobility’. By contrast, in the history of Europe in the middle ages, because the with it the sense of an inherited status in the social which was protected by law, the history of nobility led to studies on the family, the development ,of monogamous marriages, inheritance and primogeniture, church and the legitimization of temporal authority, the emergence of knights and the ideal of chivalry, a term noble carried order, a superior position More relevant is the fact that in the period between the tenth and thirteenth centuries some chroniclers referred to them as Turks, others did not. Jûzjânî, for example, never identified K alaj as Turks, but he was also as careful not to refer to them as h ’Afghans’; they were always a category apart from the Turks, Tâjîks and Afghans. Yet, one of the ways the Delhi Sultâns of the K alajî dynasty expressed status was through a Turkish h titulature—for example, Ulug , Alp, Yug h rush, Qutlug h h Tegin, Ikit. The confusion concernis perhaps best resolved by Mahmûd Kashg ing the backgound of the ârî who did not h include the K tribes, but listed them among the twentyalaj among the twenty-two h four Og ûz-Turkmân tribes, where Turkmân meant, ’like the Turks’. In other words, Kashh K a laj did not belong to the original stock of the Turkish tribes, but had h &g acirc;rî felt that the h associated with them, and therefore, in language and dress often appeared ’like Turks’. Much of this information has been summarized by V. Minorsky, ’The Turkish dialect of the K alaj’, h Bulletin of the School of Oriental (and African) Studies, 10, 1940, pp. 426-34; and in his , pp. 347-48; but see also, R. Dankoff, ’Kashghari on the udûd al-’âlam commentary on the H tribal and kinship organization of the Turks’, Archivum Ottomanicum, 4, 1972, pp. 32-33; and other than C.E. Bosworth, The Ghaznavids: Their Empire in Afghanistan and Eastern Iran, Beirut, Librairie du Liban, 1973, pp. 35-36, 210-18; see idem, ’Al-Xwarazmi on the Peoples of Central Asia’, Journal of the Royal Asiatic Society, 1965, pp. 8-9; G. Doerfer, aladj’ in Encyclopaedia of Islam, Leiden, E.J. Brill, second edition, 1956, Vol. 4, pp. h ’K 917-18. It needs to be also noted in passing, that although Habib finds the apocryphal genealogical claims of the K aiajîs of Malwa from Qulij K h ân, the son-in-law of Chingiz h ân, amusing (’Formation of the Sultanate ruling class’, p. 4, fn. 18), we should remember h K the significance that the title , ân/güregen the ’royal son-in-law’ had for the Mongols, and h gurk the emphasis that Timûr placed upon it as a principle legitimizing his right to authority (amongst others, see, Gerhard Doerfer, Turkische und Mongolische Elemente im chen, Wiesbaden, Franz Steiner Verlag, 1963, Vol. 1, pp. 434-35, s.v. ’qûrîqân’ ; and Beatrice F. Manz, The Rise and Rule of Tamerlane, Cambridge, Cambridge University Press, 1990 reprint, pp. 14-16, 57). The apocryphal genealogy of the K alajîs of Malwa claiming descent h from a ’royal son-in-law’, should be seen instead as an ideological effort to claim greater status in north India especially after the removal of the Sayyid dynasty (855/1451) which had in its own time claimed a right to rule through its Timûrid connections. At least one of the Sayyid Sultâns, K ân (817-24/1414-21), had the k h idr K h utba read in the name of Shâh Rûk h h b. Timûr (for an English translation of the khutba see T.W. Arnold, The Caliphate, London, Routledge and Kegan Paul, 1965, pp. 113-14; and V.V. Barthold, ’Xalif i Sultan’, in Sochineniya, Moscow, Akademiia Nauk S.S.R., Institut Narodov Azii, 1966, Vol. 6, pp. 48-49). Hepthalite confederacy. alaj h K ûz h Og Neupersis- Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 37 study of the manor and serfdom.27 In other words, the study of the nobility did not merely belong to the realm of political or administrative history. Since it was an integrated aspect of the social and cultural life of the period,, the nobility could also provide the entry into the study of a larger associated universe. It is not my claim that the usage of the term ’noble’ should make it possible for scholars, of the middle ages to research in north Indian history the same set of questions that their colleagues in European history are engaged with. My argument is to the contrary: since the term ’noble’ is borrowed with such a limited set of implications from European history, and since it sits so uneasily in the social milieu of pre-Mughal India, it should not just be used with fair greater care, but that the term itself obscures rather than reveals the unique elements of medieval social and political life .21 There is no doubt that during the Mughal period the 27 Marc Bloch, Feudal Society: Social Classes and Political Organisation, trans, L.A. Manyon, Chicago, University of Chicago Press, 1961, Vol. 2, pp. 283-344, summed up the scholarship of his generation. For useful overviews of current research on nobility in western Europe during the middle ages, see Lawrence Stone, ’Prosopography’, in F. Gilbert and S. Graubard, eds., Historical Studies Today, New York, Norton, 1972, pp. 107-40; George Beech, ’Prosopography’, in James M. Powell, ed., Medieval Studies: An Introduction, Syracuse, Syracuse University Press, 1976, pp. 151-84; Bernard S. Bachrach, ’Some observations on The Medieval Nobility: A review essay’, Medieval Prosopography, 1, 1980, pp. 15-33. Three collections of essays provide an introduction to the major issues of interest in the field: Fredric L. Cheyette, ed., Lordship and Community in Medieval Europe: Selected Readings, New York, Holt, Rinehart, and Winston, Inc., 1968; Timothy Reuter, ed. and trans., The Medieval Nobility, Amsterdam, Europe in the Middle Ages, Selected Studies, Vol. 14, North-Holland Publishing Company, 1978; and Georges Duby, The Chivalrous Society, trans. Cynthia Postan, Berkeley, University of California Press, 1977. For nobility in Germany see in particular, , Oxford, Clarendon Press, 1985; and John Benjamin Arnold, German Knighthood 1050-1300 B. Freed, ’Nobles, ministerials, and knights in the Archdiocese of Salzburg’, Speculum, 62, 1987, pp. 575-611; and idem, ’The formation of the Salzburg Ministerialage in the tenth and eleventh centuries: An example of upward social mobility in the early Middle Ages’, Viator, Medieval and Renaissance Studies, 9, 1978, pp. 67-102. For France see Constance B. Bouchard, ’The origins of the French nobility: A reassessment’, The American Historical Review, 86, 1981, pp. 501-32; idem, ’Consanguinity and noble marriages in the tenth and eleventh centuries’, Speculum, 56, 1981, pp. 268-87; Georges Duby, The Early Growth of the European Economy: Warriors and Peasants from the Seventh to the Twelfth Century, trans. Howard B. Clarke, New York, Cornell University Press, 1974; idem, The Age of the Cathedrals: Art and , trans. E. Levieux and B. Thompson, Chicago, University of Chicago Society, 980-1420 Press, 1981; idem, The Three Orders: Feudal Society Imagined, trans. A. Goldhammer, Chicago, University of Chicago Press, 1982 reprint; and idem, The Knight, the Lady and the Priest: The Making of Modern Marriage in Medieval France, trans. B. Bray, New York, Pantheon Books, 1983. 28 Probably the most confused usage of ’noble’ is to be found in the work of Nigam, op. cit. According to him the status of the ’nobility’ was not determined by birth (pp. 2, 97) or feudal privileges (pp. 93-103), but the author does not find it necessary to explain why the individuals he was describing should be then referred to as ’nobles’. To add to the confusion Nigam goes on to say that the nobles also constituted one element of the ’bureaucratic’ organization of the Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 38 ’ennobled’ by the pâdishâh, but we tend to assume that the political status derived through service with the state automatically provided the officer with a position in local society. On the other hand, if the term noble implies an inherited social and political status protected at least by customary law, then the ’noble’ in Mughal India was not the man+abd§r whose property could be escheated, but the zamîndâr chieftain whose rights were protected, amongst other things, by the watan jdgir.11 Despite his possible low ranking within the echelons of mansabdirs, it is not unlikely that the amir who was also an autochthon carried with him a potentially greater political and social clout. In other words, the understanding of Mughal mansabdârs as ’nobles’ only provides us with an entry into an imperially prescribed system of ranks; it does not enlighten us in any way about the manner in which north Indian society understood and respected hierarchical distinctions in the Middle Ages.30 In the light of my particular historiographical concerns here, it is certainly a complete inversion of social hierarchy to refer to slaves as ’nobles’. As I mansabdârs were Sultanate (see pp. 106-10). K.A. Nizami, ’The early Turkish Sultâns of Delhi’ in Muhammad Habib and K. A. Nizami, eds., A Comprehensive History of India: The Delhi Sultanate, New Delhi, People’s Publishing House, 1982 reprint, Vol. 5, pp. 224-26, goes further and refers to ’an all-India military-cum-administrative service manned entirely by foreigners ... Turkish slave officers and Tâzîk’ ; for similar sentiments see also Muhammad Habib, ’Intoduction to Elliot and Dowson’s History of India, Vol. 2’, in K.A. Nizami, ed., Politics and Society during the Early Medieval Period, New Delhi, People’s Publishing House, 1974, Vol. 1, pp. 103-10; Irfan Habib carries the same confusion in his ’Formation of the Sultanate ruling class’ when he refers to the manner in which ’nobles’ were ’transferred’. 29 On the watan jâgîr see Ali, op. cit. , pp. 79-80. Note also Firdos Anwar, ’Implementation of escheat under Shah Jahan, some implications’, Proceedings of the Indian History Congress, 52, 1992, pp. 266-73, where the author mentions in passing that ’in India nobility had no base in land and derived power mainly from its military contingents ...’ (p. 269). 30 As scholars researching the medieval period, we should perhaps also worry about the disjunction between the social and the political in our historiography. Our current emphasis upon a study of the state and its ’nobility’, which had only an exploitative relationship with the masses, allows for a facile reinterpretation of our data by communalists to suggest that the Mughals were, and remained, ’foreign conquerors’ in ’India’. Coercive power was, however, only one element of Mughal governance, and other than mansabdârs many other social groups participated for different reasons in supporting the ’apparatus of the Mughal empire’. We know very little about these groups, their social backgrounds or their links with the institutions of the state. Useful early attempts were Iqtidar Alam K an, ’The middle classes in h the Mughal empire’, Proceedings of the Indian History Congress, 36, 1975, pp. 113-41; and Norman Ziegler, ’Some notes on Rajput loyalties during the Mughal period’, in John F. Richards, ed., Kingship and Authority in South Asia, Madison, University of Wisconsin Press, 1978, pp. 215-51. Some of the historiographical problems in using the term ’nobles’ were also noticed by Frank Perlin who did not finally alter his own usage. Frank Perlin, ’Of white whale and countrymen in the eighteenth-century Maratha Deccan: Extended class relations, rights and the problem of rural autonomy under the old regime’, Journal of Peasant Studies (henceforth JPS), 5, 1978, p. 224, fn. 19; see also his criticism of Irfan Habib in ’Concepts of order and comparison, with a diversion on counter ideologies and corporate institutions in late pre-colonial India’, JPS, 12, 1985, pp. 111-15. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 39 hope were they to point out in the remainder of the paper, however, military slaves the favoured subordinates of Shams al-Din Iltutmish precisely because were not ’nobles’. III The Mu‘izzi and Shamsi bandagân must have been acquired in a variety of ways. It is likely that the majority of the slaves may have been captured in war during the campaigns of the Sultans in the marches of Afghanistan and north India. But the contemporary sources of the early thirteenth century only provide us with information concerning a relatively small number of the Sultin’s larger bandagin retinue. None of these were, curiously enough, seized as captives. Chroniclers such as Fakhr-i Mudabbir and Juzjani provide us information only about those slaves who were purchased, and it was some of these slaves who were later deployed in the more strategic commands within the Sultanate.&dquo; Some of these military slaves, like Taj alDin Ilduz, were purchased when they were extremely young, khurd s§1 bud. Others, like Qutb al-Din Ai-Beg and Iltutmish himself, were acquired as youths.32 Their ages and method of acquisition notwithstanding, all bandagan through a similar course of training before they were eventually raised up the ranks. The intention was to create new identities for the slaves where their dependence upon their master was absolute. To achieve this, the alienation of the slave from his natal surroundings and the related absence of social moorings in his new environment were crucial. In their stead, newly fostered ties and opportunities in the service of the monarch enticed the slave into an exotic world. The Sultan sought to ensure that his dependents from alien cultural environments ’fit’ into their new social and religious surroundings. As a result, together with the crucial training in arms and deployment with the went 31 The richest source for information on the Mu’izzî and Shamsî slaves is Jûzjânî ., abaqât T and Mudabbir, Tarîk abaqât-i Nasiri, Jûzjânî provides . In the twenty-second section of the T h the biographies of twenty-five Shamsi slaves, some of whom were important in Iltutmish’s reign, others who became influential in the years after their master’s death. Here Jûzjânî also mentions the agents through whom Iltutmish purchased these slaves. (This information has been summarized together with the career of the bandagân in Table 1 of this paper.) Even where complete information is not provided concerning the nature of their acquisition, we can assume for the larger part that these slaves were also purchased because they were Turks. Since neither the Shansabânid Sultanate of Ghûr nor the Delhi Sultanate bordered Central Asia, these Turks must have also been brought to Ghazni and Delhi through intermediaries. For an account of the process of enslavement and subsequent sale, see the example of Iltutmish in Jûzjânî, ., abaqât Vol. 1, p. 441, and compare with an earlier example: the T Ghaznavid Sultân Sebuktegin’s personal account, in M. Nazim, ’The Pand-Nâmah of Subuktigin’, Journal of the Royal Asiatic Society, 1933, pp. 609-14, 621-23. 32 On Ildûz, Ai-Beg and Iltutmish, see Jûzjânî, ., abaqât Vol. 1, pp. 411, 416 and 442, T respectively. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 40 important that the slave be instructed in the religion and of his new world. This religious instruction was introduced after the slave had been captured or purchased by the Sultan, and this was certainly true for those bandagdn acquired as children. Otherwise, at least in the case of those slaves that were purchased as youths, this aspect of the training of the slave would have been carried out by the slave merchant or a previous master. A slave who possessed preliminary military and religious training commanded an excellent price in the slave market. Outb alDin’s value, for example, accrued from the religious training that he had received earlier from the noted Hanafi jurist the Qidi al-Quddt Fakhr alDin ’Abd al-’Aziz Kufi.33 This preliminary introduction to the Islamic world was of enormous value, for it implied that the slave possessed the rudimentary etiquette to be allowed into the imperial court where the Sultan could interact with his slave without any social discord. A secondary effort was also directed to seduce the loyalty of the slaves by holding out promises of wealth and high office possible in the service of the monarch. It helped to draw the contrast between the lives of the slaves in their former environment while they were free and as slaves in the service of the Sultan. Fakhr-i Mudabbir presented the imperial case faithfully. Referring to Turkish slaves, the author noted that in the steppes the Turks were nondescript, indistinguishable from each other in power and wealth, but when they came within the Muslim world they became commanders and generais.’’ The ideal distance between the steppe and the ’Islamic world’ that Fakhr-i Mudabbir was hinting at was not merely spatial but also cultural. At least according to that author the ability to renounce heathen practice and natal roots were ethnic qualities peculiar to the Turks. He noted that, ’when their hearts turn to Islam, chunan dil Muslimânî banihand, they do not remember their homes, their place of origins or their kinsmen, khanah wa ji’i wa aqribâ besh yid nakunand.’35 I am sure, however, that few slaves, Turks or non-Turks, were quite able to see the experience of enslavement as a chance to make good in new surroundings. Fak_hr-i Mudabbir, writing for Qutb al-Din Ai-Beg, a patron of Turkish, slave antecedents, definitely overstated his case; there were examples of discord in the imperial court from the Ghaznavid period caused by Turkish slaves who lacked sufficient decorum.36 Yet, Fakhr-i Mudabbir was correct within a more limited context. The process of training was certainly aimed at creating a bandah who was familiar with some of the norms governing Muslim society, was loyal to his master and army, it was etiquette - 33 34 35 36 The Ibid., Vol. 1, p. 416; and Mudabbir, Ta’rik ., pp. 21-22. h Mudabbir, Ta’rik h , p. 36. . Ibid. Note the example of Asihtigin G âzi h Ghaznavids, p. 57. and Eryaruq for the Ghaznavids. See Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. Bosworth, 41 of promotion that opened up for him in imperial notable successes. Of all the Sultdn’s slaves, some did learn the etiquette of their master quickly enough so that the monarch felt comfortable in their company. Here the purchased slaves had a definite advantage. Since they were a financial investment (in contrast to war booty) the merchants sought to improve their value in the slave markets by imparting them some training, and the Delhi Sultan certainly spent greater sums on their upbringing to ensure that his investments remained profitable. The starting point to a position of trust and command for the bandagân was an opportunity to personally serve the monarch. During this period the slave would display his personal worth and ability to follow the instructions of his master. The master in turn would establish close personal ties with his subordinate and win his trust and affection through kindness and choice rewards. This was a period of nourishing or fostering a slave, and the Persian chronicles used the verbs parwardan, to foster, and tarbiyat (kardan), to educate or nourish, to indicate the process through which slaves came to be regarded, in time, as foster sons. Fakhr-i Mudabbir noted that after a period of close association, Mu’izz al-Din ’adored [Qutb al-Din] as a beloved son, farzandi’yi ’aziz girâmî, [and] went to extremes in his rearing and education, baparward wa dar tarbiyat-i u mubâlighat namûd. ’37 The establishment of close dyadic bonds between the slave and the master coincided with material rewards where the bandah was gradually raised up the ranks and eventually entrusted with greater responsibility. Ilduz, for example, was exalted, buzurg gardânîd, step by step, pas martabeh, until he was older, churn buzurg shud, when he’received a command and governorship, am§rat wa wilayat rasid. By 602/1206 he was the chief, mihtar, of the Turkish Maliks and the most distinguished of the slaves, buzurgtar-i bandagdn, of SultAn Mu’izz al-Din.11 Similarly, the slave Bahi’ al-Din Tughril was raised in status, buzurg gardinidah bud, after (appropriate) training, batarbiyat.39 There is greater detail about Ai-Beg. After he had been distinguished from the commonlity of the slaves and made to serve in the SultAn’s court, his was a faster rise in the ranks: from a leader of troops, sar-i khayl; to Amir of the stables, Amir-i âkhûr; to in-charge of the foragers on campaign, bar sar-i ’ultifgi; until finally he was given the military assignment, iqtâ’, of Kuhram in the Punjab.&dquo;’ The material advancement of Shamsi bandagin followed similar principles. The slave Malik Taj al-Din Sanjar Kezlik Khan (number 1 in Table 1), for example, had been bought as a child and was brought up and nourished (parwarish) in the royal residence together with the eldest Prince grateful for the avenues service, and there were ... 37 38 39 40 Mudabbir, . , h p. 22 Ta’rîk Vol. 1, p. 411. Ibid., p. 421. Ibid., pp. 416-17. Jûzjânî, , abaqât T Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 42 Nasir al-Din Mal).mûd.41 He was then appointed as the supervisor of the SultAn’s kitchen (châshnîgîr), and was promoted thereafter to commander of the horses (amir-i âkhûr).42 In an identical fashion, the slave Malik ’Izz Tughân Khan Tughril (number 7 in Table 1) graduated from senior cup-bearer (saqi’yi khi4i- ), to senior keeper of the royal writing case (sar-i dawâtdâr), to supervisor of the Sultdn’s kitchen (cl1âshnîgîr), to commander of the horses (amir-i ikhi2r), until he finally received a military and administrative assignment, iqtâ’. 43 The length of time spent in close association with the Sultan and the nature of his subsequent deployment were two fairly reliable indicators of the seniority of the bandah within Iltutmish’s slave corps. Without doubt some of the senior Shamsi bandagân were individuals like Malik Taj al-Din Sanjar Kazlak K_han; Malik Saif al-Din Ai-Beg; Malik Hindu Khan Mu’ayyid al-Din Mubarak and Malik Ikhtiyâr al-Din Qardqush (numbers 1, 4, 9 and 10 in Table 1). All of these had been bought and trained by Iltutmish before he became Sultan. Because of this early bonding and service it is also not surprising that when their master became Sultan they were trusted and deployed by him in some of the most al-Din crucial posts within his realm. When Juzjani chronicled the reign of the Shansabinid Sultan Mu’izz alDin Ghuri, he referred to the small cadre of elite slaves as the Sultan’s bandagdn-i kh§§§ (the senior, literally, the special slaves). Members of this group were deployed to perform a variety of sometimes overlapping sets of tasks: they were appointed as commanders of strategic military assignments, iqtâ’s; or used as military personnel and generals within the central contingent of the Sultanate forces, qalb; or were given important ritual positions in the court or household of the monarch, sar-i jândâr, amir-i Ikhi2r, kl1azînadâr.44 Through their period of training and close association with their master, the Sultan had ascertained the personal qualities of his slaves and the extent to which some of them would make reliable subordinates. 41 ’Ba-u [Malik Nâsir al-Dîn Mahmûd] yek jâ dar hujr-i daulat parwarish yâft.’ Ibid., Vol. 2, p. 4. 42 43 44 Ibid. , p. Ibid 13 Jackson, ’Mâmluk institution’, p. 347, argues for a three-tier hierarchy: those with iqtâ’s, those in the qalb , and the domestic pages in the court. Jackson’s analysis once again followed his model of Mamluk Egypt. Following Ayalon’s description of the k âsakîya (bodyguard, h select retinue, pages), Jackson suggests that these were either members of the qalb or pages, but were slaves who had not yet been given military assignments. David Ayalon, ’Studies on the structure of the Mamluk Army’, Bulletin of the School of Oriental and African Studies, 15, 1953, pp. 213-16. The Delhi Sultanate in the early thirteenth century, however, was still not as structured as the one in Mamluk Egypt, and seniority within the Sultân’s dispensation was not linked with administrative functions. This permitted the seeming anomaly where a Shamsî slave such as Hindû Khân (number 9 in Table 1), who never received an iqtâ’ in Iltutmish’s reign, could nevertheless be regarded as a confidante and trusted slave of the Sultân (for full details, see following pages). Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 43 Thus, bandagdn were given responsibilities, and with it the capacity to take independent decisions, only after the Sultan was assured that the inde- initiatives of the slaves would coincide with the interests of the commodification of the slave, his alien status in local society, the process of bonding through parwarish, and tarbiyat, the rewards of increase in rank and responsibility gradually distinguished the bandagan-i kh4vx from the larger cadre of military slaves. The method of training and gradual deployment of the Mu’izzi and Shamsi bandagin coincided at many points with the normative process described by the Seljuq wazir Nizdm al-Mulk in his Siyisat nâmah. The wazir noted: pendent master. The This is the system which was still in force in the time of the Sgminids (204/819-395/1005). Slaves were given gradual advancement in rank according to their length of service and general merit. Thus after a slave was bought, for one year he was commanded to serve on foot at a rider’s stirrup .... When he had done one year’s service with boots, the tent leader spoke to the chamberlain and he informed the king; then they gave him a small Turkish horse .... In his third year he was given a belt to gird on his waist. In the fourth year they gave him a quiver and bow case .... In his fifth year he got a better saddle and a bridle with In the sixth year he was made a cup-bearer or waterstars on it bearer.... In the seventh year he was a robe-bearer. In the eighth year they gave him a single apex, sixteen-peg tent and put three newly bought slaves in his troop .... Every year they improved his uniform and embellishments and increased his rank and responsibility until he became a troop-leader and so on until he became a chamberlain.45 .... Mu’izzi and Shamsi practice, however, was not nearly as structured as Nizdm al-Mulk might have wanted it to be; Mu’izzi slaves were frequently promoted much faster than the Seljuqid wazir’s recommendations. Qutb al-Din Ai-Beg, for one, jumped ranks in a year that the Seljuq wazir would have stretched out at least over a decade.46 Shams al-Din Iltutmish’s ’system’ differed even more dramatically, since not all of his Turk and other bandagdn were recently enslaved. From Jûzjânî’s biographies it is evident that at least three slaves had served previously in Afghanistan and north India and were sold to Iltutmish by the heirs of their masters. Kabir Khan Ayiz al-Mu’izzi had been a slave of the Mu’izzi notable Nasir al-Din 45 Nizâm al-Mulk, Siyâsat nâmah, Paul, 1978, pp. 103-4. 46 In 587/1191, when he trans. Hubert Darke, London, Routledge and Kegan was the commander of the stables, amîr-i ûr Qutb al-Dîn h âk , received the additional charge of leader of the foragers on the campaign into Khwarazm, and then the revenue assignment, , â’ of Kuhram. t iq Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 44/ Husain Amir-i Shikar,&dquo; Malik Nasir al-Din Ai-Tamar al-Bahâ’i48 had been slave of the Mu’izzi bandah Bahd’ al-Din Tughril, and Malik Nusrat alDin Ta’yisi’ al-Mu’izzi had apparently been a slave of Mu’izze al-Din himself, sold to Iltutmish under unclear circumstances. Through these purchases, Iltutmish went against the traditional practice of buying slaves that were natally and socially uprooted. These three slaves had already established political and social ties with individuals in the area of their deployment, and because of their connections they were not as dependent upon their new master for privileges. Significantly enough, Malik Kabir Khan Aydz al-Mu’izzi provides the only example of a Shamsi bandah-i kJ1â$$ who had his privileges reduced. In 625/1227-28 he was appointed as commander of the town and fort of Multan together with its surrounding territory and towns, qasabat. Multan was a plum appointment ; the city straddled the major trade routes into Sindh, Afghanistan and eastern Iran, and it was also a strategic frontier outpost against Mongol incursions into the Delhi Sultanate. Circa 629/1231,49 however, Malik Kabir Khan was removed from Multan and given (the town of?) Palwal for his personal maintenance, ba-wajh-i m§’yabt§j hal. Kabir Kh£n’s loss of power and privileges would only be reversed in the reign of Rukn al-Din Firuz Shah (633/1236-634/1236). Clearly an ’imperfectly’ recruited slave corps defeated the purposes for a 47 Malik Kabir K ân Ayâz al-Mu’izzî’s name is also included in the list of Iltutmish’s h notables as ’Izz al-Dîn Kabîr K ân without the al-Mu’izzî nisbah. Jûzjânî noted explicitly in h Kabîr K , Vol. 2, p. 5) that he was a slave of Nasîr al-Dîn. Traditionabaqât. ân’s biography (T h ally, the slave, manumitted or otherwise, included in his own name the name of his master as the nisbah of affiliation or belonging. Thus Iltutmish did not hesitate to publicly proclaim himself . bî See from many examples the inscription of Iltutmish dated 608/1211: M.M. t al-Qu Shu’aib, ’Inscriptions from Palwal’, Epigraphica Indo-Moslemica, 1911-12, p. 3. But then why did Malik Kabîr K ân the slave of Nasîr al-Dîn use the nisbah Mu’izzî? Jûzjânî may well h have confused masters. To obscure matters further Jûzjânî does mention other Maliks with the nisbah of Nâsirî in his list of Shamsi notables: Malik al-Umarâ’ Sunqar Nâsirî and Malik Bidar Kûlân—Muhammad Kulan Turk Nâsirî in a variant manuscript (T , Vol. 1, p. abaqât. 450, fn. 6). We know nothing of their social backgrounds. If they had been slaves their nisbah of affiliation, Nâsirî, indicated that they had been closely associated to a patron with that name. This could have been Malik Nâsir al-Dîn Husain Amîr-i Shikâr, or even, possibly, Nâsir al-Dîn Qubâcha. 48 The editor of Jûzjânî, , abaqât. Habibi, read the name as Ai-Tam? (AITM) Bahâ’î, and T the Malik’s name was also included in the list of Iltutmish’s notables (Vol. 1, p. 451) as Malik Nasîr al-Dîn AITM Bahâ’î. In the list of Iltutmish’s nobles Habibi also mentioned the alternate reading Ai-Tamar in a variant manuscript. I have preferred the Turkish reading AiTamar. See ibid., pp. 450-51 and fn. 6. In the list of Iltutmish’s notables (ibid., p. 451) another dependent of Bahâ’ al-Dîn Tughril is also mentioned: Malik ’Izz al-Dîn Tughril Bahâ’î. It is possible that Jûzjânî did not k ow Malik ’Izz al-Dîn personally, or that the M n lik a was not important enough in Iltutmish’s reign, or in that of his successors, to merit a separate biographical reference. Nevertheless, he must have carried enough prestige to be recognized in Iltutmish’s list of notables. 49 Ibid., Vol. 2, p. 6, noted that his removal from Multan took place after two or four (lunar) years of his appointment. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 45 which it had been raised, and Malik Kabir K...hân’s insubordination was therefore quickly dealt with. However, rather than casting doubt upon the effectiveness of the bandagân-i 5hamsi, the case of Malik Kabir Kh£n enlarges upon the pressures upon Shams al-Din to not pay any particular attention to ’tradition’ in the purchase of his bandagân. The urgency with which Iltutmish sought to recruit and deploy military personnel, especially just before and after his accession, might also explain the surprising presence of Malik Hindu Khan Mu’ayyid al-Din Mubarak al-Khâzin amongst the Shamsi bandagdn-i khlixx. Malik Hindu Khân, to my knowledge, was the first indigene to be advanced to the status of bandah-i khass since Tilak, the slave of Sultan Mas’ûd Ghaznawi (421/1031-432/1041),50 and would be followed subsequently.by the slave K_husrau Khan Barwari elevated by Mubarak Shah Khalaji (716/1316-720/1320).5 The circumstances influencing Hindu Khdn’s recruitment and promotion differed from the other examples of ’Hindu’ slaves, as did his conduct which, unlike that of K_husrau Khan Barwari, remained exemplarily loyal to his master. It was the unique qualities of the bandagân-i kJ1â$$ that were sought by Iltutmish to relieve the pressures of deploying reliable subordinates over newly conquered territories at some distance from Delhi. Uchch, for example, was conquered in 625/1228 and, when the royal forces withdrew, its charge was handed over to the Shamsi slave Malik Taj al-Din Sanjar Kazlik Khan (number 1 in Table 1). Malik Taj al-Din’s governorship came 50 On Tilak see C.E. Bosworth, The Ghaznavids, pp. 101, 127-28. Irfan Habib notes that Hindû K ân may not necessarily be an indigene, but possibly a Turk since ’the name Hindû is h also found amongst Turks, Hindû meaning black and so an alternative to the Turkish word , popular in Turkish names’: ’Formation of the Sultanate ruling class’, p. 10, fn. 60. qarâ However, Hindû and qarâ were not synonyms for ’black/strong’. In Persian poetics and prosody, however, Hindû might be metaphorically used for black as Annemarie Schimmel pointed out several years ago. See Annemarie Schimmel, ’Turk and Hindu: A poetical image and its application to historical fact’, in Speros Vryonis Jr., ed., Islam and Cultural Change in the Middle Ages, Wiesbaden, Otto Harrassowitz, Fourth Giorgio Levi Della Vida Biennial Conference, 1975, pp. 107-26. Nor am I sure how Habib read ’Mahûr’ in Jûzjânî, , abaqât. T Vol. 2, pp. 11, 44, 62; the printed text clearly spells it as MHR without any vowels. That MHR should be read as Mathûr or Mathûra does not automatically follow. In fact since Jûzjânî describes MHR as , aba a small town, it is possible that the town of Mathûra is not s qa the one that the author was referring to. 51 I have deliberately excluded from the category of Shamsî bandagân-i ss â the example h k of the eunuch, khwaja sara, ’Imâd al-Dîn Raihân who was also of Hindu origin. Jûzjânî recognized ’Imâd al-Dîn Raihân’s prominence in the politics of the Delhi Sultanate only in 650/1252 during the reign of Sultân Nasîr al-Dîn Mahmûd. There is no information, however, concerning his enslavement and early service. His importance in the early years of Nâsir alDin’s reign (644/1246-664/1266) suggests that he was purchased, like his rival Balban, sometime in the last years of Iltutmish’s reign (died, 633/1236). As a result he would have been too â His rise to prominence, unlike Tilak or K h k young to have been a Shamsî bandah-i ss. usrau h h &K acirc;n, was not a result of the favour and trust of the monarch. ’Imâd al-Dîn rose to power, much like Balban, through his own successful ability to utilize the opportunities for political manipulation present in the unstable years of Iltutmish’s successors. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 46 end with his death in 629/1231-32, upon which another slave, Malik Saif al-Din Ai-Beg-i Uchch (number 4 in Table 1), was appointed as governor of the region. Malik Saif al-Din remained as governor of Uchch through the remaining portion of his master’s reign. The case of Multan was identical. After its conquest, its charge was handed over to the slave Malik Kabir Khan Aydz al-Mu’izzi (number 2 in Table 1). After he was removed in disgrace, circa 629/1231, the charge of Multan was handed over to the slave Malik Ikhtiyar al-Din Qardqush Khan Ai-Tegin (number 10 in Table 1) who continued as governor through the remainder of Iltutmish’s reign. The strategic locale of both Uchch and Multan seemed to necessitate the appointment of trusted subordinates from the very outset. Three of the four amirs (numbers 1, 4 and 10) appointed to Multan and Uchch were old slaves of Iltutmish dating back to a period before he had succeeded in becoming Sultan. Not accidentally they were regarded as his seniormost slaves and therefore appointed to positions of trust from which-with ,the exception of Kabir Kh£n-they were never removed during the lifetime of their master. The close nexus between strategic commands and the deployment of senior slaves is also well brought out by the case of Lakhnauti. Here, unlike the case of Uchch and Multan, it was the free Amir ‘Ala’ al-Din Jani who was appointed governor of Lakhnauti after the death of Prince Nasir al-Din Mahmud in 628/1230-31.52 Subsequently, sometime in 629-30/ 1231-33, Malik ’AIA’ al-Din Jani was dismissed, ma’zul shud,53 and in his place the slave Malik Saif al-Din Ai-Beg (number 5 in Table 1) was appointed. Unlike Malik Jani, Malik Saif al-Din’s governorship was a huge success, and the Sultan appreciated his remission of elephants to Delhi sufficiently to reward him with the title Yughântat. Saif al-Din established the tradition of a bandah as a governor of Lakhnauti for the remainder of Iltutmish’s reign. After Saif al-Din’s death in 631/1233-34, the governorship of the province passed into the hands of the slave Malik ’Izz al-Din Tughdn Khan Tughril (number 7 in Table 1). In contrast to Habib’s to an 52 Jûzjânî, , abaqât. Vol. 1, pp. 438, 448, 453-54. There has been considerable confusion T concerning Nâsir al-Dîn Mahmûd’s death. Jûzjânî mentioned (Vol. 1, p. 447) that Nâsir alDin died in 626/1229. Later in his chronicle (Vol. 1, p. 454) he mentioned that the prince died in Lakhnauti a (lunar) year and a half after (b’ad az yek sâl o nim) the arrival of the Caliphal envoys in Delhi, 626/1229, which would imply his death sometime in 627-28/1230-31. The likelihood that 627-28/1230-31 is the correct date is incidentally corroborated by Jûzjânî who notes that Balka K aljî rebelled in Lakhnauti immediately after the prince’s death (Vol. 1, p. h 448), and that Iltutmish campaigned against the rebel in 628/1230-31. At the termination of the expedition, Iltutmish appointed ’Alâ’ al-Dîn Jânî as governor of Lakhnauti. Finally, the inscription on the entrance to the prince’s mausoleum in Delhi, notes the date of construction as 629/1231-32, which would make it about a year and a half after Nâsir al-Dîn’s death. See J. Horowitz, ’The inscriptions of Muhammad ibn Sam, Qutubuddin Aibeg and Iltutmish’, Epigraphia Indo-Moslemica, 1911-12, p. 24; and S.A.A. Naqvi, ’Sultân Ghari, Delhi’, Ancient 53 India, 3, 1947, p. 5. Vol. 2, pp. 9-10. Jûzjânî, , abaqât T Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. /47 argument that the Shamsi bandagdn were frequently ’transferred’ from iqt§’ to iqt§’, the facts would suggest that Iltutmish deliberately chose not to move some of his senior slaves from their assignments. The Shamsi bandagân were the most favoured group in Iltutmish’s dispensation precisely because, contrary to established historical opinion, they were not a Turkish ’bureaucracy’ .54 Rather than professional service regulations between an officer and the state, it was dyadic bonds created through careful fostering, parwarish, and education, tarbiyat, that influenced relationships between the master and the slave. ’Promotions’, when they were given, were rewards for ’approved/agreeable service’, khidmat pasandida kardah bûd/5 and while there might be a range of ’approved actions’, these were ultimately judged against the cultural expectations of a slave’s selfless service in the cause of his master. For that matter, despite NizAm al-Mulk’s advice, there were no time-bound ’promotions’ within this cadre. Some slaves seemingly ’languished’ with the same responsibilities for years; even when Malik Hindu Khan (number 9 in Table 1) was appointed khazinadir, treasurer, he continued as tashtdir, ewer-bearer, for the remaining years of Iltutmish’s reign.56 Jdzjdni presented this information as an evidence of the affection borne for the master by the slave and the honour given to the bandah who was allowed such close proximity and intimacy, qurbat tamâm ddsht, with the Sultin over an extended period of time. Malik Hindu Khân’s career ’languished’ only if we understand it from the anachronistic perspective of a bureaucratic system. If Iltutmish’s bandagin-i k_hass were given some of the more honoured and strategic responsibilities within his patrimony, the junior slaves were not ignored. Their deployment seems to have been the occasion when they ’won their spurs’ in positions of more limited responsibility such as shahna or superintendent. In contrast to the muqta’, the military commander of a region who possessed considerable independent initiative, the shahna, a superintendent, had a more precise responsibility. He was normally detailed for specific tasks under the direction of the Sultan, an officer through whom the Sultan maintained direct control over certain establishments or crucial functions of the government.5’ Before his appointment to the command of Multan and the receipt of the title QarAqush, the previously 54 See the writings of Muhammad Habib (op. cit .), h K a liq Nizami (Religion and Politics; ’The early Turkish Sultans’), S.B.P. Nigam, op. cit., and Irfan Habib (’Formation of the Sultanate ruling class’; ). Iqtâ’s’ ’ 55 See for example, Jûzjânî, , abaqât Vol. 2, pp. 9, 10. T . ., p. 19: ’ta âkhir-i ’umr dast az tashtdârî na-dâsht’ Ibid 56 57 Although there is insufficient evidence to draw a parallel, there are certain features that are common between the areas where sha nas were used in the Seljuqid state and the early h Delhi Sultanate. For the Seljuqids see Ann K.S. Lambton, ’Internal structure of the Seljuq Sultanate’ in J.A. Boyle, ed., The Cambridge History of Iran, Vol. 5, Cambridge, Cambridge University Press, 1968, pp. 244-45. The office itself is of a Chinese origin ) shao-chien and ( was introduced into the central Islamic lands by the Qarâ Khitâ’î. See Doerfer, Turkische und Mongolische Elemente, Vol. 3, pp. 320-21, s.v. ’sihna’. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 48 discussed Malik Ikhtiydr al-Din (number 10 in Table 1) had been the shahna of the khâlisa in the vicinity of Tabarhind, the revenues of which were reserved for the Sultan. While the presence of unalienated khdlisa lands itself indicated an attempt by the Sultan to increase his revenue from taxation on the produce of land, in contrast to an income largely from plunder and tribute, it is significant that the administration of this land was under the command of a shahna who would eventually become one of the Sultdn’s bandagdn-i khtiss. Junior slaves were also appointed as shahnas to the armoury, zarr§dkh9nah, of Budaon (number 13 in Table 1), of the river and the boats, shahnah-i bahar wa kishtihi (number 15 in Table 1), of the imperial stable, shahnah-i ikhur (number 14 in Table 1). In other words, there does seem to be a correlation between the increase in the bandah’s responsibilities and the bonding between the master and the slave; as the Sultdn’s confidence in his junior slaves increased, they were given greater responsibility and political initiative. The distinctions between the manner in which some slaves were treated over others, can be related to the different types of individualized relationships that they established with their master. From the larger cadre of the Shamsi bandagin, only some were regarded as the Sultdn’s special, honoured, khass, slaves. ’ IV The sense of a larger corps of military slaves within which there was an elite cadre privileged by the trust reposed by the Sultan was also recognized by the mid-fourteenth century historian Diya’ al-Din Barani. He noted that Sultan Ghiyas al-Din Balban (664/1266-686/1287) was one of the Shamsi slaves, bandah-i az bandagân-i Shamsi bud, who was manumitted together with the ’forty’ Turkish slaves, wa dar miyiln-i bandagân-i tûrk-i cl1ihilgânî azâd shuda. 58 Manumission, of course, did not alter the juridical tie of dependence, wala’, or the emotional bond of association that tied the slave to his master. 59 Instead it was a reward given to only some (bandagan-i tûrk-i chihilgânî ), from the larger body of slaves (bandagin-i Shamsf in return for faithful service. The bandah-i k_hass, the future Sultan Shams alDin Iltutmish, had been thus rewarded by Qutb al-Din Ai-Beg, and according to Barani, Ghiyas al-Din Balban, together with the other chihilgini, belonged to a similar privileged category. Gavin Hambly was the first scholar to understand the cl1ihilgânî as a 58 Barani, Ta’rîkh-i Firûz Shâhî ., ed., Khan, p. 25; ed., Rashid, p. 30. a valuable discussion of the ties between the freed slave and his patron see, Paul G. Forand, ’The relation of the slave and client to the master or patron in medieval Islam’, International Journal of Middle East Studies, 2, 1971, pp. 59-66; for the technical position of the freed slave in Islamic law see Joseph Schacht, Origins of Muhammadan Jurisprudence, Oxford, Clarendon Press, 1975, p. 161; and The Encyclopaedia of Islam, second edition, s.v. 59 ’ For ’Abd’ and ’G ulâm’. h Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 49 privileged cadre of slaves within the larger retinue of bandagân-i Shamsi. 60 Hambly argued that the bandagân-i Shamsi included in the twenty-second section of Juzjani’s Tabaqât-i Nd$iri were twenty-five of the forty members of the chihilgini. He based this conclusion on the argument that since Juzjani chose to narrate their biographies they must have been important slaves of Iltutmish. Moreover, most of them shared the essential qualities attributed to the çhihilgâni by Barani: ’Turks, Shamsi slaves, holders of high offices and titles’. 61 He went on to add that since two of the chihilgani identified by Barani, Ghiyds al-Din Balban and Shir Khan Sunqur, also had their biographies in Juzjani, the bandagdn in the Tabaqât-i N4viri must be a part of the ’forty’ elite slaves of Iltutmish. In identifying Balban as a member of the çhihilgânî, one part of Barani’s explanation implied that those Shamsi slaves (like Balban) who held high offices and titles in the reign of Sultan Nasir al-Din Mahmud (644/1246664/1266) held lesser but still important positions in Iltutmish’s reign. Hambly’s argument relied on this element of Barani’s explanation. Juzjani’s criteria in narrating the biographies of select Shamsi bandagdn, however, differed completely from Hambly’s claim. Juzjani had been very explicit that he chose to give biographies of only those slaves to whom he was personally beholden, and that these accounts should be seen as grateful offerings for dues, add’i I;zuqûq, owed by the author to his important benefactors.’ As a result there is little reason to expect that Juzjani included in his chronicle only those slaves who had been important during Iltutmish’s reign. If we refer to Table 1 it becomes immediately apparent that of the twenty-five Shamsi slaves, eight (numbers 11, 16, 17, 19, 22, 23, 24, and 25) held no positions of privilege during Shams al-Din’s rule. Ghiyas alDin Balban, around whom hung Barani’s tale of the ’forty’ slaves, was purchased only in 630/1232-33, which was towards the very close of Iltutmish’s reign (died, 633/1236). It is not likely that he was manumitted during the remaining three (lunar) years of Shams al-Din’s life, for he was certainly not a senior slave. He was only a kJ1asadâr (falconer) by the end of Iltutmish’s rule, and received his first advancement only in Sultan Radiyya’s service (634/1236-637/1240). Until that time he had still not been trusted with a military, administrative command. Nor could someone like Malik Ikit’ Khan Saif al-Din Ai-Beg Khita’i (number 16 in Table 1) have been a Shamsi bandagdn-i kha~~. He was a lowly bodyguard, jândâr, in Iltutmish’s reign, and had only progressed to the post of chief bodyguard, 60 61 Hambly, op. cit., pp. 57-62. Ibid., p. 62. Vol. 2, pp. 2-3. Habibi read the title of the slave as Bat (BT?) K ân Saif al-Dîn Ai-Beg K h itaî, and h noticed the presence of different readings in various manuscripts. In one of the variant manuscripts, this was written as AIKN, Ikin, and Jackson suggested the Turkish Ikit as a abagât. Vol. 2, p. 28, T possible reading. I have followed Peter Jackson here. See Jûzjânî, , and Vol. 1, p. 476 and fn. 11; and Jackson, The Mongols and India, p. 92, fn. 137. 62 63 abaqât. T Jûzjânî , Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 50 sar-i jdnddr, in ’AIA’ al-Din Mas’ûd’s dispensation (639/1242-644/1246). For that matter Juzjani has no record of Shir Khdn’s (number 23 in Table 1) deployment during Iltutmish’s reign which was an astonishing omission if Shir Khan was indeed an important slave commander in 633/1236. If the held important commands because of Shamsi bandagin-i k_hasslc_hihilg&ni Sultan Iltutmish’s trust and patronage, then, at least all of the twenty-five slaves mentioned in Juzjani’s biographical section, including Balban, were not the same as Barani’s ’forty’. Although Muhammad Habib never seriously studied the question of the cJ1ihilgânî, in a brief aside in his ’Introduction to Elliot and Dowson’ he explained that forty was merely ’a formal number’, and the number of cJ1ihilgânî was probably far lower. 64 What Muhammad Habib was alluding to was the fact that within Biblical and Qur’anic contexts, ’forty’ was a synonym for ’many’ and should seldom be taken literally. There are innumerable examples of the usage of ’forty’ in this context, and in a form much like the ’cJ1ihalgânî’, the Book of Dede Korkut referred to Dirse K_han’s ’forty warriors’.&dquo; In both cases ’forty’ should be understood as a metaphor for ’a large (unspecified) number’ and Barani’s usage of the term cJ1ihalgânî is clearer when interpreted in this sense.66 Barani was accurate in defining the cJ1ihalgânî as an elite cadre within the bandagin-i Shamsi, 64 Muhammad Habib, op. cit., p. 107; Muhammad Habib’s arguments were misrepresented ’The early Turkish nucleus in India’, Turcica, Revue d’etudes Turques, 9, 1977, p. 102, fn. 19. Aziz Ahmad incorrectly cited Habib’s estimate of the number of ihilgânî as more than one thousand. Recently Irfan Habib, ’Formation of the Sultanate h c ruling class’, p. 16, endorsed his father’s interpretation. 65 The Book of Dede Korkut, trans. Geoffrey Lewis, New York, Penguin Books, 1982 reprint, p. 28 and fn. 3. 66 There have been several interesting efforts to account for the ’forty’ notables of Iltutmish. The most imaginative exercise is by Khurram Qadir, ’Amiran-i Chihalgan of northern India’, Journal of Central Asia, 4, 1981, p. 97, who calculated that ’Iltutmish had four qadis, two wazirs, twenty-one amirs and ten offspring [sic. Iltutmish, in fact, had nine children]. If one were to include the three Mu’izzi contenders for the throne of Aram Shah, the list of individuals thus involved in the power struggle during Iltutmish’s reign comes to forty.’ Jackson’s comments, ’The Mongols and India’, p. 208, fn. 82; and ’Mamlûk institution’, pp. 345-46, however, are more interesting. He suggested that rather than ’forty slaves’ the ’use of that each [slave] was [a] commander of a the distributive numeral ] gânî suggests [ company of 40 mamlûks.’ Jackson suggested further that the Amir h aacirc;na &T blk in Mamlûk Egypt who was a commander of forty horses might provide a parallel to the c ihilgânî in h India. But, if there had been a tradition of numerical commands within the early Delhi Sultanate army like the Egyptian Mamlûk Sultanate, it is inexplicable that Jûzjânî, who paid considerable attention to the careers of the bandagân-i Shamsî , made no reference to it whatsover. For that matter, it seems too far fetched to assume that Fak r-i Mudabbir, Hasan h Nizami, and Amir K usrau would all ignore such a system of hierarchical commands until h Barani, writing more than a hundred years after Iltutmish’s reign, mentioned it in passing. On the other hand, a near contemporary of Barani, ’Isâmî h în ed. A.S. Usha, t al-Salâ Futû , ( Madras, University of Madras, 1948, p. 122) also mentioned ’forty turkish slaves’ during the reign of Iltutmish. ’Isâmî did not use the distributive numeral with forty, he mentioned the slaves as ’c , and the context made it clear that the author implied not forty ihil bandah-i tûrk’ h but an indeterminate number. by Aziz Ahmad, ... Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 51 effectively a group of Shamsi bandagân-i kJ1â$$. These would be a group of slaves, trusted and commended for their actions, raised above others by Shams al-Din Iltutmish, and manumitted as a reward for their service. Barani, however, was inaccurate in suggesting that Balban was a member subsequent scholars mistaken in interpreting his usage of ’forty’ too literally. Diya’ al-Din Barani used cl1ihilgânî in two ways in his Ta’rikh. The first was an explanation that within the larger body of Shamsi bandagân there was an elite cadre, senior in status to the other military slaves. Juzjani’s evidence concerning the organization of the Shamsi bandagân has shown that Barani was entirely correct on this point. Barani seems to have been unsure of the exact number of these elite slaves in the reign of Iltutmish, and he therefore used ’forty’ to signify an indeterminate but large number. Second, Barani was also seeking to explain the process through which some of the bandagdn-i Shamsi managed to prolong their stay in power by removing other competitors during the interregnum between Sultan Iltutmish and Balban (633/1236-644/1246).6’ Balban had excelled in the politics of manipulation characteristic of this period, and had come to effectively control power within the Sultanate from 647/1249. One way of explaining Balban’s influential position in Nasir al-Din’s reign was by suggesting that the Shamsi slave had always been an elite slave, who came to power as his peers, the senior (forty) Shamsi bandagân died. Thus, while Barani was entirely accurate in understanding the hierarchical element within the Shamsi slave system and in identifying the cl1ihilgânî as an elite cadre within the bandagân-i Shamsi, effectively a group of Shamsi bandagin-i khass, he was inaccurate in suggesting that Balban was a member of this of this group, and group. V The coincidence in were empowered our sources to act as the concerning the logic by which elite slaves agents of the Sultan is noteworthy. My emphasis upon the bonds which tied the slaves with their masters was partly in an effort to explain how the unfree could be given command over the free. The relationship of subordination, nourishment and loyalty which bonded masters and slaves together, allowed bandagin to play crucial, strategic and unsupervised roles within the Sultanate. In an interesting development, however, when the Delhi Sultans started deploying fewer military slaves from the fourteenth century onwards, their reliance upon larger numbers of free notables did not mean a concomitant redefinition of bonds which tied master to subordinate. Instead, in the discourse of the court, the paradigm of subordination rather ironically 67 See Baranî, Ta’rîkh i Firûz Shâhî ed. S.A. Khan, p. 27, ed. Rashid, p. 32: ’and after the death of Sultân Shams al-Dîn Iltutmish his ’forty’ Turkish slaves were successful, ’kamiyâb gashtand’ in their conflict with the free Amirs. Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution. 52 continued to be the ’noble’, selfless conduct and loyalty of the slave to his master. Thus, generations after the Shamsi bandagin-i khp§ were deployed as instruments of coercive political power, in the Tughluqid period the free-notable Mujir Abu Rija declared that: ’I am a faithful slave, bandah-i n8k-khw§h of the king ... I am like a slave, ready to serve, [for you] I will sacrifice all of my family and possessions, khân u mân.’68 This was not a minority opinion either; the great mansabdirs during the reign of Akbar and Jahangir protested their loyalty, and proudly proclaimed themselves unfree as bandagin-i dargih.1 In a political culture where a large number of the elite were slaves, or freemen who sought to appear as slaves, the term ’noble’ is either a misnomer or in need of a dramatic redefinition. Certainly, in the thirteenth century, while the Shamsi bandagin-i khâ$$ may have been useful in controlling newly conquered, dispersed territories, the governed did not necessarily appreciate the conduct of their governors. That the local commanders may have lacked the moral right to be regarded as the leaders, least of all the ’noble’ element in local society, is suggested by this plea which was addressed to Hamid, a freeman and a scribe who served a bandah: You are a Oh Hamid! why are you standing before this man? learned man (’ilmi) and he is ignorant (jahil), you are a freeman (hurri) and he is a slave (bandah), you are a pious man (,5alihf) and he is an uncultured sinner (jüsiq). 70 ... It remains to be noted that while Hamid’s conscience may have forced him to give up service under the Turk, similar compulsions were not evident in the case of Fak_hr-i Mudabbir or Juzjani, both of whom continued to serve their slave masters. Nevertheless, in the plea addressed to Hamid which was mentioned in the malfuzat of Nizdm al-Din Auliya, a source which is not a court chronicle, we have evidence of discordance, the ambiguity with which subjects sometimes judged the moral qualities of their leaders. By the parameters of the qualities possessed by a noble and noble conduct, to the observers of the day the Shamsi bandagan remained slaves. This should hardly be surprising; it was because of their distinct unfree status in the first place, that the bandagin had been deployed by Iltutmish. ° 68 ’Isâmî, Fulûh , în p. 398. t al-Salâ 69 See John F. Richards, ’The Formulation of Imperial Authority under Akbar and Jahangir’, in J.F. Richards, ed., Kingship and Authority in South Asia, Madison, University of Wisconsin, South Asia Center, 1978, pp. 252-85; and ’Norms of Comportment among Imperial Mughal Officers’, in Barbara Metcalf, ed., Moral Conduct and Authority, Berkeley, University of California Press, 1984, pp. 255-89. 70 Amir Hasan Sijzî, Fawâ’id al-Fu’âd , ed. Khwaja Hasan Thani Nizami Dihlawi, Delhi, Urdu Academy, 1990, pp. 342-43. The Delhi edition of the Fawâ’id al-Fu’âd is an exact duplicate (including pagination and critical commentary) of M.L. Malik’s edition published from Lahore (Lahore, Malik Siraj al-Din and Sons, 1966). Downloaded from http://sih.sagepub.com by SUNIL KUMAR on May 14, 2008 © 1994 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.