Marathas, Marauders, and State Formation in Eighteenth-Century India 9780195633863, 0195633865

Citation preview

Marathas, Marauders, and State Formation in Eighteenth-Century India

STEWART GORDON

Delhi

Oxford University Press Bombay

1994

Calcutta

Madras

Oxford University Press, Walton Street, Oxford 0X2 6DP Oxford New York Toronto Delhi Bombay Calcutta Madras Karachi Kuala Lumpur Singapore Hong Kong Tokyo Nairobi Dar es Salaam Cape Town Melbourne Auckland Madrid and associates in Berlin Ibadan

© Oxford University Press 1994

Copyright in the individual essays vests in the author.

ISBN 0 19 563386 5

Typeset by Guru Typogntph Technology, New Delhi 110045 Printed at Rekha Printer! Pvt. Ltd., New Delhi 110020, and published by Neil O’Brien, Oxford University Pie** YMCA Library Building, Jai Singh Road, New Delhi 110001

/

A uOlj/sM/- |

Contents Acknowledgements

vii

Introduction

ix

1. Scarf and Sword: Thugs, Marauders, and State-formation in Eighteenth-Century Malwa

!

1

2. The Slow Conquest: Administrative Integration of Malwa into the Maratha Empire, 1720-60,'

23

3. Legitimacy and Loyalty in Some Successor States of the Eighteenth Century

64

4. Forts and Social Control in the Maratha State

82

5. Recovery from Adversity in Eighteenth-Century India: Re-thinking ‘Villages’, ‘Peasants’, and Politics in Pre-modem Kingdoms

99

6. Kinship and Pargana in Eighteenth-Century Khandesh J. F. Richards, co-author

122

7. Bhils and the Idea of a Criminal Tribe in Nineteenth-Century India

151

8. Burhanpur: Entrepot and Hinterland,1650-1750

163

9. Zones of Military Entrepreneurship in India,1500-1700

182

References

209

Glossary

212

Index

217

Acknowledgements I wish to thank the following scholars who discussed and critiqued early drafts of various articles: Donald Attwood, Frank Perlin, Richard Barnett, Richard Tucker, John Richards, Eleanor Zelliott, John Broomfield, Nicholas Dirks, and A. R. Kulkami. In India, I wish to thank the sponsors of various seminars at which I was allowed to present papers, and the director and staff of the Pune Daftar and the National Archives. In England, the staff of the India Office helped find arcane, pre-colonial material, Finally, I gratefully acknowledge the American Institute of Indian Studies, the Fulbright Foundation, and the Social Sciences Research Council for support for my studies over the past twenty-five years.

Introduction In the twenty-three years between the first and the last of these essays, almost all the things we historians ‘knew’ about eighteenth-century India have turned out to be false or inadequate or self-serving Orientalist notions. Perhaps some recounting of issues will help position these essays in the turmoil of the times. Two decades ago, the overriding historical periodization doomed the eighteenth century; it traced two centralized, bureaucratized empires— the Mughal and the British—and a period of chaos and anarchy (the eighteenth century) in between. After the Mughal period, historians found decline in just about everything, from trade to political institutions. Art historians found decadence. By now, it is impossible to avoid noticing how self-serving was the notion that Britain rescued India from itself, putting an end to the ‘chaos’ and ‘anarchy’, and stamping out ‘barbarous’ practices like thugi and sati. The first essay in this collection presented my doubts about this periodization as I re-examined the evidence on thugi. No one seemed to typify the prevailing view of the eighteenth century more than the Marathas—known mostly for turbulent warfare and long­ distance raids. What did not fit with this image, however, was the huge archive in Pune, filled with millions of documents of an enormously sophisticated revenue administration. As I and other scholars began to examine these materials, we understood why the British colonial admin­ istration had closed this archive to all historical researchers. Here was no evidence of ‘chaos’ or ‘decay’, but a vibrant administration recording the minute details of rural and urban life and committed to agricultural development. A similar process of discovery has taken place at every

archive of regional successor states of the eighteenth century, for example, Lucknow, Hyderabad, Bikaner, and Jaipur. The struggle to understand Maratha institutions and processes, so tantalizingly laid out in the documents of the Pune Daftar, represents the themes of the next three or four articles. With each one, some long-held assumption of the eighteenth century seemed to fall by the wayside. Instead of having no administrative system, the Marathas’ system was so sophisticated that Pune bankers could base rates of interest loaned to the

x / Introduction government on calculated risk and reliability of collection, area by area. Rather than no rules for succession, there were, in fact, complicated and widely-shared modesof legitimacy and loyalty. Overall, as other scholars found in other areas, there were considerable continuities with the Mughal Empire. Another of our most cherished notions from the colonial period was the primacy of the village. Empires supposedly washed over them, providing no essential services, and leaving them unchanged. This idea was so pervasive and unchallenged that it affected the unconscious construction of post-Independence research topics in a variety of fields. Anthropo­ logists did ‘village studies’, which treated the village as a self-sufficient unit suspended in space and time. Economists studied the pre-modem village, assuming a self-sufficient unit without money or markets. Political scientists mourned the invasion of ‘modernity’, factionalization, and conflict which disturbed the villages’ ancient sleep. The documents of the eighteenth century have forced several scholars, including me, to re-think relations between pre-colonial governments and villages, and the presence and importance of governmentin the countryside. One of the core issues of this re-thinking is monetization. It has been very hard to sustain the notion of an isolated, self-sufficient village in the face of the overwhelming evidence of money use and monetized contracts in a whole range of villages across Maharashtra. These concerns led me to examine some aspects of trade which brought money into the countryside in Central India. Others have undertaken similar studies for central Maha­ rashtra, Tamilnadu, the Punjab, and the Ganges valley. Similarly, a quarter century ago, scholars were quite sure they knew what caste was—at minimum, a group with endogamous boundaries, self-governing institutions, occupational specialization, and distinctive religious and cultural customs. Caste, like the village, floated in a time­ less, changeless ‘structure’ unaffected by politics or the larger world. None of this is supportable now that scholars have confronted the docu­ ments of the eighteenth century. The very words we use to describe castes—Maratha, Rajput—have histories of change and conflict, ragged edges and ambiguities which reflect the politics of the day. We know now that the closure of castes was one of the principal ‘divide and rule’ tactics of the British in the nineteenth century. It was only in the course of the recovery and examination of the documents of the eighteenth century that scholars realized that caste never has been separated from politics, never removed from the exercise of power. This is not a modem or twentiethcentury creation in India.

Introduction / xi These considerations led me to re-think—in the most recent essay— the whole nature of the political process in the pre-colonial period. Like ‘the village’, and ‘caste’, kings and kingship seemed to float in a timeless textual present in scholarly writing on India. The more I examined the documents, the more I realized that these vast Indological generalizations were simply not describing anything I was seeing. Rather than general Brahmanic processes, here were processes specific to the ecology of Maharashtra and service in the Deccan sultanates which merged peas­ antry, kingship, and soldiering. Other scholars have in recent years made similar discoveries in the documents of North India, Rajasthan, and Tamilnadu. Overall, India’s history seems much richer for the end of notions of unchanging village, immutable caste, and timeless textual kingship. Specific ecologies, and political processes, and military cultures give India’s pre-colonial past a complex mosaic out of which scholarship is teasing common and contrasting patterns from the documentation of regional archives. I feel privileged to continue this process with scholars both inside and outside of India.

Chapter One

Scarf and Sword: Thugs, Marauders, and State-formation in Eighteenth-Century Malwa* I ntrod uction

Thugi1 captured the English popular imagination in the nineteenth century as did few features of India. Bracketed with suti (widow selfimmolation) and infanticide, it proved the ‘backward’ state of India to the Utilitarians, and its need for Christianity to the Evangelicals. Priests thundered of the need for missionaries and Benthamites argued for better courts and jails. For many Englishmen, Thugi was only the most hideous and bizarre of those strange religious perversions which were the essence of India—making it alien, evil, incomprehensible, and frightening enough to be titillating. European official accounts of Thugi and later histories reveal almost as much colour as the novelists; yet ^mong them there are also striking inconsistencies. Let us compare descriptions of Thugi by authors whom we might justly expect to be authoritative. First, Percival Spear, a twentieth-century historian: The dislocation of society drove adventurous, hopeless or embittered spirits to a lawless life. They formed the material for princely armies or robber bands, each of whom recruited from the other as fortunes rose and fell. The landless or uprooted man looked for a leader and reckless from despair was the typical figure of the time. A specialized form of these men were Thugs, robbers and ritual ♦The ideas of this paper were first presented at the Michigan Rotating South Asia Seminar in March 1969. With the comments and criticism of that group and the generous assistance of J. H. Broomfield it has been substantially revised to its present form. Reprinted from The Indian Economic and Social History Review, vol. VI, no. 4, December 1969. 1 The problem of transliteration is a plague whenever foreign terms are used. Words known in the West in one form—such as Thug, Thuggi, Bhil, Rajput-I have retained in that form, though the toms appearing in quotations will vary, i.e.: Thag, Thagi, Thaggi. Place names, especially in the section on the geography, of Malwa I have taken from James Tod Annals and Antiquities of Rajasthan and John Malcolm A Memoir of Central India.

2 / Marathas, Marauders, and State Formation murderers, who rose to prominence in these times and spread across central India to the terror of travellers and peaceful men.2

Second, James Sleeman, a military officer in India, and author of Thug, or a Million Murders (1993): [The Thug] regarded the stalking of men as a higher form of sport. . that fiend in human form, luring his victims to their doom with soft speech and cunning artifice, committing the coldblooded murder of every man he met, saint or sinner, rich or poor, blind or lame____The taking of human life for the sheer lust of killing was the Thugs’ main object: the plunder, however pleasant, being a secondary consideration.3

Third, the characterization of Thugi in R. V. Russell’s comprehensive Tribes and Castes o f the Central Provinces (1916): Thug is listed between Teli (oilmen) and Turi (a cultivating caste) and the Thugs were consi­ dered a caste or ‘community of murderers’ just as the Teli were a com­ munity of oilmen.4 These accounts are quite incompatible. First, on the fundamental moti­ vation of Thugs. To Spear, they were essentially rational men, though ‘hopeless’ and ‘embittered’, who were driven by conditions in society and for lack of other income to murder and rob. For Sleeman, in contrast, the basic motivation was psychological blood-lust. A Thug was a ‘fiend in human form’ who regarded the ‘stalking of men as a higher form of sport’. It follows that, according to Spear, plunder was the central objective, while Sleeman states that the plunder was a ‘secondary consideration’. Second, Spear’s image of men of diverse occupations and origins driven to Thugi by social dislocation is incompatible with the Russell des­ cription of Thugs as a caste ‘or community of murderers’. Either Thugs were or were not a ‘caste’ with all that it implies: a high degree of here­ ditary entrance, commonality of pool of marriage partners, and some consensus or place in the ritual hierarchy. Third, Russell and Sleeman are at loggerheads on who the Thugs would kill. For Russell, the elaborate set of omens and prohibitions (against murdering women, Brahmins, etc.) constituted the essence of the Thug caste. Sleeman baldly says a Thug committed ‘coldblooded murder on every man he met’----2V. A. Smith, rewritten by Percival Spear, Oxford History of India (Oxford, 1958), pp. 575-6. 3James Sleeman, Thug, or A Million Murders (London, 1933), pp. 3-5. 4 Robert Russell and Hira Lai, Tribes and Castes of the Central Provinces of India (London, 1916), p. 558.

Scatf and Sword / 3 Finally, to Spear, Thugs were a product of a particular political and geographic setting. He refers to dislocation in ‘these times’ (the eigh­ teenth century) and ‘central India’. For Sleeman and Russell, Thugs were assumed to be an ubiquitous, permanent, and immutable feature of Indian life. Beyond these inconsistencies, more perplexing still is the fact that if we turn to the verbatim testimony of men arrested as Thugs, rather than European interpretation, we find that none of these three contradictory assertions is adequate to explain the social organization or motivation as revealed in the testimony. Consider first the definition given by Percival Spear: Thugs were a special type of ‘landless or uprooted men looking for a leader and reckless from despair’. From a reading of Thug pre-trial testimony, it is clear that these men were not ‘landless or uprooted’; in fact they were only part-time murderers-having regular, permanent villages and occupations which included cultivating, trading, or even serving as local officials in revenue collection. Many cases were reported of whole villages coming out to defend an accused Thug against British capture.5 We might expect Thugs, as ‘uprooted’ men, to be roaming the country­ side, pariahs of society; nothing of the sort is true. This very same European-collected testimony reveals a patterns of continuing relation­ ships between accused Thugs and local zamindars (we shall return to a precise definition of these zamindars). Numerous men who turned state’s evidence describe such arrangements. In return for the promise not to commit robbery or murder in his area, and a percentage of the booty from all expeditions elsewhere, the local zamindar not only left the bandits alone, but protected their homes and families while they were on the roads. Regular rent rolls detailing this relationship were known in the early nineteenth century.6 Were Thugs as Spear suggests, ‘looking for a leader’? Probably not. Some men arrested were of criminal families going back several genera­ tions. However, in addition to these bandit chiefs, other types of leaders emerge from the evidence: ‘(1) The man who always has at command the means of advancing a month or two’s subsistence to a gang; (2) A very wise man whose advice in difficult cases has weight with the gang; (3) One who has influence over local authorities or the native officers of the courts of justice; and (4) A man of handsome appearance and high s Charles Hervey, Some Records of Crime (London, 1892), 1,439. 6 Edward Thornton, Illustrations of the History and Practice of the Thugs (London, 1837), pp. 473-5.

4 / Marathas, Marauders, and State Formation bearing who can feign the man of rank well’ might combine the roles of all the leaders.7 If anything, there were too many leaders; bands were constantly fragmenting as new men proved their prowess and gained followers. We fare no better if we turn from this twentieth-century historian’s characterization to the formulation of a ‘community of murderers’ in Tribes and Castes o f Central India. The same European-collected evidence belies this thesis. Minimally, caste implies some commonality of potential marriage partners, some consensus on a place in a ritual hierarchy, and a great degree of hereditary entrance. None of these hold for the men arrested—they came from dozens of different castes. Hindus and Moslems ate, slept, and worked together in some bands.8Many of the men arrested were not members of long-standing criminal families, but were recruited later in life.9Also, murder and robbery was not necessarily a hereditary occupation; sons of some accused Thugs chose to support themselves in agriculture or as religious ascetics. Let us apply this European-collected data to the theory advanced by our remaining ‘authority’, James Sleeman, who, as we have seen, des­ cribed Thugs as ‘brought up in a faith which regarded the killing of men as legitimate sport’. The testimony provides numerous examples of men who were not ‘brought up’ as Thugs at all, but first joined a band as men in their twenties or thirties.10It also shows that plunder was the main moti­ vation, in contradiction to Sleeman’s statement that ‘it was secondary’. The accused men took to the roads, not in search of just any traveller, but specifically looking for banker’s agents, treasure-carriers, rich traders, and wealthy mercenary soldiers or pilgrims. Routes were chosen where booty was expected, and bodies of victims were severely mutilated, espe­ cially when anticipated booty was not found. Elaborate scales of division of booty are recounted in pre-trial testimony, a greater percentage for those actually doing the murders, and the head of the band, somewhat less for guards, gravediggers, and information-gatherers.11 STEREOTYPES AND AN AMBITIOUS OFFICER

These inconsistencies among various theories about Thugs arise in part from the common Anglo-Indian process of appropriating an Indian term 7Ibid., p. 23. * Caleb Wright, Lectures on India (Boston, 1849), p. 177. This ‘problem’ is discussed in Russell, Tribes, vol. IV, pp. 562-7 and in Hervey, Records, I, 274-5 f. 9James Sleeman, Thug, p. 73. 10Thornton, Illustrations, pp. 346, 367-8. 11James Sleeman, Thug, p. 79.

Scarf and Sword / 5 and using it to make sense of unfamiliar and inadequately understood social institutions or groups of people. These appropriations and conse­ quent distortions were of several kinds. For example, a specific office or title was made to stand for all titles with superficially similar functions. In a recent volume, Walter Neale has discussed the distortions to the term zamindar arising from differing Indian and British conceptions of the function of land.12 A second important type of distortion occurred when a regional or tribal name became a stereotype. In this dual process, the terms ceased to mean all people from a specific area or tribe, and came to refer to any person possessing a set of characteristics attributed to people of that place or tribe. For example, consider the word ‘bhil’: originally referring to a specific group of tribals, it came to mean any group of turbulent woods dwellers: The term Bheel [is] applied as a general name to all the plunderers who dwell in the mountains and woody banks of rivers in the Western parts of India; not only Bheelalahs and Coolies, who have an affinity to them but many others have been comprehended in this class. But these are in no manner (beyond the common occupation of plunder) connected with the real Bheels, who have from the most remote ages been recognized as a distinct race, insulated in their abodes, and separated by their habits, usages, and forms of worship from the other tribes of India.13

The term ‘Mehwatee’ had a similar ambiguous history. One tradition attributes the term to a name; it came to mean anyone operating—in this case plundering—in the style attributed to the people of the Mahi River region in Eastern Gujarat regardless of whether they, in fact, came from there.14 Another tradition attributes the term to residents of the lands of the Meos, pastoralists of Southern Haryana. Looking at the name ‘Thug’ in this light, we find that it was not a place or a tribal name, but a common eighteenth-century Hindi word meaning a cheat or trickster—anyone from the perennial practical joker, to a sleight-of-hand artist or a coinage swindler. As a verb, it can even be translated: ‘to amaze’. Oral tradition supports the position that the prin­ cipal nam ing of the word Thug was not even robbery, much less a particulgj! style of robbery. Among some 2000 fragments of oral tradi­ 12Walter C. Neale, ‘Land is to Rule’, Land Control and Social Structure in Indian History, ed. Robert E. Frykenberg (Madison, Wisconsin, 1969). 13John Malcolm, A Memoir of Central India (London, 1832), pp. 517-28. 14Ranvir P. Saxena, Tribal Economy in Central India (Calcutta, 1964), p. 5. Note the beginnings of the stereotyping process in Malcolm, Memoir, pp. 174-6 and its continuation in Hervey, Records, n, p. 511.

6 / Marathas, Marauders, and State Formation tion—both in vernacular and in translation—from Central India, I find many stories about robbers, but none specifically about Thugs.15 Through the first half of the nineteenth century, we can follow this process of stereotyping in the British official writing. Increasingly and steadily, differences between bands called Thugs were ignored, and similarities between them became exaggerated. Their fragmented orga­ nization came to be represented as a dangerous, widespread conspiracy. First described by a Madras medical officer, Dr Sherwood, in 1816 (in an article in the Madras Literary Gazette16), the Thugs virtually dropped from official British notice until the mid 1830s, except for a few careful British observers who saw great diversity in Sherwood’s so-called Thug bands. These observers noted that, some bands were Hindu, some Moslem, while some were mixed caste and community. Some attacked openly in large groups, and did not bury their victims; others attacked only at night and took great care with burials.17 In the early 1830s, a few military officers observed that numbers of sepoys who had been discharged with pay failed to return the following season. Further checks showed that they had not arrived in their home villages. At that point an ambitious military officer, William Sleeman, decided on reading the Sherwood article that these sepoy disappearances were caused by an India-wide Thug brotherhood of murders. He set out to prove this thesis to the Government of India, but more than five years of official scepticism greeted his efforts. As George Bruce, twentiethcentury historian of Thugs, explains: Sleeman found himself up against a British wall of avowed disinterest, even hostility towards any organized investigation. Government House, and most of his fellow magistrates argued that if Thugs existed, and there was no proof, they were members of a religious fraternity; and Company policy was not to interfere. Thus, apart from a few notable exceptions among fellow magistrates, Sleeman was almost alone in his determination to stem the flow of murders.18

More than any other person, William Sleeman is responsible for the stereotyping of the word Thug. Many of our ‘primary’ documents on Thugs were produced by Sleeman in this period, and they are anything but 15Principally in Stith Thompson, The Oral Tales of India, Indian University Folklore Series, no. 10 (Bloomington, 19S8). 16Quoted in full in George Bruce, The Stranglers (New York, 1968). 17Extract of a letter from the Magistrate of Etawah, 1816, in Thornton, Illustrations, pp. 322-4. 18Bruce, Stranglers, p. 36.

Scarf and Sword / 7 unbiased. By omission of differences and emphasis on similarities, he repeatedly made the case for a widespread Thug conspiracy. Sleeman’s eventual success in persuading the Government of India is a gauge of the triumph of the Evangelical, crusading philosophy of British Indian administration over the ‘preservation-of-Indian-customs’ philosophy of the Orientalists. Official support for Sleeman’s campaign came less than two years after Macauley’s Minute on Education. Even after Sleeman made a name and office—The Thagi and Dacoity Department—for himself, it was necessary for him to keep the Govern­ ment of India reminded of the widespread Thug conspiracy; only this conspiracy justified the special judicial practices, blanket military aid, pay for spies, and special treaties with the native states allowing for free pursuit by the Department. Charles Hervey, Sleeman’s successor in the Thagi and Dacoity Department, was quite explicit on this point: What is association or combination for criminal purposes? It is a secret society centred in certain well-understood oaths and rules, that bind and guide all those admitted into it, as agents for carrying out unlawful deeds in combination one with another-members with members, tribal congeries with their congeners, leaders of gangs with leaders—all the threads of whose action are drawn to those set rules and customs (each tribe having its own distinctive modes of guidance and procedure,) as the guiding centre of each organization, all mutually depending upon each other, and, as well, upon that ‘governing focus’ (so to term those rules and principles which guide all,) as between all the atoms of the several confederacies. If, then, there were no such organizations, the secret associations concerned, would crumble away; and it is to reduce them to that extremity, to the effacement that the agents of the Government, that is the special Thuggee Police,

are called into existence, themselves acting in like manner upon their own set rules, by which to enable them to do so.19 He farther argued that it was the organization and its ‘crimes, secret, dark­ ly planned, esoteric’, that demanded the special legislation allowing for the death penalty or transportation for association alone. The ‘Thugi’ laws—resisted by many judges—provided ‘special punishments on proof, not of any specific act taken by itself, but of association . . . on proof of a general charge, that the arraigned party “belonged to a gang of habitual dacoits (or Thugs, etc.); while engaged in dacoities, etc.” irres­ pective of whether the convicted party was or was not present in any one of the specific instances which supported the general charge upon 19 Hervey, Records, II, pp. 445-7. One Such ‘Thugi’ law is quoted in James Sleeman, Thug, p. 134.

8 / Marathas, Marauders, and State Formation which he was convicted.’ \ . it was for association per se that he [a Thug] had been sent up for trial!’20All the reports of the Thagi and Dacoity Department—whether by Sleeman, Hervey or others—supported this conspiracy position. ' In spite of thirty years of Thagi and Dacoity Department reports and popular books such as Caleb Wright’s A Description o f the Habits and Superstitions o f Thugs (1846) and Edward Thornton’s Illustrations o f the History and Practice o f the Thugs * 1837) which took uncritically—often verbatim—from Sleeman’s writings, the term was still not fully stereo­ typed until at least the late 1860s; that is, even within the government hierarchy there was no consensus on the answers to the question ‘Who or what is a Thug’ and ‘What are the characteristics by which one knows them’? In 1867, Charles Hervey, then head of the Thagi and Dacoity Department, berated fellow government officials who could not seem to recognize a case of Thug murder when they saw one, some correctly recognizing Thuggee in instances which were palpably the deed of experts, although death should not have taken place; others only doing so where death had resulted; some classing certain murders as cases of ‘Thuggee’ without reference to the means resorted to in the perpetration thereof; others who wholly pass by cases of poisoning whether followed by death or not, although they bore evidence of being the acts of class criminals; some who restrict their notice to selected cases only of its occurrence, passing by other similar instances; some who endeavour to distinguish between different degrees of poisoning, some calling ‘murder by poison’ Thugee, others not doing s o . . . others who lump all such kindered offenses under round numbers without any narration of the attendant circumstances, contented only with quoting against them the sections of the Penal Code under which they were triable or were tried.21

Now perhaps we can understand the confusion in the various theories of Thugi we have examined; the late-nineteenth-century official, the military officer, and the twentieth century historian are surely to be forgiven if British officials at mid-nineteenth century—contemporary to the Thug phenomenon—could not figure out what a Thug was! The only people who seemed positive that they could tell a Thug when they saw one were from the Thagi and Dacoity Department. This self-assurance is deceptive. If, for example, we take the ‘con­ spiracy’ position as constantly articulated by the Thagi and Dacoity Department and analyze it with verbatim Thug testimony collected by the “ Ibid. 21Hervey, Records, I, pp. 50-1.

Scarf and Sword / 9 Department, we find no evidence of widespread organization. For example, Malwa Thugs were unaware of the practices of stranglers in Bengal or the Deccan. Spoils were never shared with any far-flung Thug hierarchy. There was no consultation even within a single region on departure times or division of prime highways among bands. All these decisions were reached locally between band and leader. Discipline was never dispensed by a broad hierarchy; dissatisfaction with performance by a member merely meant expulsion from the band; dissatisfaction with an inept leader merely meant that his followers drifted off into other bands. Numerous examples exist of men—satisfied with their booty— returning home in the middle of an expedition.22 Indeed, the British method of capture, using confessions of men who turned state’s evidence against other gangs and individuals, suggests most strongly that there was no central organization. Reprisals were not taken against the witness or his family once the immediate gang was captured. One need only recall what the Mafia traditionally does to ‘sqealers’ to see the low level of Thug organization by comparison. Thus, the confusion of later historians regarding Thugs ultimately comes from the unsupported theories generated by the Thagi and Dacoity Department itself. Attempts to resolve contradictory attributes of a ' Department-invented class of murderers alleged ‘Thugs’ have led into a welter of vague, contradictory, and inadequate characterizations. For lack of an adequate definition, the term ‘Thug’ will hereafter appear within quotation marks. Clarity will not come to this term from more re­ reading of the Department reports, unless the biases and assumptions are known and challenged. A theory that is the product of a study of eighteenth-century Indian institutions and social and political conditions, rather than Victorian morality, is called for. The remainder of this paper is an attempt in that direction

FROM THE CULTURE OF THE YELLOW SCARF TO THE BANDITS OF MALWA

Even if ‘Thugs’ were not tied to an India-wide organization but operated in a local and fragmented manner, did they perhaps form a culturally distinct group? Do contemporary depositions reveal, for example, that all ‘Thugs’ recognized a unique set of customs and spoke a distinctive argot? 22 The evidence is most extensive in William Sleeman’s Ramaseeana (Calcutta, 1836) and Rambles and Recollections of An Indian Official (London, 1915).

10 / Marathas, Marauders, and State Formation The answer, rather surprisingly, is negative. There was apparently no clear Thug subculture and many characteristics attributed to them alone—such as invoking the deity before starting on an expedition, mak­ ing objects involved in the occupation sacred, failing in with travellers by deception, omens and signs, the specialization of tasks within gangs— prove to be common to many criminal groups. For example, invoking the deity: Robbers and murderers of all descriptions have always been in the habit of taking the field in India immediately after the festival of the Dasahra, at the end of October, from the sovereign of a state at the head of his armies, down to the leader of a little band of pickpockets from the comer of some obscure village. All invoke the Deity, and take the auspices to ascertain his will, nearly in the same way; and all expect that he will guide them successfully through their enterprises, as long as they find the omens favourable. No one among them ever dreams that his undertaking can be less acceptable to the Deity than that of another, provided he gives him the same due share of what he acquires in his thefts, his robberies, or his conquests, in sacrifices and offerings upon his shrines, and in donations to his priests.23

This paper reports preliminary rather than completed research, and much remains to be done on the comparative symbolism and organization of criminal bands (and all habitually-travelling groups). At this stage, I do not want to reject the possibility that all leaders or bands called ‘Thug’ by the British shared some common symbols from a larger Indian tradition, such as left-handed Tantrism, but the only distinctive and unique features that I find in all Thug groups so far studied are the use of a scarf to strangle victims and the occasional eating of gur, a coarse sugar. The former is not a sure feature, since groups sometimes also used poison. A sticky bowl of sugar and a limp yellow scarf are slim evidence indeed for calling ‘Thugs’ a ‘trade union’ or a ‘fraternity of murderers’. If we are stymied in a cultural definition of ‘Thugi’, a geographical approach proves more fruitful. If we plot the places of residence of the Thugs the British captured, these appear to be predominantly in Malwa, which the British called Central India. This area, in the eighteenth century, was the haunt of many sorts of marauding groups, big and small, which attacked not only within Central India, but also in Oudh, Bengal, Rajputana, Hyderabad and Mysore. All these groups, whether they struck whole towns, villages, or bands of travellers, depended on the weakness of the authority of any government above the local level. The local watch 13William Sleeman, Rambles, pp. 296-7.

Scarf and Sword / I I had no jurisdiction outside his immediate few square miles; so the marauding groups could escape after looting the village or murdering the travellers. Capturing the marauders was also beyond the means of those immediately affected—the victims’ relatives or business asso­ ciates. Only an overarching, non-local government could possibly mount the large-scale military or police expedition needed to suppress roving bands.24 The freedom of operation of such bands in eighteenth-century Malwa is striking; this freedom strongly suggests the relative strength of local as opposed to overarching governments in the Malwa region which, viewed from a dynamic perspective, as we shall see, suggests a hypothesis of state formation. This theory, it is hoped, will provide an adequate conceptual framework for understanding the phenomenon that has been called ‘Thugi’. First, however, let us place these ‘Thugs’ and other marauders in a political and geographical setting. MALWA AND THE POWERS THEREIN

Malwa—it calls to mind a vague area south of the Jumna river and east of Gujarat. Some lack of clarity is quite understandable, since the boundaries shifted with successive empires, Malwa being predominantly a rich province and seldom the seat of empire. In Mughal times, the province of Malwa extended well into Rajputana, including Harrowtee (later split into Bundi and Kotah) and part of Mewar (Oudipur), and encompassed the hilly tracts of Rath Bagur and Kantul, running south from Oudipur to the Vindhyas and separating the Malwa plateau from Gujarat. At times ‘Malwa’ also included parts of Bundelkhund to the east. The Malwa plateau, as opposed to the political division ‘Malwa’, is quite easily defined. It is a table land approximately 2000 feet high bounded on the northwest by the hills on the west bank of the Chambal, on the north by the Jumna, and on the south by the Narmada. The eastern boundary is a hilly tract running north-northeast from the eastern edge of Bhopal separating it from Gondwana and Bundelkhund. The western 24 Such pursuit, and the consequent general safe movement on the roads, was clearly in the interests of such an overarching government, for a number of elementary reasons: (1) Safe movement of land revenue, predominantly in cash, but also in kind from the hinterland to the capital: (2) Easy communication from the capital to the monarch’s agents in outlying areas; (3) Rapid movement of armies to rebellious areas without expending men and equipment fighting thieves and plunderers en route; (4) Sure import into the capital of certain necessary articles—such as horses or guns—and certain desirable articles—such as fine cloths, jewels, spices, etc.; (S) Increase in revenue fromtolls paid by traders, and (6) Religious merit and an increase in symbolic authority from building tanks and caravanassaries.

12 / Marathas, Marauders, and State Formation boundary is likewise a hilly tract—some fifty miles wide—running north from the Vindhya range between 74 and 75° east latitude. These hilly tracts on the east and west of Malwa (as well as the Vindhya range) are occupied by tribals distinct from the people of the plateau proper in colour, language, customs, religion, and style of farming—the tribals using shifting agriculture. On the plateau, the rainy season comes in June or July and through August and September it remains mild (72°-7° F); from October through December to February, it gets progressively colder and often goes below freezing at night. March through May finds hot weather, days in the 90s and above, nights in the 70s and 80s. In this climate, the 'loose rich black loam* produced, in the eighteenth century, cash crops such as cotton, sugar-cane, opium, indigo, and tobacco and large crops of grain; particu­ larly wheat, but also gram, peas, jowry and bajrie. Many areas had both a Kharif and a Rabi crop. The basic feature of the eighteenth-century historical process taking place within this geographical framework was the multiplicity of states, a hundred by one count, as many as six hundred by another. Any attempt to summarize the political history is clearly impossible in a paper of this scope. (See Raghubir Sinh, M alwa in Transition o ra Century o f Anarchy: the First Phase 1698-1765, Bombay, 1936, for the histories of some larger states). Perhaps a few dates and events will be some use in under­ standing the model which follows. Malwa was conquered in 1387 by Muslim invaders from the north and became a province of the Delhi sultanate. The next hundred and seventyfive years saw a three-way battle between Delhi, some Rajput houses and independent Muslim kingdoms. The Plateau was attacked by Sher Shah and decisively conquered by Akbar in 1561; many Rajput houses (which were to later achieve prominence) became established as small states in the next fifty years with the help of Akbar and the Mughal court. A ‘second wave’ of Rajput invaders came into Malwa starting in the 1670s. They were collaborators with the Mughals and received grants of land in Malwa in reward for their service. These Rajputs considered themselves purer than those who had come a century earlier and unlike their predecessors, they did not inter-marry with the local populace. These second-wave Rajput kingdoms were, thus, not well established when the Maratha invasions began in the late seventeenth century: [In states] which had been founded in the last years of Aurangzib’s reign—and the number of such States was the greatest in Malwa—the founders as well as their early descendants did not get enough time to organize their States nor to secure their hold over the lands and the people within the State, as they were kept

Scarf and Sword / 13 busy in the distant South with the Imperial army. And the States yet unorganized and unstrengthened could least help the Empire in times of disorder, for they had first to grapple with the question of their own existence.13

The series of Maratha forays into Malwa (which included upwards of 12,000 horses) began in 1698, had a temporary hiatus between 1703 and 1713, then began again in earnest. Yearly raids became regular, and by 1725, the Peshwa had appointed tribute collectors in Southern Malwa. (Incidentally, other invaders were also taking over areas of Malwa at this timed. DostMuhammad Khan Rohilla laid the foundation of Bhopal State between 1709 and 1717). After decisively beating the Mughal army in 1728 and settling with Jai Singh of Amber, the Peshwa granted land to his major generals, the Pawars and Holkar, in Malwa all through the 1730s. The conquest of the rest of Malwa followed during Nadir Shah’s invasion in the North. Finally, the legal recognition of Maratha supremacy was given by the Emperor when he granted the peshwa Naib Subahdari rights over Malwa, and with the death of Jai Singh in 1743, no power was strong enough to oppose them. In the 1740s, Rajputana was the scene of Maratha invasions. Often called in by one side in a succession dispute, one general or another quickly established rights to tribute. By the mid 1750s, Kota, Bundi, Mewar, Marwar and other Rajput states were committed to give tribute to the Marathas even if they strongly resisted paying it. Ajmer, the strong­ hold in central Rajputana, was turned over the Marathas in 1756. A new opponent appeared in the 1750s, the Afghan Abdali. The initial invasions of 1750 and 1752 were followed by much larger-scale invasions in 1757. Maratha power north of the Chambal River was virtually elimi­ nated even before the military disaster of Panipat in 1761. Dozens of Maratha chiefs died on the field at Panipat and a whole new generation figures in events after 1765. These include Mahadji Shinde, Ahalyabai Holkar, Tukoji Holkar, Zalim Singh Jhala (of Kota) and the Pindari leader, Amir Khan. Only by the late 1760s was Maratha power restored; and the invasions of Rohilkhand, of Jat territory near Delhi, and of Oudh began. The death of the Peshwa in Poona in 1772 and support by the British Bombay government of one of the claimants (Raghoba Rao) set off a decade of war and negotiation which has been well chronicled elsewhere. (For Sindia’s part in it see S. N. Roy, A History of the Native States of 25 Raghubir Sinh, Malwa in Transition ora Century o f Anarchy: The First Phase 16981765 (Bombay, 1936), pp. 15-16.

14 / Marathas, Marauders, and State Formation India, vol. I, Gwalior (Calcutta, 1888). The result was a draw; status quo ante prevailed at the Treaty of Salbai in 1781, with the important ex­ ception that the Company opted out of all affairs in Central India, Delhi, and Rajputana. The period from the Treaty of Salbai to 1795 was the ascendency of Mahadji Sindhia; he was the arbiter of Delhi affairs, and the strongest power in the Maratha confederacy; after his defeat by the Rohillas in 1788, he became one of the foremost military powers in India, via his shift, with the help of the French adventurer De Boigne, to a Europeantype infantry and artillery. The mid 1790s clearly mark the end of an era. Ahalyabai Holkar died in 1795, and Tukoji Holkar followed her two years later; thereafter the Holkar kingdom was in chaos. Chote Khan, minister of Bhopal and an important figure of stability died in 1795. Both the Peshwa and, more importantly, Nana Famavis—the holder of any real power at Poona— died in the middle of this decade. Mahadji Sindhia’s death preceded Nana’s by only a year and De Boigne—Sindhia’s general—left India in 1796. Finally the active involvement of the British in the affairs of indigenous states began once again with the ‘protection’ of the Nizam in 1795. A STRUCTURE AND A PROCESS26

Let us look at Malwa a few years before these morbid events. Beginning with a static view, Central India in the late eighteenth century had a five­ tiered power structure. The Peshwa of Poona stood at the top. His real power in day-to-day affairs was small, but his authority to grant or 26 The following were used extensively in developing both the static power structure and the dynamic process models: Broughton, Thomas D. Letters Written in a Mahratta Camp, rev. ed. (London, 1892). Cohn, Bernard S., ‘Political Systems in Eighteenth Century India: The Benares Region’, Journal of the American Oriental Society, CXXXn (July-September, 1962), pp. 312-20.

Forster, George,AJourneyfromBengaltoEnglandthroughtheNorthernIndia... (London, 1808).

Franklin, William, Military Memoirs of Mr George Thomas. . . (Calcutta, 1803). Ghosh, B. British Policy towards the Paihans and Pindaris in Central India 1805-1818 (London, 1963-4). Habib, Irfan, ‘Banking in Medieval India’, Contributions to Indian Economic History, vol. I (Calcutta, 1960), pp. 8-14. Luard, C. Eckford, ‘Some Views of an Indian Ruler on the Administration of an Indian State’, Asiatic Review, New Series, xxn (1926), pp. 278-98. Malcolm, Sir John, A Memoir of Central India (London, 1932).

Scarf and Sword / 15 withhold sanads on land conquered by Maratha chiefs, was very impor­ tant. Also, the major chiefs—Sindhia, Holkar, and the Pawars—were both legally generals in the Peshwa’s army and owed their initial rise to the Peshwa’s patronage. There are numerous examples of Maratha chiefs returning to Poona at the Peshwa’s behest to engage in political and military actions. The second level was the Maratha conquerors, such as Sindhia and Holkar. Current literature identifies an ideal-type of Central Indian Maratha administration, which included both local and state revenue and treasury officials, a political department, registry officials, the prince’s household, and pay-and-muster officials for the army. The state also supported a large military establishment, perhaps numbering up to 20,000 infantry, 10,000 cavalry, and 5000 garrison troops. A third or less of this army would be under the direct control of the prince; the rest would be under subordinate Maratha chiefs who as assignees of lands for support of troops constitute the next, or third, level in the power structure. The prince had virtually no control over internal administration of assigned lands. These subordinate generals were bound to the prince by no ties of family or clan. East and west of the Chambal River, below the Marathas stood some fifteen major and many minor Rajput clans some of whom had come as conquerors some centuries before and obtained Mughal sanads over their possessions; others were of recent origin (see above, p. 18). Cousins, brothers, kinsmen held separate estates, over which a senior member with large holdings had some influence. Beyond some common finances and obligations in the field to the head of the clan, estate management was [An Officer of the East India Company]. Origin of the Pindaris (London, 1818). Roy, Surendra Nath, A History of the Native States of India, vol. I, Gwalior, (Calcutta, 1888). Sardesai, G. S., ‘The Rise of Mahadji Sindhia’, Modem Review, CXXV (March, 1944), pp. 2 0 9 -u . Sarkar, Jadunath. Persian Records of Maratha History, vol. I, Delhi Affairs (Bombay, 1953). Sen, Suiendranath. Administrative System of the Marathas (Calcutta, 1925). Sinh, Raghubir. Malwa in Transition or a Century of Anarchy: The First Phase 16981765) (Bombay, 1936). Sinha, Har N. Selections from the Nagpur Residency Records, 3 vols. (Nagpur, 1950).

Tod, James. Annals and Antiquities of Rajasthan (Calcutta, 1898). Twining, Thomas. Travels in India a Hundred Years Ago, with a Visit to the United States (London, 1893). Tytler, Alexander. Considerations on the Present Political State of India (London, 1816).

16 / Marathas, Marauders, and State Formation autonomous. The military contingent might range from under 1000 for a small state tike Banswara to 15,000 for Kota, the largest independent Rajput state close to the Maratha dominions. These troops were usually scattered among many estates and concentrated at the fort which was a universal feature of such estates. Finally, below the Rajputs was one more level before one reached village administration. The men at this level were called zamindars of parganas or watandars. They were large local landholders, who con­ tracted the revenue for an area of perhaps twenty square miles and extracted a fee from other cultivators. They maintained a small body of troops (25-150) and often fortified either a village or their own homes. These watandars were definitely drawn from a variety of castes and were—in some sense—a holdover from the Mughals since most held sanads from the Mughal emperor, though a few had sanads from the Peshwa, for special services. Power-structure analysis, though neat, and often done for various parts of India, is, I think, substantially misleading. Without an understanding— not of the structure—but of the process by which these various relations developed, we cannot hope to understand the importance of any specific group, like a small ‘Thug’ band. The diagram which follows is a flow-chart, representing the general process of state formation in Central India in the eighteenth century. The dynamic is the universal desire for maximum stable land revenue, not land as such, and maximum political control, that is reducing the largest possible number of people to clientship or dependency, relations. The counter-dynamic is the short-run need to pay troops. Groups and indi­ viduals entered this on-going system at various times throughout the century. Malher Rao Holkar and Ranoji Shinde raised their troops and moved to the ‘invasion’ stage by the 1730s, while Amir Kahn entered some forty years later. Note that there are two entrances (at the top), and only one exit—establishing a ‘pax’ at the bottom. The majority who did not reach the ‘pax’ stage were thrown back on plunder as the only means of paying the troops. If we look at Central India at any given moment, we find individuals and groups at all stages in this process; some were besieging strongholds, some negotiating with bankers, some plundering their neighbours, some finding allies among Rajputs or Bhils, and a few ruling peaceful and prosperous states. Let us by way of illustration follow the fortunes of an individual through this system. He was presented at court—perhaps the Peshwa’s— through the influence of a patron. Proof of ability lead to increasing num-

Scarf and Sword / 17 Individual service

4

Money and "symbolic I— ^ sanction" from outside power

4

raising a body of followers Invasion • Plunder Extortion ___ ■Robbery

clan, tribal or religious

group with some leadership

1

»selling services to some power

- expedition to pay troops

expedition called by holder of “symbolic sanction1

Coming to terms with local powers . with allies oust present W— holder and establish right

to some percentage of harvest Finding a banker who w» advance cash against future revenues

U

by force or fraud gain a stronghold; collect “protection" from nearby cultivators hold the fort; keep down subordinate powers when main army is in the field. Some regular revenue collection reduce subordinates forts; increase own share of revenue

troopsmutiny

’ no place to— retreat to; or store supplies ■

-----►“Protection" by larger power

no revenue mutiny desertion to pay •main —------- t plunder home — troops 's . districts “protection" by larger power subordinates seize _ revenue

t mutiny of own troops "assign" large part of jands to military commanders

pawn lands for cash on some part of holdings turning subordinate powers into officials

sporadic revenue collection

establish a “pax"

Stable revenue to pay troops

assuring rights of chosen successor

------- * revolution from below

faction-fighting over succession may destroy state

troop mutiny-

18 /Marathas, Marauders, and State Formation bers of troops under his control. The individual ruler he served decided that an area, either unconquered or in a state of rebellion should be invaded. Our ambitious parvenu was given just enough money to raise troops, rather little since mercenaries furnished their own horses and equipment, and a ‘symbolic sanction’, for example a sanad to collect chouth in the area, or the ‘right’ to quarter his troops on the conquered lands. Invasion and plunder followed, with the extortion of towns or villages which paid not to be plundered. He then came to terms with local powers. This is meant quite literally. For example, the invading Marathas offered lower land revenues than the prevailing Muslim administration. The Rajputs switched allegiance quite readily. The invader needed to use them to collect revenue. The aid of these lower-level powers might be active, on the side of the invader, or passive, merely withholding aid from the formerly dominant power. The next stage—and it is a crucial one—was finding a banker to advance cash against future revenues. Until some banker considered the invader’s claims legitimate enough and his prospects for collecting regular revenue bright enough, he was forced to plunder somewhere every few months to pay his troops, especially if his ‘initial sponsor’ made no payments beyond the first one. The advance by the banker was on a yearly basis (it was a form of risk capital) and allowed the undertaking of a siege, a long-term proposition often requiring heavy equipment Reliance on a banker became more and more necessary as the eighteenth century progressed because continuing support from the original source of troops and ‘symbolic sanction’ (either the Peshwa or the Emperor at Delhi) became less and less dependable. For example, visits by the Peshwa’s Vakils to the North became more sporadic and less likely to be accompanied by any military force. The steps from the ‘seizing of a fort’ to reducing ‘subordinates’ forts’ represented increasing stages of likelihood of regular revenue collection. For every successful movement to a later stage, there was the alternative of failure. Reaching any stage did not imply a stable situation, as there was still not enough land revenue to pay troops. We shall return to the impli­ cations of this. ‘Reducing subordinates’ forts’, may be viewed as a transi­ tion stage; it was possible to pay troops out of land revenue, but equally possible that plundering had to be resorted to. The next three stages represent the exit from this system of necessary plunder. They differ only in the degree of development of government infrastructure. The ‘turning subordinate powers into officials’ stage is, perhaps, best exemplified by the lands held by De Boigne, Sindhia’s

Scarf and Sword / 19 general, in the late 1780s and early 1790s. He had a regular revenue administration and audits; he developed agriculture, cast guns and protected the land from marauders. In contrast, the ‘pax’, the next stage, required many more government functions— measurement-type revenue system, regularizing dues from travellers, minting money, a state justice and police system, regular diplomatic relations with other states, cash salaries for troops, continuity of members of administration anfl neutral­ izing internal marauding groups by grants of land. Examples of a kingdom as stable as this are hard to find: Ahalyabai Holkar’s kingdom would probably qualify, Zalim Singh’s Kota might, as would Sindhia’s domin­ ions for short periods. From initial invasion to stable exit was at least a thirty to forty year process with heirs completing the process started by the father; only Zalim Singh did it in his own lifetime. The line running from the bottom to the top of the chart, on the left-hand side, represents the service which had to be given to the power that furnished the initial troops. For example, the Peshwa might call Sindhia to Poona to undertake a joint mission against Tipu or the Nizam. This call might come at any time and involve a war during which the newly conquered area might rebel. Thus, the call to service could entail a loss of several stages (move­ ment upward on the chart). What is the utility of this rather complicated model? First, I think it organizes and makes sense out of what has always been assumed to be a completely chaotic area in this period. Second, there is evidence that at least some people at the time were operating with this process in mind: [Mahadji Sindia’s] genius saw that to realize his plans, the mere predatory hordes of the Mahrattas could never prove adequate. It was a circle of plunder, and, as one country was exhausted, the army had to march, with numbers increased by those whose condition their success had made desperate, to ravage another. They had, in their first excursions, little or no means of reducing forts; nor did their system of war admit of protracted hostilities in a difficult country, and against a resolute enemy. These wants were early discovered by their enemies. The

Bheels from their mountains, and the Rajpoots and others from their strong holds (which were multiplied by fortifying every village), not only resisted, but retorted upon the Mahrattas, by laying waste their lands, the wrongs they had suffered. This evil was only to be remedied by a regular force. We are distinctly informed, that its existence led Madhajee Sindia to determine upon the measure he now adopted, of raising some corps of infantry.27

Third, it suggests that ‘states’ and ‘marauders’ were not different in kind, 27John Malcolm, A Memoir of Central India (London, 1832), pp. 126-7.

20 / Marathas, Marauders, and State Formation but only in relative degree of success in conquest, revenue collection, and infrastructure-building. All were involved in the same process, with the same ends, using the same sources of legitimation. The striking feature of the second half of the eighteenth century is not the existence of this process, but the shift from a predominance of relatively stable states to a mix containing great numbers of marauding groups. I would now like to lay out a hypothesis to explain this change, which in the end returns to our beleaguered term ‘Thug’. The hypothesis is that after the battle of Panipat in 1761, few Marathas accompanied the Maratha chiefs when they once again established their power north of the Narmada river. The Maratha chiefs were, therefore, forced to rely on mercenaries (some of whom received the new name ‘Pindari’): The strength of the Maratha Cavalry continued to be its most distinguishing feature dll about the year 1750, when contact with the French and British armies discovered [sic] the superior advantages in modem wars of regularly-trained infantry batallions protected by artillery, the third arm in modem warfare. The success of the English and the French induced the Maratha leaders to have recourse to this new agency and for the first time we find mention of the Gardis or the trained batallions. The weakness of this new addition to the military force

consisted in the fact that, unlike the Mavales or the Shilledars, who each owned his plot of land and served the State, not as mercenaries, but as militia, the Gardis were mercenaries pure and simple, made up of foreign recruits of different nationalities, who had to be paid fixed salaries all the year round, and only owed loyalty to the commanders who paid them their wages.28

Other powers already in the area then augmented clan troops with greater and greater numbers of mercenaries. Th demand for cash rose astronomi­ cally, since mercenaries had to be paid virtually year-round and could not and would not be paid in small grants of land. The continued success of the British in using trained mercenaries only made the situation worse: to compete, indigenous states had to arise European-type infantry, which cost from four to ten times as much as native cavalry. For example, 1000 European-type infantry troops cost 26,000 Rs per month in 1795.29 Perhaps more importantly, because of the new European-type artillery, the military advantage shifted drastically from the defensive to the offensive power. Hill forts, especially small ones, provided little protec­ tion, and long sieges were replaced by intense cannonades and rapid * M. G. Ranade, ‘Introduction to the Peshwa’s Diaries’, Journal o f the Royal Asiatic Society, Bombay Branch, XX (1900), pp. 456-7. 29Suiendranath Sen, Administrative System of Marathas, rev. ed. (Calcutta, 1925), p. 445.

Scarf and. Sword / 21 breaching. Thus mobile marauding groups—with efficient artillery— competed militarily at least on equal terms with established states; leaders of states who could formerly retire to hill forts now had to meet the marauders in the field, with baggage trains of supplies and expensive mercenaries and artillery: Under the later Peshwas, these forts appear chiefly to have served the double purpose of State granaries and State prisons. State prisoners were sent to the forts for custody, and the condemned criminals of both sexes were sent there for penal servitude. In the latter half of the century, the forts are chiefly mentioned in this connection. Against the more improved means of warfare represented by the artillery, these hill-forts ceased to be valuable for the purposes of defence, and in many places were neglected and allowed to go into disrepair. In the war with the English, the forts offered little or no protection, and submitted without firing a shot.30

All this cost money. More and more states were forced into the position of being unstable, in the sense that the regular revenue was insufficient to pay their troops. Every year these states had to use the troops in one or more of three ways, (1) forcing more revenue from their own lands, (2) attacking neighbouring lands, or (3) allying for the purpose of war and plunder with another power. Every decision to go on a campaign was like a chance card in Monopoly. It could bring gratuitous revenue, but the risks were very high. If it went on too long, the leader might be forced to plunder his ally’s lands in order to pay his troops, or mortgage the revenues of his own lands to a banker at a great discount, for ready cash. There was also a high probability of revolt from below every time the major part of the army left its home territories. Thus there was operating a ‘plunder dynamic’ of the powers higher on the chart (‘marauders’) trying to seize the treasury and standing crop of powers lower on the chart (‘states’) while attempting to establish longer-term rights to revenue. Powers lower on the chart—having more stable revenue—were often forced to plunder when taxes and tribute were insufficient to pay the troops (for example, because the standing crop had been seized by marauders). Those hurt worst in this process were at the lower levels of the power structure, previously analyzed. The overlord (perhaps a Maratha prince or his assignee) was constantly trying to increase his percentage of the crop (often by military means) while the rest was likely taken by marauding bands. Reliance on income from local produce became more and more chancey. Therefore, every level in the power structure tended to link up with and 30 Ranade, ‘Introduction’, p. 459,

22 / Marathas, Marauders, and State Formation give protection to some groups of part-time marauders, to have a non­ local source of revenue. These ranged in size from Pindari armies of ten thousand horse to a motley handful of thieves. John Malcolm, the first British administrator of Central India writing in the 1820s enumerated eleven types, or more properly styles, of marauders in late-eighteenthcentury Malwa. To each was attributed certain characteristics of organi­ zation and operation, by the stereotyping process discussed above. Most of these styles—including that identified by Malcolm as Thaggi-were those of locally-recruited, locally-based marauders.31 Malcolm empha­ sized the point here. Almost every large village which retained its inhabitants subsequent to the ravages of Jeswunt Row Holkar and the Pindaries had a band of this description either living in it or in communication with the Potail; and the latter received, for the countenance and support he gave them, a fixed share of the booty.32

Research to date, thus suggests^that what the British saw as ‘Thug’— ‘a national fraternity of murderers’—consisted of a small core of families members of which had been murderers for several generations. These and many other men recruited bands in their local areas during the severe dislocation of the last decade of the eighteenth century and especially after the British defeat of the Marathas in 1803. After 1815, the British were occupied with establishing regular relations with various levels of the prevailing power structure and with destroying the large scale marauders (‘Pindaris’ and tribal groups); smaller-scale groups (such as ‘dacoits’ and ‘Thags’) flourished, preying on the traders and pilgrims travelling on roads which were somewhat less dangerous because of the elimination of the large-scale marauders. It was the writing of William Sleeman and the Evangelical, crusading tone of the British Indian administration of the 1830s that played up these locally-organized, smallscale marauding groups (given the name ‘Thugs’ by the British) into a hideous, widespread religious conspiracy, somehow typical of India and Indian ‘national character’. We cannot and will not know the nature of the ‘Thugs’ or any other marauding group of the eighteenth century until we return them to a historical and geographic setting, and view them in the context of the on-going structure and processes of power.

31 Malcolm, Memoir, pp. 174-91. 32 Ibid., pp. 173-4.

Chapter Two

The Slow Conquest: Administrative Integration of Malwa into the Maratha Empire, 1720-60 TOE EIGHTEENTH CENTURY, THE MARATHAS, AND MALWA

The eighteenth century in India has generally been described as a period of great turbulence, characterized by march and counter-march, rising and failing fortunes, and bewildering political intrigue. Many historians, focusing on this aspect, have dismissed the century as merely an un­ savoury hiatus between the collapse of the Mughal Empire and the rise of British domination. Yet there was more to the century than the march andcounter-march of armies. The other aspect of the period was the emer­ gence of strong successor states in Gujarat, Bengal, Oudh, Malwa, Hyderabad, Mysore, and the Punjab. Recently, historians have begun exploring these successor states, looking both back towards the Mughal administrative and ideological heritage and forward towards their role as princely states in British India. There are also important issues within the century itself, such as the role of successor states in developing regional language and consciousness, and successor states as channels of eco­ nomic and social mobility. For historians prone to see only ‘chaos’ and ‘anarchy’ in the eighteenth century, the Marathas more than any other group typified the period. Thompson and Garratt, for example, termed them ‘cool and insatiable robbers’, ajudgement based perhaps on their raids on Mughal supply lines and extractions of naked tribute (termed chauth, or one-quarter).1It seems supremely ironic that these supposedly archtypical marauders, the Marathas, of all the successor states should have left the most complete administrative record, permitting the broadest questions and the most detailed answers. The research for this study was carried out in Poona during 1969-71 with the support of a Fulbright-Hayes Language Fellowship. I wish to thank B. D. Apte, A. R. Kulkami, J. H. Broomfield and M. Gluchkov for their support during the project. Reprinted from Modem Asian Studies, n, 1 (1977). 1 Edward John Thompson and G. T. Garratt, Rise and Fulfilment of British Rule in India (London, 1934), p. 63.

24 / Marathas, Marauders, and State Formation This record remains almost unknown. When the British captured Poona, the Maratha capital, in 1818, they seized an enormous mass of administrative papers which began around 1725 and continued until the end of the eighteenth century. In the nineteenth century, these papers were rearranged to suit British needs (primarily as source material for settling land disputes in Maharashtra) and closed to scholarship. Early in this century a few historians were allowed access and several volumes of transcribed documents were printed, but it is only since independence that their diversity, geographic scope, and wealth of detail have become known. This mass of documentation, now partially indexed, permits the asking of fundamental questions about the nature and dynamics of eighteenthcentury successor states. We can ask: What constituted ‘conquest’ in the eighteenth century and how was it accomplished? How long did it take and how much were different elements of the indigenous society affec­ ted? What was the interface between conqueror and indigenous popula­ tion? We can also ask basic questions about the ideology of the Maratha Empire. What were the aims of government? How much continuity was there with the Mughal Empire, and how much innovation? The geographical framework in which these questions will be consi­ dered is the Malwa plateau. This region, whose northern rim is fifty miles south of Agra, is an oval tableland about one hundred and fifty miles by one hundred and twenty miles. It is bounded on the east and west by hills and on the south by the Vindhya mountains. Several features make it particularly interesting for the study of conquest. Malwa is a wellidentified ‘march area’; the one main road from the North (Agra, Delhi) to the South (the Deccan, Surat, Bombay) runs through the eastern portion of the plateau. It carried caravans and armies a thousand years before the Maratha conquest. Malwa’s population represents the gradual accretion of traders in towns along the road, cultivators attracted by the rich, black soil, and conquerors hoping to rule both. It is a complicated mix of Jats, Ahirs, Rajputs, tribals and some twenty other castes and communities. As a whole, the plateau is, therefore, without strong regional or religious identity. MUGHAL MALWA

At the opening of the eighteenth century, Malwa had been part of the heartland of the Mughal Empire since Akbar’s conquest in the 1560s. It had an administrative structure typical of areas firmly under control and valued for revenue production (some 10,000,000 rupees in 1707).

The Slow Conquest / 25 Overseeing the whole province was a subahdar, advised by senior administrative, religious, and police officials, all resident at a major town, such as Ujjain (in southwest Malwa). The plateau was further divided into ten sarkars and these were the functional centers of Mughal power. Some administrative decisions came from them and large bodies of troops were garrisoned in them. They were always strategically located. Mahatpur and Handia guarded the great Mughal trunk road from Agra to the South. Chanderi (in eastern Malwa) was a major trade and manufacturing center for silk. Mandu, Raisen, and Kanauj were fortresses guarding access from the South. Mandsaur (in western Malwa) flanked the major road to Rajasthan. Below the sarkars were three hundred and one parganas. Each, in theory, functioned as a revenue unit of some one hundred to two hundred villages and had a permanent Mughal revenue official and police office; in the absence of Mughal records, however, we cannot be certain of the pervasiveness of this low-level administration. Of the indigenous local elite, the zamindars, more is known.2In Malwa the vast majority of the zamindars under the Mughals were indigenous revenue officials with some local power. These zamindars gained their rights by three methods: colonization of fallow land, conquest, or grant by a previous empire or the Mughals. The vast majority were Rajputs. Historical evidence on them is extremely scarce, the main source being Muslim court chronicles. Still, U. N. Day, in Medieval Malwa, discusses pockets of immigrant Rajputs even before the founding of Dilawar Khan Ghuri’s kingdom in 1387.3 Such movement from Rajasthan was promoted by Ghuri’s successors, and the Rajputs were used for local control, though the methods are unclear. Exactly where they settled and their relations with the dominant Muslim state are, as yet, unanswered questions. All we can say is that the immigration continued through the fourteenth and fifteenth centuries. When the Mughal Empire was established in Malwa in 1560-80, several thousand were resident on the plateau. 2 My discussion of the Mughal use of the term zamindar, and his rights and duties, draws on the following recent research: N. A. Siddiqi, ‘The Classification of Villages under the Mughals’, The Indian Economic and Social History Review [I.E.S.H.R.], I, 3 (January* March, 1964), pp. 73-84; S. Nurul Hasan, ‘The Position of the Zamindars in the Mughal Empire’, I.E.S.H.R., 1,4 (April-June, 1964), pp. 107-19; B. L. Grover, ‘Nature of Dehat-iTaaluqa (Zamindari villages) and the evolution of the Taaluqdari system during the Mughal age’,I.E.S.H.R., n, 2 (April, 1965), pp. 166-77, and n, 3 (July, 1965), pp. 259-88; and N.A. Siddiqi, ‘Land revenue demand under the Mughals’, I.E.S.H.R., n, 4 (October, 1965), pp. 373-80. 3Upendranath Day, Medieval Malwa: A Political and Cultural History, 1401-1562 (Delhi, 1965), p. 22.

26 / Marathas, Marauders, and State Formation The Rajputs filled the position the Mughal Empire termed chaudhri, whose duties were to assist the pargana-level Imperial officials with settlement and collection. Each was held responsible by the state for revenue collection, and for law and order in his area (called a talukka). To fulfill these duties, as well as to provide local protection and subdue the local population, zamindars maintained a small fort and could muster one to two hundred irregular soldiers consisting of relatives and retainers. In return for these obligations, these zamindars were entitled to various sorts of emolument—rent-free lands, a percentage of the revenue, miscellaneous cesses—amounting to 5 to 10 per cent of the government’s share of the revenue. The rights were hereditary (though the Mughal Emperor asserted the right to regulate succession); and they could be sold, transferred, or mortgaged. On the lands the zamindar personally owned and cultivated (or rented out), he paid the government taxes, like any other cultivator. From the mid-sixteenth century onwards, the Mughal empire created many new zamindars where none earlier existed. Firstly, zamindar rights were granted to persons colonizing new land and persons who received maintenance grants (madad-i-maash) to fallow land. Secondly—and much more importantly—the Mughals continued the medieval Khilji Kingdom’s policy of settling Rajputs in non-Rajput areas, to diminish rebellions and ensure regular revenue collection. Jhabua state, in southwestern Malwa, is a good example of this process. In the first decades of the sixteenth century, two Rajput brothers, Jaswant Singh andRam Singh, were living on family lands near Ajmer (Rajasthan) which had been granted to their father. Ram Singh’s son enrolled in the Mughal army under Akbar and received the first, rather vague, grant in recently-conquered Malwa. Keshodas, Ram Singh’s grandson, quite successfully continued the tradition of Mughal service. Keshodas had, in 1572, been attached to the retinue of the young prince Salim, afterwards the Emperor Jahangir. He distinguished himself in the campaign in Bengal (1584) and was in recognition of his services granted five villages in Hindustan and ten districts in Malwa. After the accession of Jahangir (1605-28)) he was employed to subdue the turbulent free-booter Jhabbu Naik of Jhabua, Thana Naik of Thandla, and others who infested the south western districts of Malwa and especially Lakha Naik and Chandrabhan (Rajput) of Dhulet who had attacked and murdered the son of the Gujarat governor. Keshodas reduced these men to order and came into possession of their territory which included the districts of Jhabua, Thandla, Bhagor and Ramgarh.4 4Central India State Gazetteers,\ (Lucknow, 1907-9), pp. 518-19.

The Slow Conquest / 27 Thirty years of free-booting and anarchy followed Keshodas’ death in 1607, until the state was rebuilt by his grandson, again with the patronage of a Mughal ruler, this time Shah Jahan. Throughout the remainder of the sixteenth century, the family peacefully ruled their tribal subjects. Ratlam, another Rajput state (45 miles northeast of Jhabua) has a similar history, except that Mughal service started only in the time of Shah Jahan (c. 1630), half a century later than the Keshodas’s service under Akbar. The Mughal grant that started Ratlam state dates from about 1680.5Thus, Mughal patronage and Rajput settlement in western Malwa continued throughout the seventeenth century. Raghubir Singh, eminent historian of the Malwa region, asserts that the process of using Rajputs as a collaborating elite generally accelerated during Aurangzeb’s reign (1658-1707).6 Thus; at the opening of the eighteenth century, Malwa had a Rajput elite; some were long-standing residents in Malwa with centuries-old ties to certain areas. Others, though of long residence, had moved about in administrative positions and received estates under Akbar. Some received grants after Akbar, but many settled in Malwa only under Aurangzeb. These Rajputs were still culturally distinct, did not inter­ marry with the local population, and had tenuous ties to the peasants below them. For the Mughals, they played a crucially important role in indirect rule. They were a foreign, collaborating elite, responsible for revenue collection and law and order. Singh succinctly sums up their position: ‘The new set of Rajputs and their newly granted landed-estates supplied the necessary factor of stability and support to the Empire. They were the creations of the Mughal Empire, and as such very rarely stood in opposition to the Imperial authority.’7 Overall, then, the Mughals can be viewed as town-based and fortbased; their administrative and military apparatus was concentrated in ten strategic sarkar-garrisons around Malwa. They were, thus, well suited to static, positional warfare. Three features of this system are worth noting. Firstly, these sarkar towns, though well defended, represented substan­ tial concentrations of movable wealth, both in the government treasury and among the private traders who thrived under Mughal protection. Secondly, this structure was crucially dependent on roads for communi­ cation between foi^, collection of revenue, and supplying the garrisons. 5Ibid., pp. 315-23. 6Raghubir Singh, Malwa in Transition, or A Century of Anarchy: The First Phase, 16981765(Bombay, 1936),pp. 13-15.SeealsoK. R. Qanungo,S/ua

Banswara

• MANOSAUR SIIAMAU

I

Pipiauda • Sa,|ana

Narsinghgarh

| Tai • ■ • ^ I* Jacxa \ . Kachtod

SARUNGPUR

% • »Mahxjpur

Dhamoa I

• Raih

Shahiahanpgf Shujaipuf Ouiana

is'amnagar

y

RATLAM

bh ilsa

Raised

23“

22°

Oct., 1728 — — —

Approximate Plateau Border

2 ,. 74

75'

76

77°

76

Fig. 1: Places of Maratha Tribute Collection, 1729

79

32 / Marathas, Marauders, and State Formation military leaders, such as Bhikaji Shinde, Sultanji Shelke, Santaji More, and Udaji Pawar. Village names, pargana names and a few names of indi­ viduals were indiscriminately mixed in the papers, a practice atypical of later documents. This format suggests a simple listing of sources from which Maratha patrols succeeded in collecting tribute. Further features of the documents also suggest this transitoriness of collection. The total amount from nineteen parganas was quite small— slightly over Rs 500,000-since the area included major cities like Dhar, Indore, and Ujjain. Less than one-twentieth of the Mughal collection, the Maratha total was closer to ‘cash on hand’ than the much larger amount of a full year’s land revenue. Finally, internal evidence proves that the Marathas lacked any local-level administration of their own; for amounts not collected by military patrols, they were dependent on local collabo­ rators for much of the collection. Next to six items of debit was written the name. Nandlal, who was then Mughal Mandloi of Indore; several years earlier he had switched loyalties to the Marathas. In the first half of the 1730s, forays using military commanders as revenue collectors were increasingly successful and covered wider and wider areas, as shown in Figures 2 and 3, maps of places of actual collec­ tion of 1731 and 1732. The documents of these years show a thorough familiarity with the western half of Malwa; names are generally accurate and—for the first time—there were estimates of revenue based on the pre­ vious year’s collection. The random mixing of village and pargana names gave way to a regular ordering by parganas. The estimates were given to the nearest ten or hundred rupees, rather than the crude thousands of pre­ vious documents. The total collection of 1732 was over a million rupees, still only 10 per cent of the best Mughal collection from all of Malwa, but over twice the Maratha collection of 1729 and over 75 per cent of the estimated revenue.15 Let us pause now (say it is 1735), looking out over the ramparts of one of the few safe Mughal forts, and consider these Marathas and the kind of demands they were making. At the Mughal court, in Delhi, the agents of the Peshwa continued to assert claims to chauth (one quarter of the Mughal revenue) of Malwa. This ‘diplomatic offensive’ was an old one; it began in 1717 during routine negotiations between Shahu, the formal head of the Marathas and a Mughal representative. The Peshwa put forward the same claims to Malwa in 1719, and once again during the 13 Places of collection for 1731 are taken from a receipt Poona Daftar. P.D.H.R. no. 179. Information on 1732 is located in a group of accounts concerning the division of the actual collection, between the Peshwa and his generals. P.D.H.R. no. 154.

The Slow Conquest / 33 26°

JHANSi & IC H H A

25*

24°

Deoiia

Banswara

I—

• m andsaur / SUAMAU

Narsmghgarti

I 12 • •

Pitfauda

a

•

*

l>Jaora

Dhamoa

I

RATLAM

,Jd#£^

*

*

Du*,aha • ls,4mn? #'

• Suodarsi

•isanj. .Kjyah.

B H 1LSA

*

Shafyahanpur Sftupipur

\ . Kaehfod

• Ram

SAR LING PUR

•

g

u

• Rateen

..ehJS T V

23°

22 °

-

— — —

Approximate Plateau Border

2r 74 '

75 °

76°

77°

Fig. 2: Maratha Tribute Collection, 1731

78 °

79 °

34 / Marathas, Marauders, and State Formation 26“

NARWAfl JHANSI SHIVPURI OfiCHHA

Kotaras Budha Dongar

25°

Bhadaura ^ * . Kaiabag • ^HANDERI Shahdaura ^

Pacrxx

SIRONJ.—

24°

Deoiia

Banswara

I

P'Oiauda

SARUNGPUR ... »

I* Jaora \»


> C

CL P.K. 30. Rajamandal memorandum of warning, 1751.

Recovery from Adversity in Eighteenth-Century India 1 109 him to pay the revenue which had been fixed. He declared that he was unable to pay the revenue. The government fixed the settlement according to the year’s report of actual measurement of land under cultivation. Accordingly the revenue is fixed at Khandeshtaka 251.29 Other similar documentation gives the details that the pargana-level officials were directly involved in the report of the land actually remaining in cultivation, from which a new settlement was negotiated. One is perhaps worth quoting, in full: Letter of agreement. Jahankhan patil and Badghe patil and Sand and Yes Chaudhris of Kasbe Satatgav of pargana Adilabad in 1751 in writing gave a letter of agreement that the Peshwa has called us and ordered us to settle the revenue. We replied that we are unable to pay. Considering future prosperity [of cultivation], Gangapant, representative of the Deshmukh, appealed for a remis­ sion of the revenue and Hammat Khan agreed to stand surety along with the chogule [or chaudhri] for the revenue at ‘fixed’ Kandeshitakas [a monetary unit] 2-0-72, for the current year. For our share [of the revenue] if there is any alteration, we will be held as outlaws. So this letter of agreement is written.30

From other documents of the period, we can get a sense of the effects of this conflict. For example in Adilabad pargana, there were 4 deserted villages in 1748 when hostilities began; by 1751, 12 were deserted. Nearby Erandol pargana had 2 deserted villages out of 139 in 1748; by 1750 it had 6. Other parganas fared somewhat worse. In Sendume, the ‘patil and jemadars owing to the Mughal disturbances ran away. Thus, they were absent [at the time of the settlement]. At the end of the year, they were consulted and agreed to pay 7500 rupees.’31 Many deserted villages were also noted in Nasirabad, Nandurbar, and Ner parganas.32 Other parganas experienced more fighting and more severe disruption. Nimbait, for example, by 1753, had 23 of its 51 villages deserted and the whole of its remaining revenue taken up with military expenses.33Spora­ dic fighting continued through 1754 and 1755, especially around Malgav fort, still held by the Nizam’s troops. An estimate of 1755 conceded that the pargana was ‘badly disorganised’. Neighboring Dhulia pargana was much the same. Of 125 villages, 8 had been deserted in 1750; by 1753, 31 were deserted, and in 1755—after 29P.K. 128. Letter of agreement, 1751. ’op.K. 189. Taleband, 1753. 31P.K. 201. Jhadti of Prant Khandesh, 1751. 32Ibid. 53P.K. 187. Taleband, 1753.

110 / Marathas, Marauders, and State Formation nearly five years of fighting—there was still Mughal resistance. The Peshwa wrote to his newly appointed collector, as follows: If you do not do the settlement, the villages will be deserted. There is one Musad Khan Faruki who takes money from the people and uses it. Therefore, he should be punished. Otherwise, government takes the loss. People are not prepared to remain there. Government accepts the loss if more troops are required.34

Other areas fared even worse. The whole of Sarkar Baglan, in south­ west Khandesh, was looted by Damaji Gaekwad. In Pargana Songir, Damaji Gaekwad looted all the villages of this pargana and took money from them, so that all the villages became deserted. No taxes could be collected, therefore receipts = 0.35

Several sarkars between Khandesh and Malwa were particularly vulner­ able to attacks by Bhils, as these areas bordered ‘buffer zones’. These sarkars became completely deserted.36 The most interesting feature of this period of adversity was that it was reversed within a decade, most areas returning to full prosperity and cultivation. Before discussing these processes of recovery, let us consider the general features of the downward slide. In the 1750-5 period, adversity began with two specific, armed groups challenging Maratha authority. These were the remaining Nizam’s troops settled on estates aided by Mughal fort commanders and the BandeGaekwad faction who also held estates in Khandesh. Fighting began over attempts to confiscate these estates and continued with each seasonal revenue collection. For a time the local rebels were successful; they seized a substantial portion of available revenue, looted villages, and denied revenue to government. Officials became discouraged, unable to meet their contracts or risk giving out development loans. Villagers contracted for less and less cultivated area, retreating to more-protected fields. Information reaching the center became less frequent and less exact. Next, villagers began to flee the continued fighting and individual villages became deserted. Finally, whole parganas became deserted, some of them from depredations of local marauders such as Bhils, rather than the war itself. How, then, was the situation reversed? Probably foremost was defeat­ ing the Mughals and the rebels. This meant season after season putting MP.K. 217. Rajamandal, Appointment Letter, 1755. 33P.K. 201. Jhadti, 1751. 36Wad. Letter 57 and 58, p. 41.

Recovery from Adversity in Eighteenth-Century India /111 several thousand cavalry into the field, chasing them down, engaging them, and extracting treaties, hostages, and sureties. A central govern­ ment incapable of mounting this much force would simply have lost Khandesh to competing local forces. Meanwhile, government revenue officials went from village to village re-negotiating revenue contracts, and giving assurance that government demand would be lowered to reflect the conditions. Likewise, they re-negotiated terms with pargana officials who had fled. We also know of their coercive activities. They imprisoned village officials who supported the enemy or refused to acknowledge Maratha authority. The amount of village desertion varied greatly from area to area. In parganas where only a dozen or so villages were deserted, we can reasonably expect that the population was lodged with relatives in surrounding villages.37 Village exogamy, the common marriage pattern, would given an especially wide distribution to the wives’ networks. At any rate, there is no evidence that the village headman led the village away, or could call it back. More devastated regions required more government effort to repopulate them. Consider Baglan which was looted by Damaji Gaekwad. In 1754, the Peshwa’s daybook records the following request from his kamavisdar in Baglan: You have requested that the dams having given way, the garden crops do not get water in Baglan region and therefore requested permission to repair the dams. It is thereby ordered that for the purpose of repairing the dams Rs 5000 be paid yearly for 5 years from the revenues of the pargana.3®

Areas completely deserted required recruitment and repopulation, usually done by a special Maratha government official. The Peshwa wrote to one such official, as follows: the Sarkar Handia in Prant Nimar of the Khandesh District was, owing to the inroads of the Bheels, lying waste and desolate and was overgrown with forest. You were, therefore, formally ordered to take steps for populating the tract in question. This, by dint of personal exertion and enterprise, you have been able to effect. You have now come to Poona and requested that in consideration of your long services and your large family, a new watan (grant) may be given to you. Considering that you are an old servant of Government, capable of 37 Short-range migration would be in keeping with the general rule that people move as short a distance as possible. See Michael J. Greenwood, ‘Research on Internal Migration in the United States: A Survey’, Journal of Economic Literature, XIII, 2 (June 1975), p. 398. 3*Wad. Letter no. 391.

112 ! Marathas, Marauders, and State Formation performing any duty entrusted to you, and that you have hitherto served faithfully and are ready to do so in future. Government is pleased to bestow on you a new watan of Sarmandloi and sarkango of the above tract to the South of Rewa.39

Would that we had more details of this process, but it was certainly not the simple return of villagers to their ancestral fields. None of these repopulation documents mention the return to their original population, only favorable taxation terms to attract ‘settlers’. Crucial to repopulation was control of4he Bhils. Bhils (and tribals generally) have had a bad press, emphasizing only their raiding activities. The situation was, however, much more complicated. Bhils were the most vulnerable element in society; they lived on the most marginal land with the most variable water supply. In peaceful times with adequate monsoon, most tribals in Khandesh and Malwa were, from the point of view of production, agricultural peasants.40There was, however, always a section who lived by slash-and-bum agriculture or hunting-and-gathering in forests or hilly tracts. Elphinstone, in his famous report on areas con­ quered from the Peshwa well summarized this dual character: They are a wild, predatory tribe; and through they live quietly in the open country, they resume their character, whenever they are settled in a part that is strong, either from hills or jungle.41

Adversity on the plains meant that government was involved in fighting more serious rebels. Bhils from the hills, under their own chiefs, issued forth for raiding and probably recruited plains Bhils as conditions worsened. They cut roads, looted caravans, closed passes, and pillaged villages. It was up to any government to protect its villagers, and thereby its revenue, from these local marauders. The Mughals had tried with moderate success to settle minor Rajput lineages over them. Unfortu­ nately, when the Mughal government was weak, Bhils were convenient recruits into zamindar armies. Both Mughal and Maratha governments tried to insure the loyalty of Bhil families by granting haqqs or rights. Often these haqqs consisted of minor rent-free land or a certain amount of grain at harvest, plus clothing, blankets, or sandals. In return, the Bhil family agreed to provide some police duties if they lived in a village, protect passes if they lived in a forest, or live in outposts surrounding a 39Ibid., Letter no. 57. 40See K. Suresh Singh. 'State Formation in Tribal Societies: Some Preliminary Observa­ tions’, Journal of Indian Anthropology, 6 (1971), pp. 161-81. 41Mountstuart Elphinstone, Report on the Territories Conquered From the Peishwa (Calcutta, 1821), p. 2.

Recovery from Adversity in Eighteenth-Century India / 113 fort and warn of attack.42 The Marathas also tried to gain the loyalty of Bhil chiefs by grants, which as Malcolm explained in his report on Central India, ‘assign land at a fixed moderate assessment (sometimes far below revenue), [they] are seldom given but to Rajpoot lords or Bheel chiefs whom it is desired to conciliate and induce to cultivate the soil, or who engage to give protection to certain limits, on condition of this favour.’43 Overall, the Marathas seem to have relied heavily on military patrols to defeat or kill those Bhils who raided villages. In western Khandesh, there are regular notices of these patrols, with Bhils captured, tried, and executed throughout the 1760s and 1770s. The Marathas regularly collected a special tax, called Bhilpati, from those areas threatened by Bhil raids to pay for additional troops.44 The point of this little excursion into Bhil country is to emphasize that if a whole region became depopulated, even for a few years, it was no longer simply a matter of the villagers coming back. There were larger political, economic, and military reasons why the village was abandoned in the first place, and only active political and military intervention by government could reverse the situation. This, as we have seen, required I) defeating the major rebels; 2) a concerted, successful policy—whether conciliatory or punitive—against local marauders; 3) appointment of officials whose job it was to recruit settlers and oversee repopulation, with suitable rewards for success; 4) negotiations with potential settlers, resulting in favorable tax terms over a period of up to a decade; 5) re­ building of some infrastructure, such as dams. In addition, areas primarily in cash crops such as Malwa and Khandesh cotton depended on government’s insuring the overall safety of the roads, so that caravans could transport the crop to distant markets. All these activities were fairly successfully done by the Marathas in the 1750s and 60s; yet the results were all swept away in a major cataclysm at the end of the century, to which we now turn. A BRIEF LOOK AT A MAJOR UPHEAVAL, 1795-1820

Elphinstone’s Report, compiled shortly after the British conquered Maharastra and Central India, well conveys what they found in Khandesh: Some parts of the province are still in a high state of cultivation, and others, more 42R. N. Saletore, ‘The Bhils of Maharastra’, New Indian Antiquary, 1,5 (August 1938), pp. 322-36). 43Sir John Malcolm, A Memoir of Central India (London, 1832), I, p. 57. "P.K., 198. Account of Bhilpati, 1747.

\ \ A t Marathas, Marauders, and State Formation recently abandoned, convey a high notion of their former richness and prosper­ ity; but the greater part of Candeish is covered with thick jungle, full of tygers and other wild beasts, but scattered with the ruins of former villages. The districts north of the Tapty in particular, which were formerly very populous, and yielded a large revenue, are now almost uninhabited forest.45

Elphinstone put this destruction in three stages, beginning in 1802, ‘when it was ravaged by Holcar’s army’, followed by ‘the famine of 1803’, furthered ‘by the misgovemment of the Peshwa’s officers’, and com­ pleted by Bhils who withdrew to the mountains and raided the remaining villages. Many of these elements are familiar from the adversity of 1750-5, i.e., conflict over paramountcy leading to warfare across the province, then a drought, then local marauders—especially Bhils—raiding villages from nearby hills. There were, nevertheless, important differences, which resulted in recovery in the 1750s and destruction in the early 1800s. The first difference, and probably the most important, was the absolute decline in strength of the Maratha central government at Poona. With more land alienated to its chiefs, and, therefore, less revenue, it simply could not hire the troops to crush a revolt as it had in the 1750s. Holkar’s army ravaged and looted with little opposition. For villages, this meant not only sporadic military patrols, but a loss of faith in a government incapable of responding to revolt. The second difference was in the nature of Maratha armies, especially the addition of a large irregular following. It is now generally agreed that even in the early eighteen century, Maratha armies had small bodies of irregular, unpaid troops known as pindaris who were used mainly for looting the villages and camp of the enemy. The pindaris emerged in large numbers in the succession disputes that followed the death of three major Maratha leaders between 1795 and 1800. Armies were more and more irregularly paid; looting became more common and compensation paid to the looted village rarer. Each army entertained more pindaris who agreed to maintain themselves by looting the ‘enemy’. The downward spiral had begun. Soon, the pindaris, under their own chiefs, made deals with local landed elites, especially in the hills north of the Narmada River. The local noble known as a zamindar protected the families of the pindari chief and his followers, in return for a share of the booty and an agreement not to molest his lands.46 Warfare at its most destructive, that is against 45Elphinstone’s Report, p. 3. 46(an officer in the East India Company), Origin of the Pindaris (London, 1818), p. 97.

Recovery from Adversity in Eighteenth-Century India! 115 cultivating villages, at least passes as the front line shifts; systematized looting is perhaps the most destructive train of events than can happen to '

an agricultural society. They use sword and lance, and pillage without distinction. They move in bodies seldom exceeding two or three thousand men, and hold direct undeviating course until they reach their destination, when they at once divide into smaller parties, that they may with more facility plunder the country, and carry off a larger quantity of booty; destroying, at the same time, what they cannot remove.47

Without strong government intervention, this zamindar protected looting was an inevitable downward cycle. His fields are laid to waste, his cottage reduced to ashes and he has no alternative, but that of joining the standard of some lawless chief. Thus, the number of the Pindaries may be said to increase in the same ratio, as the means of subsistence diminish.

Malcolm found village headmen who with a few followers had begun plundering neighboring villages, using the ruins of their village as a base or hiding in the steadily-spreading forest. Besides becoming local marauders, other alternatives were, in fact, available to villagers. The first was to go under the protection of a local, armed zamindar. This meant moving family and stock to an area immediately surrounding the fort, cultivating smaller fields, and giving a percentage to the zamindar in return for protection. This was apparently a common alternative; by 1820, within the approximately 3000 square miles of Khandesh, Briggs found over 400 forts, large and small, with populations living inside or under the walls.48Malcolm, at the same time, found that many others had moved hundreds of miles to find a safe place to cultivate, but that ‘a great majority went to large towns, where they found a temporary asylum, and obtained subsistence by labouring in gardens or fields.’49Unfortunately, in the absence of census data from this period, we cannot check the actual frequency of these various alterna­

tives. CONSEQUENCES OF LONG-TERM ADVERSITY, AND SOME CONCLUSIONS

Adversity of this scale and intensity had profound long-term effects, fortunately well-documented after the British conquest of the area in 47Ibid., p. 96. Poona Daftar. Deccan Commission Files, vol. 171, Document no. 136. 49Malcolm, pp. 20-2.

116/ Marathas, Marauders, and State Formation 1818. First, there were the effects on the land. Villages, once deserted, rapidly disappeared. Scrub weeds were quickly replaced by tough, thorny bushes and trees. More importantly, when irrigation works went without repair, drairtage was altered and large areas north of the Tapti became malarial. Even decades after the British conquest, troops could not be quartered in those areas through rainy seasons. Predatory animals spread from the hilly regions onto the plains and river valleys. In two decades, cultivators—whether returning or new colonizers—generally faced con­ ditions of clearing heavy growth and starting afresh. The effects were equally profound on the people. Malcolm found three classes of cultivators in Central India during his surveys of the 1820s. Two are familiar to us from land revenue studies: ‘Junna Kursa’, who were the old or native cultivators of their own lands, parallel to mirasi tenure in Maharastra, and Taeekustee’ who were contract tenants with agreements of from two to five years, parallel to istimari tenure in Maharastra.50The third, ‘sookwasee’, were a product of adversity: These are cultivating labourers who settle, as the name implies, for one, two, three, or more years, where they expect to be best treated. They are usually men who have been driven from their homes by war, pestilence, tyranny of nilers, quarrels with their relations, or some serious misfortune. They have no immu­ nities or rights, and are much at the mercy of those by whom they are employed. It is a melancholy comment upon the past condition of Central India, that a great proportion of its husbandman are of this class.11

Nevertheless, for those still on the land, there were opportunities. Hie British government made hundreds of settlements with local powers (occupying the hundreds of forts in Khandesh and Malwa) in which land and revenue rights were recognized in return for pacifying the area: The next few years (after 1817) were spent in settling the country and repopulating villages. One of the principal means of achieving this was by granting a guarantee to small holders that their holding would be assured to them, on the understand­ ing that they assisted in pacifying the districts in which they lived. This guarantee, which secured the small thakurs from absorption by the great Darbars, acted like magic in assisting to produce order.32

It sounds so simple and obvious in this gazetteer report from the end of 50Here, I am following S. N. Sen, Administrative System of the Marathas (Calcutta, 1925), and Hiroshi Fukazawa. ‘Land and Peasants in the Eighteenth Century Maratha Kingdom’. Histosubashi Journal of Economics (June 1965), pp. 32-61. 11Malcolm, p. 27. n The Imperial Gazeteer of India, IX. New Series (Oxford, 1908), pp. 342-3.

Recovery from Adversity in Eighteenth-Century India /117 the century; yet, at the time, a new rural aristocracy was produced overnight, some of its elements of the old zamindari families, others much more recent. There were other opportunities, given vacant fertile land, even if it was covered with trees and infested by tigers. Elphinstone, in his 1820 report, suggests these possibilities. ‘A very light assessment, and the favourable terms on which waste land is granted to speculators will, it is hoped .. .not only draw back the Natives of Candiesh who have retired to Guzerat and other Countries, but even attract new settlers, from places where the population is overabundant.’53 Even with adequate tax incentives, Elphinstone is quite correct in using the term ‘speculator’ for the processpf re-populating Khandesh; it proved a risky business, at best. First, there were the politics of migration. Every ruler was aware that people and their production were much more important than land, as such, and were understandably reluctant to let Khandesh natives return. Thus, Briggs, the Khandesh commissioner in the early 1820s, received frequent applications from both individuals and village officials who were prepared to return home but were afraid of being seized and plundered if they attempted to leave the Nizam’s or Holkar’s territories. Briggs recognized the delicacy of the situation: politically, the British could not afford to anger their new native allies in Gujerat, Malwa, and Berar; not could they impoverish them and threaten their stability. Therefore, the British put no pressure on surrounding native states to allow return migration into British Khandesh.34 In India, there is no indication that peasant emigration and return has ever been anything but a complicated business. In the mid eighteenth century, for example, the Maratha general, Shinde, wrote to his officer that the Raja of Kotah (in North West Malwa) ‘has approached us for recovery of subjects who ran away to Patan. You should allow them to return only if they want to. They should be allowed to stay wherever they are happy 55 Shinde knew productive peasants and potential revenue when he saw them! In addition to these sorts of negotiations, there were negotiations with the cultivators themselves. From this distance, we can see only some of the issues, specifically those in which government was involved. Aside 53Elphinstone’s Report, p. 4. 54Poona Daftar. Deccan Commission Files no. 181, Letter 817, from Briggs to Bombay Government 55Sardar Anad Rao Phalke, ed. Sindeshahi Itihasachi Sadhne (Documents of the History of the Sindhias), 1(Gwalior, 1929), p. 32, Letter 33A.

118/ Marathas, Marauders, and State Formation from revenue arrangements and development loans, there were legal issues. Generally, throughout Maharastra, a Thurlkuree (or original cultivator) had thirty years in which to return and reclaim his lands.36 Perhaps this is the reason that so few of the new villages were built on the exact site of older villages.57On the other hand, it is well to remember that nowhere are ‘traditional’ rights established more quickly than in India. For example, in 1835, William Sykes (then Statistical Reporter to the Government of Bombay) made a tour of the northern districts of Maharastra in search of mirasi estates, the ‘ancient’ holders of land. He found that while oral tradition assigns estates to an old Maratha family, most cultivators denied the existence of the estates. In one village near Ahmadnagar, Sykes managed to browbeat the local officials into produc­ ing a document from 1777, showing the mirasi rights. ‘In 1827, there was not a single person alive, a descendant from the mirasi rights in 1777.’ Even the headman was new. Yet, and this was true of other villages Sykes examined, this did not stop any of the recent settlers from successfully claiming mirasi rights as the ancient and original tillers of the soil. An in­ credulous Sykes found that in the revenue records, ‘individuals have had Thai (mirasi estates) named after them . . . and the village authorities seem to think of the present cultivators are [sic] established on hereditary rights.’58After a period of adversity and depopulation, Sykes was watch­ ing ‘tradition’ in the making. Even if legal problems and ‘tradition’ could be circumvented, success­ ful repopulation depended in other ways on government intervention. Briggs, as the first Commissioner in Khandesh, received frequent re­ quests for grants to repair irrigation works. These were considered too expensive and routinely rejected: In many instances however I met with the most importunate petitions for repairs of the numerous fine dams in Candeish which will at some future period so naturally tend to the increase of the land revenue, but I invariably told them it must depend on the expected increase of revenue, desirable from those repairs whether or not Govt would undertake to sink so much money, and I called on each District to furnish me a list explanatory of those objects.59

This thrifty attitude contrasts sharply with the outright grants for re56Elphinston’s Report, Vn. 37Poona Daftar. Revenue File no. 3. Wingate’s Report, pp. 83-4. 38William F. Sykes, ‘On the Land Tenures of the Dekkan’, Journal of the Royal Asiatic Society, n (1835), pp. 210-11. 39Deccan Commission Hie no. 176. Letter no. 539. Briggs Report on the Revenue System of the Former Government, p. 337.

Recovery from Adversity in Eighteenth-Century India / 119 building dams given by the Marathas after the adversity of 1750-5 and probably made it that much more difficult to re-populate many parganas. In another crucial area, dealing with local marauders, the fledgling British government did rather better. They had little success with military patrols and even less with settling Bhils and others on guaranteed plots of land. Within a few years, however, the Bombay government developed the strategy of hiring some Bhils to police the rest. I had been tried before, but never with uniforms, discipline, and regular pay. Generally, it worked. The Bhil Corps gave regular employment to a significant segment of the warrior community; the rest settled into either huntingand-gathering in the forests or cultivation on the plains. The security thereby produced would have been a definite plus in repopulation attempts. Still, the whole process of economic recovery went slowly, very slowly. In 1853, during attempts to survey Khandesh. Wingate first described the river valley as having ‘the aspect of a newly peopled colony if we except those traces of a former industry to be seen in the mango and tamarind trees and the many ruined wells, which are still to be met with in the neighbourhood of almost every village.’ For the rest, ‘the whole of Khandesh may be considered as only very partially reclaimed from a state of nature’.60 In this slowly growing economy, a social structure emerged that was substantially different from pre-adversity times. In each pargana were three or four very large landholding families, invariably district or village officials. Below them were a small number of intermediate independent holders, and a great number of poor farmers, for whom the big families stood surety for revenue. It is tempting to consider the big families the heirs of successful ‘speculators’ in the business of land re-population.61 Several conclusions about villages and adversity in pre-modern India appear from this study. First, let us recall the model of village action described by Metcalf. As adversity worsened, there was always an appropriate village response. Villagers fought or fled or—as conditions declined—migrated a short distance, then perhaps a longer distance, always under village leadership and always anticipating return to ances­ tral lands as soon as conditions got better. In fact, in Central India, in the eighteenth and early nineteenth centuries, we found several kinds of “ Revenue File no. 3. Wingate’s report, pp. 83-4. 41J. F. M. Jhirad, ‘The Khandesh Survey Riots of 1852: Government Policy and Rural Society in Western India’, Journal of the Royal Asiatic Society, Parts 3 and 4 (1968), pp. 151-65.

120 / Marathas, Marauders, and State Formation adversity. Only short-term adversity embedded in relatively good times fits Metcalfs model, with the village fighting, paying, or fleeing as a kind of unit, resisting forces from the outside. On the other hand, where adversity went on for more than one or two years, Metcalfs model fails. There is no evidence that the village ‘moved o ff, to return whenever possible. As we have seen, the prbcess of repopulating deserted villages was a difficult one, highly dependent on government action, not village initiative. Government either appointed special officials to find people, settle them, and protect them from local marauders (as under the Marathas) or entertained speculators who did much the same thing (as under the British). The work of repopulation was complicated by the economic value of the cultivators in the areas to which they had emigrated; neighboring states resisted their return. Repopulation was impossible until government decisively defeated both competing powers and local marauders; it was greatly aided by positive government actions, such as development loans and rebuilding of dams. Second, the Metcalf model underplays the effects of long-term adver­ sity. The ecological effects alone were profound. With irrigation works destroyed, whole areas could become malarial, as they did in northern Khandesh. Tigers and other predators threatened the new settlements. The social effects could be even more profound. The process of a new power negotiating settlements with whomever remained on the land could, almost overnight, create a new rural elite with rights to land and revenue. Slower, but equally important, was the necessary process of using speculators or officials to repopulate areas; they, too, got rights and large estates in the process and joined the new rural elite. Destruction of records and a new population aided the claims of new cultivators to be ‘original’ holders, with all the rights that implied. Often these could be established within a generation. The results for Malwa were the negoti­ ated existence of dozens and dozens of small protected states and zamindaries. The picture in Khandesh is less clear, though the results of prolonged adversity appear to have enhanced the position of a few large landholders. In any case, redevelopment proceeded very slowly, requir­ ing nearly a century to reach the productivity of Mughal and Maratha times. Finally, let me offer a couple of more speculative suggestions that arise from the research thus far. First, banditry and migration should no longer be treated as separate issues in systems in which adversity was a regular occurrence. It seems that both could be a response to adversity, and the deciding factor was ecology. When adversity began, those, like Bhils,

Recovery from Adversity in Eighteenth-Century India! 121 living in marginal agricultural areas and adjoining hills seem to have had few options but to retire to the hills and raid the plains. Less marginal cultivators had more options, like negotiating with government for tax relief; if adversity worsened, they still seem to have had more options, such as emigrating to a town, going under the protection of zamindar, or long-range migrating, besides becoming local marauders. Perhaps these cultivators from better land had not only more resources with which to move, but more knowledge and a wider network of relations. Second, it seems important to view villagers as negotiators in the political system of pre-modern kingdoms, not mere passive observers. In good times, they negotiated the year’s revenue and area to be cultivated. They appealed for remission during short-term adversity. When the situation worsened, they ‘voted with their feet’ and left for better conditions. They listened to the offers of an official or a speculator trying to repopulate vacant land and judged the advantage of opening new land against the risks of leaving the current land. Perhaps this concern with villagers as negotiators, even if possessing limited resources and options, will focus our attention on contracts as an important feature of life in pre­ modem kingdoms. With the recent emphasis on caste, myth, arid kinship, the contract side has been largely ignored, and, thereby, we risk a one­ sided understanding of the dynamics of these kingdoms.

Chapter Six

Kinship and Pargana in Eighteenth-Century Khandesh Introduction

The ‘texture’ or composition of pre-colonial rural India remains a question central not only to historical research, but to modem policy planning. Throughout this century, for example, planners assumed that the village is—and always has been—the significant social unit in the countryside. Hence, the development of village panchayats, village development schemes, and village-oriented rural credit. More recently, pan-village units, such as block areas, have received attention and funding. They have, however, often been criticized as artificial govern­ ment impositions, with no local roots. It is in this context that attention is now focused on parganas, a very old form of supra-village local rural organization. The central question, then, is to what extent India, in the seventeenth and eighteenth centuries, was organized into small supra-village units. How did these units arise and what was their function to the central government, to an elite—if anywithin the unit, and to cultivating peasants? The answers to these questions, we believe, are crucial to our ‘mental map’ of pre-colonial India. Let us begin with a brief history of the term pargana. HISTORY OF THE TERM PARGANA

Parganas—local, contiguous groupings of ten to 150 villages—were a ubiquitous and long-lived feature of the Mughal Empire. The term is, however, much older than the Empire. Of Persian origin, the original meaning was only region or district. It was introduced into India possibly as early as 1300 in the north, reaching the Deccan with the Tugluqs in the fourteenth century.1 The term pargana is common in the Adilshahi records of southern Deccan in the fifteenth and sixteenth centuries.2 1One early reference is in the Tarikh-i-FiruzShahi, dated around 1350. H. M. Elliot and John Dowson, History of India as Told by its Own Historians, vol. 3 (New Delhi, 1964). Reprinted from Indian Economic and Social History Review, 22,4 (1985). 2Hiroshi Fukuzawa, ‘A Study of the Local Administration of Adilshahi Sultanate (AD 1489-1686)’, Hitotsubashi Journal of Economics, June 1963, p. 41.

Kinship and Pargana in Eighteenth-Century Khandesh / 123 Parganas have long been thought of as simply tappas, a much older Hindu term for a grouping of villages, renamed by the Muslim con­ querors. This view began with the earliest British studies of Maratha and Mughal administration, even pre-dating the settlement reports of the 1820s. Typical is Mackenzie’s preface to the 1809 translation of the ‘Hakikut Hindustan’, an eighteenth century Mughal revenue register:3 Thus we find a Soobah or province formed of what constituted a Dasum in the ancient registers and sometimes Naads comprehending the same purgunnahs or subdivisions under the name of mahals. . . an arrangement which it is appre­ hended originated in the ancient system peculiar to India, of settling a country by little communities in groups of villages instead of the European mode of subdividing the land into estates and farms occupied by single families___

By 1830, a consensus had emerged among the British administrators. Parganas were seen as ‘ancient’, probably pre-Muslim, and somehow the geographical expression of ‘little communities’. Elaborating this theme of ‘little communities’, research in this century on parganas has moved from a static, structural view to a more dynamic one. In the past two decades, the focus has been on the development of parganas. The most articulated of these theories is that of Richard Fox in Kin, Clan, Raja, and Rule.* In examining Oudh, Fox found, first, a long period of infiltration, colonization, and conquest of local areas by immigrant Rajput groups. The kinship of the Rajput clan defined a local area and became coterminous with a pargana. Thereafter, the boundaries of the pargana depended on the strength of the clan vis-a-vis the central government. It was only if the central government became unusually strong and the clan usually weak that others (grantees or bureaucrats of the central government) might usurp the lands and rights of the Rajput lineage of a pargana. Though based on eastern Oudh, these recent studies have oudined the organization of local parganas throughout the Mughal Empire. Before examining the specifics of our data and analysis, it is important that we have a clear picture of this generalized model of a pargana, whose main features were as follows: 1. Parganas were of ancient origin. 3India Office Library, Mackenzie Collection, General vol. 44, ‘Hakeekut Hindustan of Letchmee Narain’, trans. supervised by C. Mackenzie, 1809, p. 2. 4Richard Fox, Kin, Clan, Raja, and Rule (Berkeley, 1971). Fox’s work follows on Bernard Cohn’s two articles on the Benares area, as follows: ‘The Initial British Impact on India: A Case Study of the Benares Region’, Journal of Asian Studies, vol. 19, no. 4, August 1960, pp. 418-31; ‘Political Systems of Eighteenth Century India: The Benares Region’, Journal of the American Oriental Society, vol. 82, no. 3, July-September 1962, pp. 312-20.

124/ Marathas, Marauders, and State Formation 2. The villages represented the area controlled by a single clan/ lineage, generally Rajput. 3. The head of the clan occupied the chief position in the pargana and negotiated pargana affairs with whatever central government stood above them. Thus, parganas were the ‘building blocks’ of empire. 4. Parganas can be seen as ‘segments’, the physical manifestation of a model of Rajput kinship on the local terrain. 5. The head of the pargana held court and lived in a fortified town, the largest town in the pargana. His kinsmen attended court. 6. The head of the clan could call on his kinsmen to assemble a local military force, the largest and dominant force of the pargana except for the rare intervention of central government power. 7. Ritual flowed outward from the main pargana town, and was patronized and controlled by the clan. 8. Pargana politics were clan politics; conflict followed fissures in the clan and the jockeying of various lines and relatives. 9. The central government, through alliance and bureaucratic pres­ sure, could strengthen the head of the clan, or undercut his power. 10. During decline of crises, the state would dissolve into these pargana/lineage units.

The most important feature of this model is, of course, that lineage was the only basis for pargana organization. This assumption overlooks several questions crucial to our understanding of parganas and the local texture of pre-modem India. What was the importance of the pargana as an administrative unit, its importance to the state, rather than to its inhabitants? How much of the pargana infrastructure—human, physical, administrative—survived from one empire to the next? Was it, indeed, possible to build an empire by mobilizing heads of parganas? Can we find parganas that were not clan-based, and, if so, why? Can we trace largescale trends of the eighteenth century, such as monetization, at the par­ gana level? ADILABAD, ON PAPER AND ON THE GROUND

Two unusual, perhaps unique, sources of documentation allow these questions to be answered for a single pargana through two empires, the Equally important for his model is the following seminal article: Khashi N. Singh, ‘The Territorial Base of Medieval Town and Village Settlement in Eastern Uttar Pradesh, India’, Annals of the Association of American Geographers, 58,1968, pp. 203-20. Irfan Habib earlier set out a clear association of Rajput and Jat clans with local territorial divisions within the Mughal Empire in North India, See Irfan Habib, The Agrarian System of Mughal India (Aligarh, 1963), pp. 161-9.

Kinship and Pargana in Eighteenth-Century Khandesh / 125 Mughal and the Maratha. They both yield detailed data on Adilabad pargana, a region in the Tapti river valley, in what is now the northeast comer of Maharashtra state. On the Mughal side, the basis of this discussion is a detailed village by village report prepared by the Diwan of Khandesh province for the Mughal diwan of the six Deccan provinces at Aurangabad in a d 1699. This sixty-three page report is part of the Inayat Jang Collection, National Archives of India. As we shall see, the report has a narrow administrative focus, but excellent fme-grid detail on revenue, irrigation, and cultivated land.5 On the Maratha side, no one document contains the scope, detail, and organization of the diwan’s report. Rather, there are the day-to-day and year-to-year documents of the central administration from the first revenue collections in the 1720s through total Maratha control, in 1760. Housed in the Pune Daftar and part of the Peshwa Khandesh Collection, these include letters from the Peshwa to his agents in the field, appeals from local people, yearly and interim accounts, surveys, and instructions. The record is frustrating and fragmentary, but often gives remarkable glimpses into the ‘on-the-ground’ workings of the pargana. We have supplemented these two main sources with the British reports of Khandesh in the early 1820s, written soon after the conquest of the Deccan. Particularly useful, we shall see, is John Brigg’s map, compiled in 1821, before the parganas were consolidated into large administrative units. Adilabad is today a small rural town located twenty-three miles to the southwest of Burhanpur city in Khandesh district, in the state of Maharashtra. At the turn of the seventeenth century it was a Mughal petty administrative centre (qasbah) in Asir district, Khandesh province. The name of the town, spelled in Arabic script in the Mughal documents, is clearly Muslim in origin. Adilabad can be translated as the ‘Abode of Justice’ or even ambiguously as the ‘Abode of Idol Worshippers’.6 With the aid of the map, let us briefly tour the pargana, on horseback, as Mughal and Maratha officials did every season.7 5Documents from the Inayat Jang Collection housed in the National Archives of India, New Delhi, 111/433-92,569,571. ‘Detailed ofpargana Adilabad, district Burhanpur, suba Khandesh in accordance with thefasli year 1104.’ The 63 page report is stamped with the iron seal of Abdullah dated AH 1099 (his year of appointment). The seal is inscribed adam tabdil, indicating that the copies so stamped are copied without modification or alteration. 6Adil alone means justice or equity. Another meaning offered by Steingass’s PersianEnglish Dictionary is that of‘One who deviates;. . . one who gives partners to God, an idolworshipper’ (p. 829). 7This map was constructed from die diwan’s report, a Maratha village list of 1756, and the following early British map. India Office Library, ‘Trigonometrical Survey of the Province

126 / Marathas, Marauders, and State Formation

J E S / P A S 1 966

Kinship and Pargana in Eighteenth-Century Khandesh / 127 Starting west from the walled town of Adilabad, on the banks of the Puma River, we would ride through a broad cultivated river valley, with villages (like Dudhale) partly irrigated by wells. Typical crops in the irrigated fields might be sugarcane, opium, or fruit trees. The unirrigated fields would have, for example, wheat, channa dal, or jowar. Soon we would be near the Tapti, but we would probably ride well back from the river itself. As streams approach the Tapti, they get deeper and deeper, eventually joining the river, which is over 100 feet below the plain.8To the west, the pargana narrows to include only the river-irrigated villages along the south bank of the Tapti, for a distance of more than thirty miles. Another tour proceeds south from Adilabad town through the essen­ tially flat Purna River valley past villages like Hartala and Rulkhed to a tributary of the Purna which formed the southern boundary of the pargana. We might follow it downstream to its junction with the Puma. There we would find Bodyad, a substantial market town. The north side of the Puma was a more sparsely populated plain, cut by streams coming from the forested hills arising only five miles north of Bodvad. These steep hills divide the watershed of the Puma and the Tapti and, considering they were the home of unsubdued tribal Bhils, we would probably not cross them unless absolutely necessary. The final tour of the pargana would be north and east along the plateau overlooking the Tapti. We would find villages regularly, every couple of miles, all partly irrigated from the Tapti or tributary streams. We would, this whole tour, be following the old and well-travelled road from Aurangabad to Burhanpur and would see caravans, grain-carriers, and pilgrims. Returning from our tour, what sort of a place was Adilabad town? TAXATION AND SETTLEMENT, 1996 Adilabad town was the last major stopping point for travellers moving

north from Aurangabad to Burhanpur and thence north to Agra along one variant of the great Mughal north-south routes.9 In ad 1696 the town lay of Candeish, divided into Purgunas, completed in 1822, by Messrs. Arthur White and Jas Evers, under the direction and superintendence of Captain John Briggs, Political Agent. ’ The scale of this map is 4.125 mi. to the inch. 'Pune Daftar. Revenue File no. 3, ‘Wingate Report’, 1853, pp. 78-9. 9 Adilabad appears at 21’ N 76’ E in ‘The Deccan (West) Political 1707’ Map 14 A and ‘The Deccan (West) Economic’ Map 14 B in Irfan Habib, An Atlas of the Mughal Empire (Delhi, 1982). The route between Aurangabad and Burhanpur appears on the economic map.

128 / Marathas, Marauders, and State Formation in the jurisdiction of Muhammad Afzal, the chief local official (faujdar). It was probably his residence and headquarters as well. Local officials included four town headmen: Ratan Patel, Issa (?) Chaudhri, Suba (?) Chaudhri, and Tija Chaudhri. The local accountant (patwari) was named as Narainji. The town’s official revenue figure (jama) was set at 648,357 Mughal dams (16,209 rupees). The more meaningful ‘established collec­ tions’ figure (muqarrar hasil) was 8527 rupees.10Receipts (wasuli) at the time of the report totaled 4620 rupees with 3906 rupees in arrears (titimmah). In other words 54 per cent of the anticipated revenues had been collected. These figures refer only to the land taxes imposed on the measured agricultural area (raqaba) of the town. The latter consisted of 3111 acres of sown land (mazrui) and 2477 acres of waste or grazing land (uftada). The total also included 470 acres allotted to tax-free grants (aimma). Areas in this category were probably cultivated. The revenue burden on 3111 acres of cultivated land subject to taxation in 1696 was 8527 rupees or 2.74 silver rupees per acre. Much, if not all, of this land was under irrigation. The water came either from wells or from the streams flowing

from the Puma River according to the report." It is interesting that these and only these data were deemed essential information by the imperial administration at Burhanpur and Aurangabad. Location, identity of the jagirdar and local officials, and revenue assess­ ments and collections were obviously significant. The surveyed agricul­ tural area of the town was recorded as well as the source of irrigation for the fields. The amount of land allotted to holders of tax-free grants was listed. We do not find population figures, the number of households, or data on crops cultivated, markets held and so forth. The Mughal system of taxation did not consider population as a variable, but land and its produce. Looking outward from Adilabad town, the document allows us a limited, economic picture of the whole pargana. Late-seventeenthcentury Adilabad pargana consisted of 135 towns and villages (muwaza). Of these 124 were classified as ‘original’ (asli) and eleven ‘additional’ (dakhili) villages.12 The latter were relatively newly settled villages whose lands and revenues were treated by the imperial administration as 10The following information is taken from the sheet summarising Adilabad town, p. 2 of the Mughal report, 1/11/491. “ VI 1/492. 11See Irfan Habib, The Agrarian System c f Mughal India, op. cit., 252 n., for a definition of dakhili villages.

Kinship and Pargana in Eighteenth-Century Khandesh / 129 part of one of the older villages. The two settlements together formed a composite revenue unit. Thus, for imperial purposes at least, only 124 villages were listed in series in the documents under consideration. From the viewpoint of the state, within these boundaries Adilabad pargana constituted a distinct economic unit. The principal determinant of prosperity and production was cultivated land. Adilabad fell within the zone of regulation (zabt). This meant that village lands had been subject to cadastral survey since the mid-seventeenth-century administrative reforms in the Deccan. Under this system of measurement, survey parties (raqaba) located and measured cultivated fields and fallow lands. The data thus obtained were used in conjunction with market prices of produce to fix the assessment (jama). Let us consider the overall figures from this survey, as follows:13

Bighas (Acres) (per cent)

Sown Area (mazrui)

Culturable Fallow and Grazing (Uftada)

Grants (Aimma)

Total

149,322 (111,999) 56.9

98,699 (74,024) 37.5

14,643 (10,982) 5.6

262,674 (197,005) 100.0

796.0 (597).

118.1 (88.6)

2118.3 (1588.&)

Average per Village Bighas (Acres)

1204.3 (903.2)

13 The measured area figures have been converted to bighas from the units given in the documents. The original listing gives each area in the customary Khandesh units: 1aut=20 partan, 1partan=4 bighas, 1bigha=20 biswah (or pand). The survey bigha eventually used by the British in Khandesh district was derived from an average of the values for the extant Khandesh bigha. This equivalence was setat 10.33 bighas to 1acre or 1bigha equalled 3,739 square yards (0.75 of an acre). James M. Campbell, Khandesh District Gazetteer, 1880, p. 205. See also H. H. Wilson, A Glossary of Judicial and Revenue Terms (London, 1868), 1968 reprint edition for definitions of these units. The term partan was confined to Kunbi cultivators of the nineteenth century and has since died out in modem usage. The area totals differ slightly from the summary figures given in the report. We have gone through and converted all the numbers on a village by village basis and our own calculations. Thus the total given in the summary is 261,839 bighas instead of 262,674 given in the text table. Fractional amounts (biswahs) have been dropped in making calculations. These values for the Khandesh bigha differ considerably from either the imperial bigha-i Ilahi from Akbar’s period valued at 0.6 acres or the bigha-i daftri of the later seventeenth century valued at -.4 acres. See Irfan Habib, The Agrarian System of Mughal India, op. cit, pp. 353-66, for a fiill discussion of this question. The justification for using the Khandesh value in the

130 / Marathas, Marauders, and State Formation From the total 262,674 bighas of productive land just under 57 per cent were actually cultivated tax-paying lands. Another 37.5 per cent were fallow or grazing lands. The final category, just over 5 per cent of the total, was in the hands of tax-exempt grant-holders (aimmadars). The latter could have been either cultivated or fallow lands but was more likely to have been the former. The average village contained just over two thousand bighas (1500 acres) of productive lands. From this total twelve to thirteen hundred bighas (900 to 1000 acres) would have been actually sown in any year. Note two points about the graph on p. 379. First, villages included considerable grazing and common lands (uftada).14They were used, as they were in Europe at the same period, for a variety of village needs— grazing dairy cattle, firewood, and gathering of flowers and medicinal herbs. Both state and cultivator recognized that these lands were valuable, but they were not taxed. Second, and not obvious from the graph alone, is that only approximately one-half of the land regularly cultivated was actually cultivated in any year, with the remaining half allowed to lie fallow. This is a high percentage of fallow land for the period, and suggests low population pressure, In a situation of relatively low population density, we might expect cultivators to choose the most fertile and well-watered tracts. Revenue figures bear this out; because revenue demand per village was tied to annual serves and market prices, it is an indirect measure of productivity. We find, as expected, the Mughal demand per village on Adilabad was higher than Khandesh as a whole, and more than twice as high as nearby Malwa.15 calculation lies in the fact that the bigha values are embedded in land areas using auts and partans, that is, the original Maratha units of measure-not those of the empire. 14Crops are not designated in this Mughal record. The nineteenth-century crop pattern is well established. For the seventeenth century we do have scattered references from travellers referring to cotton, sugarcane and indigo as Khandesh crops. See Khandesh District Gazetteer, p. 250. Also Irfan Habib, Atlas of the Mughal Empire, op. cit., Map 14B, for the economic products of Adilabad and its vicinity noted as cotton and sugarcane. 15Lands classified as grants were not subject to a revenue demand. Irfan Habib, The Agrarian System, op. cit., lists the hasil-i- kamil for Khandesh at 4,080,019 rupees levied upon 6,339 villages (pp. 21,408). For Malwa the corresponding figures are 8,472,291 rupees levied upon 17,919 villages (pp. 20,408). Adilabad’s annual established collections (mugarrar hasil) were 168,539silver rupees and 11 annas in 1709. The demand for each bigha actually under cultivation in AD 1696was one and one-third rupees. That this is a relatively high imposition can be shown by several quick calculations. The total hasil figure for Khandesh province, at its highest level, for the same

Kinship and Pargana in Eighteenth-Century Khandesh / 131

PERCENT

MEASURED LANDS INADILABAD, 1696, BYUSE

(uftada) Measure Lands in Adilabad, 1696, by use

132 / Marathas, Marauders, and State Formation Both contemporary and later evidence, thus, suggests that Adilabad pargana’s arable lands were fertile and productive in this period. The Kunbi cultivators of this region were renowned for their industriousness and agricultural skill. The soil in Adilabad comprised ‘two kinds of black soil [first] the rich alluvial clay found north of [the town]’ that gave place to ‘a deep black loam [that] yields the finest crops’.16From this soil the Kunbi farmers grew the usual millets of the Deccan, oilseeds, as well as sugarcane and quantities of cotton.17The proximity of Burhanpur’s urban markets must have stimulated production of commodities for sale. IRRIGATION AND REVENUE ASSIGNMENT, 1696

The most important contribution to this agrarian productivity in Adilabad probably came from irrigation. Nearly every village within the sub­ district had access to some form of artificial water supply. In the uncertain rainfall regime of the Deccan this was a vital matter. We can quantify, to some extent, the importance of different forms of irrigation. Of Adilabad’s 124 villages, thirty-seven received irrigation directly from the Puma or the Tapti. In these villages, approximately 37,805 bighas were cultivated (not irrigated, cultivated) from a total of 148,400 bighas for the pargana. Thus, just over one-quarter of all land actually cultivated in Adilabad was located within a mile of these two rivers. Another twelve villages obtained irrigation waters from watercourses or canals (nala). Thus, over half of the villages in the pargana were beneficiaries of the long-standing system of riverine irrigation in Khandesh. This system dates from the early centuries of the Muslim conquest under the Farruqi dynasty, or perhaps even before. Local engineers and masons constructed stone dams or weirs across river courses where rocky ledges made this feasible.18 A late nineteenth-century gazetteer stressed their importance, as follows: The weirs, bandharas, must, at one time, have been very numerous. In the west there is scarcely a stream of any size without traces of them. Tradition attributes period reveals an average 643.6 rupees per year per village. For neighboring Malwa die average hasil per village was even lower at 472.8 rupees per year. 16The term uftada is discussed by Irfan Habib, Agrarian System, op. cit., pp. 5-6 and 302 n. He equates uftada with the ‘cultivable waste’ category of agricultural land in British period statistics. See Habib as well for a full discussion of subsistence grants, pp. 298-315. 17Khandesh District Gazetteer, p. 349. '•Ibid., p. 139.

Kinship and Pargana in Eighteenth-Century Khandesh / 133 their construction to the Musalman rulers___The sites of these dams were, as rule, well chosen. Except a few built straight across the stream, the dams are more

or less oblique, the watercourse issuing from the lower end___In building a dam, holes were cut in the rock in the proposed line of the wall from six to thirteen inches square, the same or more in depth, and from three to six feet apart. In the holes, stone upright, sometimes small pillars taken from Hindu temples, were set, and the dam was either built in front of these, or the stones were built into the dam---- The materials are common black basalt stone, coarse concrete mixed with small pieces of brick, and the very best cement___While the dams were built with the greatest care, the watercourses were laid out with the strictest economy, following the lie of the ground and making long bends to avoid cuttings or acqueducts.

These long gently graded watercourses allowed a slow, easily regu­ lated flow of water to the fields. The Farruqi rulers, and presumably the Mughals, appointed channel keepers (patkaris) to maintain dams and channels. They were compensated by grants of land.19 The report noted another fifty villages in the pargana drew upon wells (chah) for irrigation. In the vicinity of Adilabad, in the nineteenth century, wells varied from twenty-two to sixty feet in depth. Most were step-wells. Cultivators worked wells of this type with leather bags to draw water to the surface into earthen channels leading to the plots under irrigation. A good well permitted the use of four leather bags on a daily basis. Each bag irrigated one-quarter acre of land each day. The generally accepted maximum area of irrigation for each well was five acres of land. High yielding garden crops were best served by this type of irrigation according to the nineteenth-century sources.20 The overall importance of well irrigation obviously depended upon the number of wells in operation. The Mughal record does not provide this information, but we can surmise that the wells were numerous. In 1879, the number of wells was probably around one thousand in Adilabad section of Bhusaval subdivision.21If the same figure held in the seventeenth century the average per village for the pargana might have been eight or nine wells in operation. If the wells were confined to the fifty villages for which they are recorded this would have been closer to twenty wells per settlement. In short, we might guess that fifty to one hundred acres of prime land might have been watered by well irrigation in these villages. Management and appropriation of the considerable annual revenue 19Ibid., p. 140. “ Ibid., p. 143. 21Ibid., p. 144.

134 / Marathas, Marauders, and State Formation and Adilabad’s central location suggest the need for a substantial official presence within the pargana. The summary sheet lists several key figures: first, Muhammad Afzal, the faujdar, was an imperial officer (mansabdar), responsible for the maintenance of public order and the peaceful collec­ tion of revenues. He commanded, recruited and led a contingent of heavy cavalry. Following the usual practice, his jurisdiction probably extended beyond the boundaries of the sub-district. Second, the two Islamic judges (qazis), appointed by the regime, enforced both the creedal and commer­ cial provisions of the Muslim holy laws (the shariat). They also provided important legal and social services to the local Muslim community. Qazi Namauddin was attached to the town of Adilabad itself. His jurisdiction extended to 101 surrounding villages. Qazi Bakhtiyar Badha (?) was responsible for another twenty-three villages in the pargana. Third, the imperial intelligence reporter (waqi-nigar) submitted near-dajly reports of public events and official news directly to his superior in Ahmadabad. The emperior himself, or one of his highest ranking ministers, scrutinized local reports from all over the empire as a means of obtaining direct immediate information. Finally, Mahadev Naik, a leading member of the local Maratha aristocracy, not an imperial official per se, filled the joint position of deshmukh and qanungo. Under these designations he was responsible for ensuring the collection of imperial revenues in the pargana and for keeping accurate local records of tax collections and submissions. The first three officials were Muslim officers alien to the pargana sent on temporary duty by the imperial administration. The latter, Mahadev Naik, was a local notable whose family undoubtedly had held this position for some time. This cluster of officers, usually acting in concert, carried out vital executive, judicial, revenue, and communication functions in the pargana. As this official document makes clear, the revenues of each village and town within Adilabad pargana were assigned to an individual or an institution. Under the peculiar Mughal system of revenue assignments, termed jagirs, official revenues could be collected directly by individual officers as payment of their salary claims. These jagirdars, as they were termed, did not have to reside within the pargana or in the villages they held in jagir. But they or their agents would be present in Adilabad to collect regular instalments on the revenues. In 1996, the largest single jagirdar was the faujdar, Muhammad Afzal. He held in addition to Adilabad town itself, another fifteen larger villages for a total grant of 31,816 rupees (18.9 per cent of the total pargana income). Of the individual jagir holders listed against each village few

Kinship and Pargana in Eighteenth-Century Khandesh / 135 others were actually resident in the sub-district. Perhaps fifty officers held one or more villages in Adilabad in this listing. A more precise count is not possible because of the difficulty of deciphering some names and the tendency of the compilers to use varying forms of the same officer’s name and titles. A summary report ofjagirs for 1702, six years later, reveals that Muhammad Afzal was still faujdar with a jagir valued at 38,906 rupees. Another fifty-six officers, some serving as far away as the Hyderabad Kamatik, held jagirs in the pargana in that year.22 For the administration, the truly significant fiscal and political unit within the pargana was the village or town. At this level the most important group was not imperial officers per se. Instead, it was the body of village headmen and village accountants who together with the pargana deshmukh cooperated with the regime in keeping order and collecting taxes. The report lists 304 headmen called chaudhri or patel for the 124 villages of the pargana. The average is 2.5 headmen per village. The numbers ranged from one to as many as six in exceptional cases. The report also specified the name and title of each village accountant (patwari) as well. The repc»t lists only 133 village accountants. In nearly all the villages only a single accountant held the post in contrast to the headmen who more often shared their posts. When the entire list of these 437 men, plus the deshmukh, is read, two salient facts are inescapable. First, the empire at this level depended upon Hindu cooperation. Only one recognizably Muslim name appears—and that questionable—for one village account (one Shaikh Shakarullah, known as Shankraji). The remainder are all Maratha names ending nearly invariably with the ji honorific (Kishanji, Timaji, etc.). By contrast the jagir holders in Adilabad were nearly all Muslim. Only two recognizably Hindu, not necessarily Maratha, names appeared in the list ofjagir holders for 1696. The cultural and ethnic divide between the imperial superstruc­ ture and local policy is starkly revealed. Second, there is no indication of kinship, caste, or lineage for any o f these m en. T hese m en were treated

singly, individually, as office-holders responsible for land taxes. Caste and lineage, in spite of their local significance, were not important to the imperial administration. ADILABAD: CONFLICT AND DUAL ADMINISTRATION, 1700-50

It is important to put Adilabad into the political and military events taking place around it. Khandesh, in 1700, was a province at war. Raja Ram, titular head of the Maratha confederacy, led the first major raid into “ 1/17/95-6.

136 / Marathas, Marauders, and State Formation Khandesh the previous year. He joined several commanders, including Nimaji Shinde, who had been raiding the province for several years. The combined forces attacked the nearby entrepot of Burhanpur in 1702, almost certainly crossing Adilabad on the way.23 Until 1717, Maratha energy was spent in internecine wars. Several leaders had, however, settled in Khandesh, carving out rights and territories. Khanderao Dhabade, for example, established a line of posts along the Burhanpur Surat route and charged one-quarter (chauth) on the value of all goods on the road. Husein Ali Khan and an 8000-men Mughal army attempted to reopen this road in 1718. The decisive defeat of the Mughal army by the Marathas led to a treaty, in 1719, by which emperor Muhammad Shah, in the first year of his reign, conceded one-fourth of the whole revenues of the Deccan to the Marathas.24Thus began a new phase in Adilabad’s history. Every year, thereafter, from 1719 to 1724, the Peshwa led raids into Khandesh. The Marathas had, at this time, virtually no centralized bureaucracy; it was built through the 1720s. The Peshwa, Baji Rao, as head of the new

bureaucracy, began to receive reports and accounts from his Brahmin subordinates in Khandesh in the mid 1720s. These suggest the level of disruption caused by Mughal-Maratha warfare. Of 124 Adilabad villages, eleven were deserted in 1724 and twenty were deserted in 1725. (In nearby Bomar pargana, twelve of the forty-three villages were deserted.) Another indication of disruption is the rare occurrence in forty villages of ‘shares’ settlement (battai), a division of produce at harvest or on the threshing floor.25 Overall revenue remained low, around Rs 20,000, far below the levels of 1700-5.26The final indication that the administration was struggling in these early years is the rapid change in personnel. In 1724 and 1725, Pushoram Chimnaji was collector; the Peshwa replaced him with Ragho Krishna in 1728, and in 1729 Malhar Baskar was appointed.27Narain Dikshit followed him in 1731.28The Maratha admin23James Grant Duff, History of the Maharattas (New Delhi, reprint edition, 1971), vol. 1, pp. 223—4. See also, G. T. Kulkami, The Mughal-Maratha Relations: Twenty-five Fateful Years (1682-1707), Pune, DeccanCollege, 1983, for details ofthe military campaigns ofthis period. MIbid.,p.443. 25 Pune Daftar, Peshwa Khandesh Collection, 38, Hisseb of Babti, 1724, hereafterreferred to as P.K. “ P.K., 38, Hisseb of Babti, 1725. ” P.K., 198, Hisseb Ismali, 1729. “ P.K., 236, Hisseb of Babti and Sardeshmukhi, 1731.

Kinship and Pargana in Eighteenth-Century Khandesh /137

Fig. 1 istration was spread so thin in these early years that the collection area of

each of these men included not only Adilabad, but five other parganas clustered some forty miles to the west. Through the 1730s the pargana appears to have settled into a regular ‘dual administration’. The Mughals collected their share, and the Marathas theirs. The documents unfortunately do not give details of Adilabad alone, but continue to lump Adilabad with the earlier group of parganas— Bomar, Lohara, Yaval, Chandsur, and Pachore. Note, in Figure 1, how the collected revenue for the six parganas rose steadily through the 1730s. It is only in 1739 that we have detailed revenue documents of Adilabad alone. Recovery was largely complete. Of 124 villages, only four were deserted.29 Because Adilabad was still a functioning Mughal pargana, certain villages were excluded from revenue payments to the Marathas, and, thus, disappear from our records. Specifically, ten villages were assigned to the Mughal auditor (muzmu) and six more were assigned to unnamed jagirdars. The remaining ninety-four villages yielded Rs 96,314 of which Rs ^4,078 (one-quarter) went to the Marathas.30 All but about Rs 4000 is marked paid in the government treasury (vasul). With the 29P.K., 194,Terij, 1739. MP.K„ 194, Yadidast, 1739.

138 / Marathas, Marauders, and State Formation assigned villages, the collection was about three-quarters of the collec­ tion of forty years earlier.31 The Maratha administrative records suggest the degree of bureaucratic development and penetration. Records from 1748 list each Mughal religious grantee (aimdar) and the amount each was taxed. For example, on a grant to Mullah Hazarat Khan worth Rs 200, he paid Rs 19 to the village headman, Rs 41 to the village book-keeper, Rs 17 to the Maratha regional official (sardeshmukh) and Rs 43 directly to the Maratha government.32 The Marathas even collected a small percentage of the town taxes on commodities coming in for sale to Adilabad, and paid a percentage of the cost of the city’s ten watchmen.33 At mid century, political events once more overtook Adilabad. The Maratha central government went to war with the Nizam and his allies, the Gaikwad-Bande coalition of Maratha houses. Warfare raged across Khandesh. The seizure of the estates and town of Adilabad from Mughal grantees severely disrupted the pargana. In spite of working with admin­ istrative documents, one gets a surprisingly graphic picture of the proceedings. Throughout 1751, the Peshwa’s day diary has dozens of notations of armed cavalry sent to individual parganas or villages, often following warning letters that the entire revenue should be paid to the Marathas, not the Mughals. The troops often met resistance. Note, for example, this stem letter from the Peshwa to a Maratha apparently allied to the Mughals: Warning letter to Pamaji Mahadev. You have remained at Birkul in pargana Leling, along with your troops and you are preventing the Kamavisdar’s collecting the revenue and confiscating [the Mughal estates]. Formerly you were informed that you should not remain there___You have written letters to us, but do not expect any reply. Further, if there is any complaint against you, it will not be tolerated. By your actions, government is at a loss because revenue cannot be collected. A letter should be sent to him that he must not stay there, but go away, we give assurance of his safety.34

Stray documents give the details of the patrols; these are notations in the Peshwa’s daybook: As the pargana officials were prepared to make a settlement three headmen from the village were held in confinement and 5,000 rupees were taken from them. 31P.K,, 194, Terij of jama and vasul, 1739. 3lP.K., 194, Hakikataima, 1748. 33P.K., 194, Terij Sair and Rahadari, 1748. 34P.K., 203, Memorandum of confiscation, 1749.

Kinship and Pargana in Eighteenth-Century Khandesh /139 From village Nandgav of pargana Manikpunj. Confiscation was taken and orders were given for the release of the village headman. Rupees 701 taken. A letter of confiscation was sent to the head of Utran pargana; he was summoned and imprisoned. He was ordered to be released. This fine was, therefore, taken from him.35 By the end of the year, revenue estimates were in shambles. Appeals for remission came in from all parts of the pargana. Typical is this one with its tacit acceptance of disruption: Letter of agreement given to government by the headman Baji of village Kotli Pargana Adilabad 1751. Earlier in the year the Peshwa wrote him and ordered him to pay the revenue which had been fixed. He declared that he was unable to pay the revenue. The government fixed the settlement according to the year’s report of actual measurement of land under cultivation. Accordingly the revenue is fixed at Khandesh taka 251.36 Other similar documents give the details that the pargana-level offi­ cials were directly involved in the report of the land actually remaining in cultivation, from which a new settlement was negotiated. One is perhaps worth quoting, in full: Letter of agreement. Jahankhan patil and Badghe patil and Sand and Yes Chaudhris of Kasbe Satavgav of pargana Adilabad in 1751 in writing gave a letter of agreement that the Peshwa has called us and ordered us to settle the revenue. We replied that we are unable to pay. Considering future prosperity [or cultivation], Gangapant, representative of the Deshmukh, appealed for a remis­ sion of the revenue and Hammat Khan agreed to stand surety along with the chogule [or chaudhri] for the revenue at ‘fixed’ Khandeshi takas [a monetary unit] 2-0-.5, for the current year. For our share [of the revenue] if there is any alteration, we will be held as outlaws. So this letter of agreement is written.37 /

As the Marathas struggled to take control, we get a most intimate look at Adilabad. The officials seized the ‘dastur amal’, a crucial Mughal register showing the prevailing rates of taxation in Adilabad.38 It is worth considering this document in some detail. The first section shows the divisions of the harvest (battai), if government and grower were unable to agree on a cash taxation. For irrigated garden land (bagayat), the division was two shares for the farmer and one for the government on a wide variety of produce, such as ginger, turmeric, vegetables, safflower, 55P.K., 30, Rajamandal, Memorandum of warning, 1751. 3P.K„ 128, Dastur Amal of Adilabad, 1751.

140 / Marathas, Marauders, and State Formation tobacco, flowers, opium, sweet potatoes, dal, and irrigated wheat. On unirrigated wheat and coarse grains, the division was equal. The next section has cash amounts for some items (which are largely meaningless since we do not have prevailing market prices). The last two sections were the crucial normative guidelines for the recovery of Adilabad pargana. The first was the payment schedule for bringing uncultivated land into grain production. It consisted of a six-year stepwise division between government and cultivator that began with the five shares to the cultivator and one to the government, and ended with an equal division, in the sixth year. Equally important, the ‘dastur amal’ spelled out the pay of the pargana-level officials. The deshmukh was to receive 2.5 per cent of the revenue, while the deshpande received 1.25 per cent. Finally, pay for the non-cultivating village servants was spelled out. The carpenter, shoe­ maker, and iron-worker received small shares of the village revenue, while the barber, potter, washerman and village records keeper were specifically to receive their shares (probably non-monetized) from the village with no central government involvement. Supplementing this dastur amal were village settlement contracts. We have a complete set for 1752, one for each of Adilabad’s 106 inhabited villages. The agreement for Salogav, for example, lists fifty-six cultiva­ tors tilling slightly over nine auts (696 acres), and the cash demand. The contract has the sign (the plow) of the headman and the signature of the Maratha kamavisdar.39 The war ended on 24 November 1752, with the Treaty of Bhalke. Salabat Jung ceded to the Peshwa, among other territories, the suba of Burhanpur.40 The ‘dual administration’ ended, and Adilabad along with the rest of Khandesh became Maratha. Before we leave this history of Adilabad, let us quote, in full, the Peshwa’s instructions in 1753 to his kamavisdar, Laxman Bhikaji, working in the pargana. This year the Mughal share of Adilabad came under the jurisdiction of the Peshwa and its administration was entrusted to you. So you should do the work honestly and pay in advance, as agreed, 25,000 rupees, interest at 1 rupee per 100 per month. Interest and advance should be collected from the people [of your district]. Whenever you pay accounts to the government treasury take a receipt Mushira (fee for conversion of collected coinage to standard rupees) is 3 rupees per one hundred collected. If even the advance is not collectible in the district, government will pay the interest. So a sanad is granted. Gandadar Bhijaji should take up the work of book-keeper (phadnis) for which he will receive 200 rupees 39P.K., 128, Kaulbandi Zamin, Salogav, 1751. 40Selectionsfrom the Peshwa Daftar, New Series, I (Bombay, 1964), Letter 155, p. 140.

Kinship and Pargana in Eighteenth-Century Khandesh / 141 per year. After allowable expenses, all remaining revenue should be sent to the government.41 ADILABAD’S IMPORTANCE AS AN ADMINISTRATIVE UNIT

Consider the powers vested at the pargana level. Adilabad, throughout the period, was treated as a revenue unit. In Maratha times, it was the parganalevel administrator, the kamavisdar, who signed the revenue contracts with individual cultivators and village headmen. These yearly negotia­ tions made the pargana-level administration virtually the only on-going Maratha presence in the area. Equally important, it was the pargana-level officials who initiated appeals for remission during times of adversity— whether from drought, flood, or invasion. The headmen of villages seek­ ing remission had to come ‘face-to-face’ with the kamavisdar, usually at the head town.42 The kamavisdar’s report was the major evidence on which the Peshwa decided the case. The pargana-level administration also had judicial duties; the kamavisdar investigated disputed local rights, such as claims to fill the offices of deshmukh, qanungo, and village head­ men. These were usually decided right at the pargana level. Each year, also, the kamavisdar decided a few dozen criminal and civil cases involv­ ing everything from murder to false weights, within his pargana. This would have been an immediate judicial presence for a pargana like Adilabad, because there were no lower courts (except village or caste punchayats). The only appeal from a kamavisdar’s decision was to the Peshwa, a costly and difficult venture at Poona some 300 miles away. The kamavisdar reviewed all religious grants (aima) in the pargana; he surveyed the land, fixed taxes, and adjudicated disputes. In a similar fashion, the pargana mattered to trade. The kamavisdar regulated market towns (qasbahs), extracted the transit duties (zakat), and, again, adjudi­ cated disputes in the markets. The pargana-level administration could back these decisions with force. A small contingent, fifteen to thirty men, was routinely stationed at a pargana the size of Adilabad. In the case of Adilabad, however, more force was close by. The major fortress of Asir, with its contingent of several hundred regular soldiers lay only a day’s ride away.43 41P.K., 235, Letter from Peshwa to Laxman Bhikaji, 1753. 42For an extended discussion of this process, see Stewart Gordon, ‘Recovery from Adversity in Eighteenth Century India: Rethinking Villages, Peasants, and Politics in PreModem Kingdoms,’ Peasant Studies, viii, 4 (Fall, 1979), pp. 61-80. 41 Asir fort, and its influence: is discussed in Stewart Gordon, ‘Forts and Social Control in the Maratha State’, Modem Asian Studies, 13 (1979), pp. 1-17.

I

142 / Marathas, Marauders, and State Formation Lastly, the pargana administration was responsible for a series of locally important economic decisions. Adilabad’s kamavisdar, for ex­ ample, was responsible for offering lightly taxed five-year development loans (taqavi) used to open new land, or repopulate villages, or build dams. He also employed and supervised local watchmen (such as the ten at Adilabad city), petty tax collectors on the roads, and all the menials and troops of his establishment. The kamavisdar made local purchases of cloth and candles, paper, swords, and food, to stock and supply his office. ADILABAD ACROSS TWO EMPIRES

Let us return to a related question, raised at the opening of the paper. What were the continuities and discontinuities between the Mughal and Maratha empires in Adilabad pargana? Continuities were many and basic. In administration, for example, the Marathas adopted Mughal terminology wholesale. Documents bristle with Mughal terms like jama (settlement), teriz (receipts), izafa (cultivated). The basis of demand— the maximum expected in the best of times—derived from Mughal figures. More importantly, the method of field assessment and cash com­ mutation was Mughal. Even such basics as the number of villages, that is, what constituted Adilabad, passed directly from Mughal to Maratha. Both empires acknowledged that Adilabad had 124 villages, though at no time during our study were they all inhabited. Possession of critical imperial Mughal documents, such as the settlement rates (dastur amal) was important, nearly as important as military might, to the successful Maratha conquest of the pargana. Some of the local personnel remained the same. The Marathas acknowledged and continued all village head­ men, record keepers, and religious grantees (aima), as long as their Mughal grant was properly signed and sealed. Finally, the local head of the pargana (deshmukh) remained in the Naik family; Mahadev Naik held the office in the early years of the century, with Vasudev Naik replacing him in the mid-century documents. None of this should surprise us. After all, preceding the Maratha takeover were twenty years of dual Mughal Maratha administration. Equally significant, and often overlooked, is the long period of Muslim rule in the Deccan. Maharashtra had been familiar with Muslim admin­ istration for two hundred years before Mughal invasion of the South. The conquering Marathas had grown up for generations under a Muslim administrative system. Thus, what constituted ‘Adilabad’ was as much an adminis'trative

Kinship and Pargana in Eighteenth-Century Khandesh / 143 blueprint—a complex of normative, settlement and taxation documents, plus village registers and signed contractual obligations—as anything existing on the ground. Even when the area was depopulated (as it was between 1810 and 1818), the pargana could be re-built, even with different cultivating peasants, from the pattern contained in the docu­ ments. In spite of these continuities, there were several major discontinuities at the pargana level between the Mughal and Maratha empires. The first change is easy to see in the documents, but difficult to assess. Between the Mughal documents of 1696-1705 and the Maratha village lists of 1755, there is considerable disparity in village names. In fact, less than half the village names correlate, even when using the additional data of geographic co-ordinates. The disparity is even greater with the first British maps of the 1820s. This contrasts sharply with Rajasthan, for example, where researchers have been able to pinpoint over 90 per cent of all villages mentioned in a seventeenth-century register, with the use of modem census data and modern maps.44 The most likely explanation has to do with Adilabad’s extensive periods of adversity, 1700-20,175060, and 1795-1818. We might speculate that depopulated villages were not repopulated on the same site. The continuity of Adilabad’s ‘124 villa­ ges’ may mask considerably more movement, depopulation, adversity, and colonization than has been recognized. Further research into the village lists is necessary to establish this. In comparison to this sort of discontinuity, the others seem less significant. First, there were no more Mughal military grants (jagirs). Right to the end, the Mughal government persisted in single village grants. After the Maratha conquest, the pargana was granted and treated as a unit; revenue might be divided, but administration was not. Second, the administration, though using Mughal terminology, was Brahmin rather than Muslim and spoke Marathi rather than Persian. The difference may have been not significant to the villagers. Large sections of the Mughal, and even earlier Muslim, financial administrations south of the Narmada had been staffed by Brahmins. For example, both the Qutb Shahi and Adil Shahi revenue departments were run by Brahmins. Kohlapur Brahmins formed the fiscal administration in the Kamatak, especially in the Bangalore area; Brahmins from Ahmadnagar and northern Maharashtra were earlier recruited by the invading Mughals, to 44 Richard Saran, ‘Land Revenue Realization and Irrigation Devices in Metro Pargana, Rajasthan, 1658-1663’, unpublished paper presented at the 12th Conference on South-Asia November 1983, Madison, Wisconsin.

144 / Marathas, Marauders, and State Formation administer areas further south.45 It is at least possible that Adilabad’s villagers dealt with Brahmin collectors both before and after the Maratha conquest. Third, the credit and economic networks changed. Adilabad’s revenue flowed to Poona, rather than Delhi. The advance money the collector needed came from a growing banking community in Poona. All of this suggests that, in Khandesh, parganas had little history of their own. They Were on the receiving end of major political and military events. At the pargana level, raids, wars, and changes of empire meant destroyed and vacant villages, and depopulation, with no evidence of the re-assertion of a strong clan/lineage structure at the pargana level. LINEAGES AND PARGANAS

Adilabad was a pargana without the strong clan/lineage structure we have come to expect from recent studies of Rajput clans in U.P. We find numerous markers of the difference, but let us suggest the following: (1) Unlike Rajasthan or the Punjab, the pargana was not name^i for a dominant clan. (2) The deshmukh, the local head of the pargana, held no independent villages. The deshmukh was paid a small percentage of the revenue. Thus, even the elite of the ‘clan’ had no revenue base indepen­ dent of the central government. (3) Revenue settlements were with cultivators of individual villages, not with a dominant clan. Thus, the central government had full, yearly statistics on actual and expected cultivation. (4) Appeals for remission in bad times came from the village headman and were only secondarily routed to the deshmukh. (5) Though the deshmukh signed the pargana revenue contract, he was not ultimately responsible for unpaid taxes. Arrears were broken down village by village. (6) In times of serious adversity (1715-20, 1749-55, 1800-18), the deshmukh’s troops (if they existed at all, we have no positive evidence) were incapable of keeping local order. If central government forces were occupied in war or revolution, no local force could stop the raids of tribal Bhils, and pargana such as Adilabad went into a downward cycle of devastation and depopulation. These cycles stopped only with the re-assertion of strong central government control.46It was simply not possible to build or rebuild an empire, in Khandesh, out of negotiations with the pargana leaders. Why not? First and foremost, the deshmukhs of Khandesh were not Rajputs and not heads of Rajput clans. For example, the deshmukhs of 45Op. cit., Mackenzie, introduction, pp. 6-6. 46Stewart Gordon, op. cit., pp. 1-17.

Kinship and Pargana in Eighteenth-Century Khandesh / 145 Raver and Jamner (parganas in central Khandesh) were Reven Kunbis; Chopda’s deshmukh was a Dore-Gujar. Other Kunbis held the deshmukhis of Yaval, Amalner, and Varangao. A Mali family held the office in Erandol. In the area of Adilabad, most of the deshmukhs were Marathas.47 To see the disappearance of the clan-based pargana as one moves south, consider the crucial area between Rajasthan and Khandesh, the Malwa plateau. Across the plateau, the basis for parganas was mixed. Some were co-extensive with Rajput clans. In the sixteenth and seventeenth centu­ ries, many of these clans had been ‘planted’ by the Mughals to control local populations. Other parganas, however, had no such clan basis. They were the immediate hinterland of small trading and manufacturing towns.48However, as one reached south, to Khandesh, the Mughal Empire was notably unsuccessful in ‘planting’ Rajput clans to control or conquer local populations. Further, the Rajput ideology of lineage loyalty and service (as bril­ liantly analysed by Norman Ziegler) had little appeal south of the Narmada.49 Some Maratha houses made half-hearted attempts to assert Rajput status, but few ever married daughters to Rajput houses, and most—such as Sindhia—pursued statehood without ever aspiring to Rajput status. Curiously, the last people to whom the Rajput model appealed were the tribal Bhils of the hill areas surrounding the Khandesh valley. Some of them adopted Rajput names, dress, and a few succeeded in marrying daughters to lesser Rajput lines. Some even succeeded in carving out tiny kingdoms during the chaos of 1800-18, much as Rajputs had in Oudh three centuries earlier.50 Be that as it may, Khandesh’s pargana heads were, on the whole, Kunbis and Marathas, just as were the overwhelming majority of its village headmen and cultivating popula­ tion. 47Khandesh Gazetteer, pp. 62-3. 48The histories of several of the ‘planted’ Rajput clans are found in John Malcolm. A Memoir o f Central India (London, 1832), and Raghubir Sinh, Malwa in Transition, or A Century of Anarchy and the First Phase 1698-1765 (Bombay, 1936). 49Note, for example, the differences between Maratha and Rajput military recruitment patterns. Rajputs tended to recruit kinsmen, and could often raise an army of hundreds of armed men, through lineage channels alone. Marathas offered their ‘service’ to a leader, not necessarily a kinsman, sometimes alone, often with only immediate family, and occasionally as head of a band of non-relatives. Leaders often changed sides in factional disputes, regardless of kin ties. This pattern is illustrated by the lives of some of the well-known Maratha leaders, such as Malhar Rao Holkar, Mahadji Sindhia, and the Pawar brothers. 50Surajit Sinha, ‘State Formation and Rajput Myth in Tribal Central India’, Man in India, 4 2 ,1 (January-March 1962), pp. 35-79.

146 / Marathas, Marauders, and State Formation The final reason that Adilabad did not look like the clan-based model of parganas in the north has to do with economics. Richard Fox is explicit in the dependence of his model on a largely isolated, non-monetized, and self-sufficient region. It was in this sort of arena that Rajput pargana/ lineages could prosper, ‘segment’, and negotiate with the central govern­ ment. The kinds of goods and services which could be extracted from such an area—fodder, soldiers, grain, and occasional cash tribute-meshed perfectly with the lineage-based pargana structure found in Oudh. ADILABAD AS A MONETISED HINTERLAND

Adilabad, in eastern Khandesh, was a different story. As we shall see, it was exactly the reverse, that is, a heavily monetized, market-oriented pargana, close to a major entrepot. What were the pressures on Adilabad towards monetization? First and foremost would have to be government land tax. The evidence is highly specific and highly detailed that taxes were paid universally in cash, not in kind. Typical are a set of documents from 1769. Each is a year-end account of a single village showing the receipts and outstanding balance. These were not simply items of account. The finer grained documents show they are actual cash contracts. Khandesh bundle 128 contains the actual land measurements (zaminzhada) and the field-by-field surveys of the fields currently worked by each cultivator in the village; often these documents also record the crops grown on these fields.51 Attached documents record the type of land and the contract agreement for revenue due.52Except for special surveys arising out of disputed land, what we are seeing was described by George Giberne, one of the British ‘men-onhorseback’, sixty years later, in 1820. In a document termed Kowl Katavenee, he [the Kulkami] records the engage­ ment made with each individual cultivator or Ryot on taking possession, that he shall pay so much per begah to the Circar [government], etc.53

According to Gilberne, the Kulkarni kept a record of actual payments (tahsil). Everyone above the cultivator expected that taxes were going to 51P.K., 189, Jama vasul. 32P.K., 184, Kaulbandi, 1963. 53 George Gilberne, Report on the System of Revenue Management within the Collectorate of Candeish, 1828. The only known copy of this report is at the British Museum, London. It is catalogued as printed book, but is, in fact, a MSS, in the hand of the author. B.M. 793 m 17(11).

Kinship and Pargana in Eighteenth-Century Khandesh / 147 be paid in cash. First, the village headman most often gave the surety (or his guarantee) that the money, in cash, would be paid. The Maratha collector (kamavisdar) expected cash. The collector had to pay, in cash, a high percentage of the estimated revenue at the beginning of the reve­ nue year—50 per cent and above. This money, often substantial sums (Rs 25,000 for Adilabad), he got from the Poona banking community. The kamavisdar was allowed 1-2 per cent per month on the uncollected balance; he certainly expected and regularly collected the cash from his district to pay back the loans. Recall the Peshwa’s instructions to Laxman Bhikaji, his kamavisdar at Adilabad, in 1753. Interest and advance should be collected from the people [of your district]. Whenever you pay accounts to go the government treasury, take a receipt. Mushira [the charge to convert the currency collected to standard rupees] will be 3 rupees per one hundred collected.

At the top was the central government, which had substantial on-going cash needs. For example, all civilian officials were paid in cash; religious grants were in cash; supplies, from roof tiles in a fort, to paper, to oil for lamps—were bought with cash; all military supplies—horses, grain, tents, weapons—were bought with cash; finally, all military personnel were paid in cash. Cash needs of the Marathas, in the eighteenth century, were possibly even greater than the Mughals because a much higher percentage of the army was paid directly in cash. There is no evidence, for example, in Khandesh of soldiers getting small parcels of land in lieu of salary. Everywhere in the Martha Empire, soldiers wanted cash on the line, not land. Most military crises throughout the eighteenth century, including European, were, in fact, shortages of cash to pay the troops. The degree of monetization, in taxation, is so striking that the very few times there was payment in kind, they stand right out. In approximately 250 parganas examined in Malwa and Khandesh, there were only a few

cases of settlement in kind. Two were situations in which Maratha officials and the cultivators could not arrive at any settlement. They re­ sorted to division on the threshing floor (battai). Another case consisted of the thirty-three villages surrounding Asir fort. They were specifically settled ‘in kind’ to provide supplies for the fort. The only other time that divisions (battai) appears in the documents was during severe political or other disruption. Recall that we found battai in Adilabad during the mid 1720s and again in 1750-3, both periods of war and dislocation. A battai settlement, in these terms, is understandable, since neither government

148 / Marathas, Marauders, and State Formation nor cultivator could sign a cash contract for the coming harvest. The times were just too precarious. So, if we concede that in normal times cultivators, across the pargana of Adilabad (and across the province of Khandesh) paid taxes in cash, where did the money come from? For McLeod, who examined the countryside on horseback in 1819, this does not seem to have been a problem. On the Ryots having agreed to pay the specific sum, they repaired home, and set about finding the means of answering the demand. For this purpose, they either sold some of their grain or other articles, or raised money from the Bannian [bankers] or by other means, which they paid, as they could get it, in the hands of the Patells in the presence of the Koolkumees, Potdars, and Mahar of the Village; the Patell gave the money for examination to the Potdar or Sonar who marked and kept it.54

The Maratha materials only corroborate McLeod’s statement. The documents have dozens of creative reasons why cultivators could not pay, including plant disease, war, drought, robbery, and so on. No cultivator ever appealed for remission because of the absence or shortage of cash. So, cultivators raised cash by, as McLeod puts it, (1) sale of grain or other articles, (2) raising money from the bania (banker), (3) or other means. Let us look at number one—sale of grain or other articles. Consider the population of Adilabad as sellers of goods and services, rather than as tax payers. The evidence suggests that there was a much wider range of crops grown than has been generally assumed. All the following were listed as regularly grown and taxed in Adilabad: sugar­ cane, potato, sweet potato, eggplant, other vegetables, ginger, turmeric, opium, wheat, channa dal, safflower, flowers (for scents), tobacco, mogra, jawar, rala (Italian millet), bhadali (coarse grain), chaff and straw. Although the documents do not directly show that Adilabad’s cultiva­ tors sold their goods in Burhanpur, the circumstantial evidence is strong. All the commodities Adilabad grew were, in fact, for sale in Burhanpur. The market records of 1756 document them. Also, as you will recall, Burhanpur city was only twenty miles from Adilabad; even with poor roads, this was one day’s bullock cart ride. There were other documented ways that cash entered the Adilabad economy. A major source was the central government. Of all taxes 54 India Office Library, MSS. European D.32 (Erskine Collection), ‘Account of the Revenue System of the Maratha Country under Nane Fumavese by Lt John Macleod, 1818*, p. 492.

Kinship and Pargana in Eighteenth-Century Khandesh / 149 collected, approximately one-fifth was spent locally. These were spent on repair of roads, walls, dams, temples, and government buildings. Over and above these local expenses, the government purchased supplies— paper, candles, food. In Adilabad, this amounted to over Rs 1000 each year. Around the town of Adilabad, twenty men were hired on monthly salary as guards. Monthly wages were paid to a body of locally recruited soldiers.55We should remember that soldiers probably sent money home just as they have in every other war. This would return cash to the countryside. Local officials—deshmukhs and deshpandes—were paid by the government in cash along with other non-cash perquisites. Finally, the government put cash directly into agriculture, in the form of loans given for expansion of agriculture or for recovery from adversity (tacavi). FINAL THOUGHTS ON PARGANAS AND HINTERLANDS

The main point suggested by this excursion into eighteenth-century Khandesh is the danger of generalizing from one small region to the whole Mughal Empire. While parganas and Rajput clans might have been coterminous in Oudh, the conjunction was less frequent in Malwa, and largely disappeared as far south as Khandesh. While we indeed found walled towns, and the title of deshmukh, these were not markers of strong Rajput-style lineages. Besides conquest, there were various bases of pargana organization in Khandesh, as follows: (1) The area surrounding a town (qasbah ). The villages supplying produce and

raw materials for the manufactures of the town might form a pargana. (2) The area surrounding a fort, especially a large fort. For example, Asir fort was surrounded by Asir pargana. This was once the territory of an indepen­ dent chief, but he had been displaced before 1200, ahd replaced by the bureaucracy of each successive kingdom. (3) A development grant with attendant rights. In the two hundred years of Mughal rule, groups had peacefully migrated into areas of Khandesh, and

developed (under government patronage) the land. As Kunbis, they worked the land, and it was their leader, likewise a Kunbi, who was the deshmukh. This appears to be the pattern throughout central and eastern Khandesh.

All this suggests that we need to revise our mental map of eighteenthcentury Khandesh. Rather than isolated, non-monetised villages or regions, we need to connect villages to cities and towns. Villages, parti­ cularly in a pargana like Adilabad, were regularly growing cash crops and 55P.K., 125, Dastur Amal of Adilabad.

150 / Marathas, Marauders, and State Formation finding markets for them. Except in times of severe adversity, they sold their crops and paid taxes in cash. Cities, like Burhanpur, depended on their hinterlands, like Adilabad, for labour, raw materials, and food. Our research can no longer separate city from hinterland, or city from pargana. They must be treated as one system. Finally, what of our larger question, the basis and importance of parganas as small regional units in pre-colonial India. Two points emerge from the study. The first is the variety of contexts for parganas in different regions. In Rajasthan and U.P., parganas were the result of Rajput immigration and conquest, and the spatial grouping preceded Mughal conquest. In Malwa, the basis of parganas varied. Some were co­ extensive with older Rajput settlements, some were more recent Rajput lineages ‘planted’ to control local populations, and some were simply villages attached to forts or towns. In Khandesh, the pargana appeared largely without strong lineage underpinnings and functioned more as a government unit than as an indigenous ‘little kingdom’. We await re­ search that compares these areas to areas further to the south— Maharashtra, Nagpur, or Mysore. Second, parganas were important to local popula­ tions, regardless of whether they were Rajput lineages or government groupings. The parganas served as the local focus for tax collection, appeals, development loans, and employment. The pargana was, of course, equally important to the central government, with crucial pargana administrative and settlement documents seized by each successive empire—Mughal, Maratha, and British.

Chapter Seven

Bhils and the Idea of a Criminal Tribe in Nineteenth-Century India

The concept of a ‘criminal tribe’ was fully institutionalized in India by the end of the nineteenth century. Legislation named the tribes, defined their place of abode, their required occupation and their children’s education. It also restricted their movement and permitted special classes of evidence (such as informers) in cases involving them.1 To illustrate the process and the effects of becoming a ‘criminal tribe’ this essay will focus on the Bhils. Attempts to identify this group reveal some of the problems early British administrators solved neatly by categorizing them as a ‘criminal tribe’. In the extensive literature of the last one hundred and fifty years, Bhils have been characterized as follows: (1) descendants of Dravidians, or pre-Dravidans, or Mundas, or pre-Mundas; (2) autochthonous inhabitants of Central India or recent immigrants from Rajasthan; (3) animists, or Hindus; (4) fanners, or hunters-and-gatherers; (5) racially distinct people, or indistinguishable from lower-caste Hindus; (6) outcastes or ‘no-castes’; (7) inhabitants of hilly terrain, or of the plains.

All of this sounds confusing; I think it should. The common criteria for defining ethnicity are not readily applicable. Bhils, for example, do not have a common language; they speak some variant of the local plains language. Further, they do not share a common origin myth or pantheon of gods; their gods tend to be local or adopted from the Hindu pantheon. Even lineage names are shared with Rajputs. Still, there are at present nearly three million people termed Bhil in the census inhabiting a broad triangle touching Rajasthan, Gujerat and Central Khandesh. 1 Reprinted from Anand E. Yang, ed., Crime and Criminality in British India (Tucson, Arizona, 1985).

152 / Marathas, Marauders, and State Formation THE HISTORICAL BACKGROUND

The first—and probably the best—attempt to piece together a history of the Bhils-in the Mughal-Maratha period is in an official letter by John Briggs, first Commissioner of Khandesh, in 1825.2 Based on interviews, Briggs concluded that prior to about 1600, Bhils lived mainly in Baglana, an isolated tract in northeast Maharastra, and possibly in other pockets in the hills separating Malwa from Rajasthan. During the Mughal period some moved into areas that controlled passes into the Khandesh valley (either from the north or the south). With or without government grants, they collected levies from passing caravans and plains villages below the hills.3 Government policy fluctuated between conciliatory grants and punitive raids. When Maratha rule replaced Mughal authority in the early eighteenth century, the basic facts remained the same. Bhils continued to live in the most ecologically vulnerable environments. Their shifting agriculture was primitive; it yielded little spare resources to store for the periodic droughts in the hill areas. Those in the hills hunted with bow and arrow. The culture was largely self-sufficient, maintaining its own priests, holy places and spirit-possession ceremonies. Some came down from the hills to trade forest products at large fairs. They continued to ‘tax’ or raid both caravans and plains villages. \ To control Bhils, the Marathas tried both ‘carrot’ and ‘stick’ tactics. A\ few Bhil leaders were given grants specifying certain rights (such as duty from travellers) in return for protecting the passes. More commonly, Bhils were offered jobs as watchmen of forts or gates. Still others were encouraged to setde on the plains and become peasant cultivators. Even in the best of times, however, there were raids on the plains villages. The Maratha response depended on the effectiveness of the local official, the kamavisdar. Normally the Maratha official levied a special tax, termed Bhil-patti, on villages most threatened; the tax was used to hire more troops.4 The cavalry chased Bhils up into the hills, often summarily killing those they caught. Alternatively, they extracted hos­ tages and sureties. Throughout the period, there was a high level of mutual mistrust between the plains government and the Bhils. 2Political and Secret, vol. 60, Consultations of 22 June 1825, India Office Library, London (IOL). 3Briggs had often noted these earlier grants. See the discussions of Mupawa in Political and Secret, vol. 22,12 June 1822, pp. 3224-7. 4Poona Daftar, Peshwa Khandesh Collection, rumal no. 236, a jarma (account) of Bhilpatti in Khandesh and its expenditure for patrols.

Bhils and the Idea of a Criminal Tribe 1 153 Equally important for an understanding of 'criminal tribe* ideology is the twenty years of chaos that preceded the British conquest of 1818.s Around 1795 three important Maratha chiefs died who had extensive estates in Central India. In each case the succession was disputed but the central Maratha government at Poona was no longer strong enough to arbitrate. Warfare raged across all of Central India. Irregularly paid mercenaries looted villages throughout Khandesh. In the second decade of the nineteenth century, supra-local govern­ ment disappeared. On the plains, there was extensive depopulation. The remaining inahabitants sought refuge under the walls of forts, which proliferated. What of the Bhils in this downward cycle? Bhils were implicated as raiders in the adjoining hills. The reasons are familiar. They include weak central government control and general animosity toward peasants on the plains. In addition, there was increased ecological pressure on the hills in the period of chaos. Bhils, who had settled on the plains as cultivating peasants or watchmen, were subject to the same periodic looting as other peasants. Many retired to the safer (but also less productive) hills areas, found leaders and raided the remaining villages.

THE BRITISH CONQUEST

The British conquest of Central India was anything but simple. The final military campaign of 1818 involved over one hundred thousand troops; they attacked various heirs, members and subsidiaries of the Maratha confederacy, while nominally supporting others. After the campaign, negotiations stabilized relations between the British and the remaining native powers. The settlement of these claims and counterclaims some­ times dragged into long court cases, but, by and large, British ‘pacifica­ tion’ techniques were successful. The policy rested on an elaborate network of subordinate treaties which guaranteed the rights of the larger Maratha houses Shinde, Holkar, Parwar. More importantly, it guaranteed the lands and rights of the subordinates of the Marathas, mostly small Rajput houses. The immediate result of these treaties was peace on the plains of Khandesh and Malwa, an area ravaged by two decades of war. The short term prospect was indirect administration, a welcome solution for an area the Bombay government had neither money nor manpower to administer directly. 5For details on this period, see chapter 5 of this volume.

154 / Marathas, Marauders, and State Formation The new government was to be strictly modeled on Maratha lines. Mountstuart Elphinstone, Commissioner of the Deccan in 1818-19, then Governor of Bombay (1819-27) defined this goal carefully: The basis of my system is that nothing ought to be altered without great and apparent necessity, at all events not until we thoroughly understand the present system, and see how it works. I am convinced that the Maratha system, if cleared of abuses and vigorously acted on, will be very well for the people, and I should dread to see judges and chaprasses almost as much as missionaries. The country ought not to be made over to the regular Government until we have determined by experience how it ought to be governed.6

The new British government was to explore the contradictions in Elphinstone’s statement for several decades, especially the contradiction between conserving the Maratha system and clearing abuses. It was to be a ‘Maratha’ system, without a functioning king, or the patronage, kinship and friendship loyalties which had formed the ever-changing factions of the previous century. From the outset it was clear that conserving the Maratha system meant retraining a whole class of Brahmin local administrators, none of whom could tolerate the Bhils. Henry Pottinger, then Commissioner of Khandesh, summarized the situation in 1822 as follows: I have found very considerable difficulty in bringing the Kumavisdars and their assistant collectors to understand the delicate nature of their interference with the Bheels. Those men had been accustomed some from long habit, and some from imbibed prejudice, to esteem the Bheels a set of desperate and unprincipled marauders, with whom it was derogatory to government, and unbecoming their own individual station, and even caste, to hold intercourse, and towards whom it was equally absurd and unnecessary to preserve any good faith beyond that dictated by the exigency of the moment.7

The Brahmin administrators naturally saw the settled, agricultural peas­ ants of the plains as subjects deserving of attention. Plains peasants after all, paid taxes, were generally predictable, and incidentally supported caste, dharma (sacred duties) and the Brahmins. No empire was founded on the Bhils; they ate meat, had no particular caste distinctions, supported no Brahmins and moved about in the most puzzling ways. The British administrators had much innate sympathy for their Brah­ 6Cited in Rustom D. Choksey, Mountstuart Elphinstone: The Indian Years, 1796-1827 (Bombay, 1971), p. 224. 7Political and Secret, vol. 66, 22 June 182S, ‘Replies to Queries regarding the Bheels furnished to the Commissioners in January, 1822’.

Bhils and the Idea of a Criminal Tribe / 155 min subordinates’ position. Across all of Central India, they had created new jurisdictions, from large raja-doms to tiny zamindaris. Their model was, of course, the landed nobility of England and their hope was that the new aristocracy would stabilize the country and make natural allies for the new British rule. These subordinate treaties unintentionally defined a ‘heart-land’, the same plateau and river valleys as Mughal and Maratha times. Defending these heartlands and the transportation lines between them made the hill people enemies and created a precarious situation. First, the British were conserving a Maratha administration which had never evolved an effective method for dealing with Bhils and other hill groups. Second, their Brahmin subordinates hated Bhils. Third, they were legally and ideologically committed to defending the valleys, so vulner­ able to attack from the hills. Finally, they were understaffed and aware that their tax base was a recently ravaged and depopulated country. BRITISH POLICY, 1820-30

British policy proceeded in four overlapping stages. Initially, the British responded to attacks on caravans and cattle-lifting with military forays into the hills. In the short run, the policy failed for a number of tactical reasons: (1) A large foray could not move quickly enough through the forested hills to catch the Bhils. (2) Small forays were often unable to cope with rapidly-assembled groups of one hundred to two hundred Bhils. (3) Pursuit was thwarted by the many boundaries with native states. (4) Troops could only be sent into the hills in dry season because the monsoons stopped all operations. (5) Diseases, especially malaria and cholera, were common in the hills, so troops could not be permanently stationed there.®

In the second stage, the new British men-on-the-spot, particularly John Briggs and Henry Pottinger, tried a different strategy. Having learned that Bhils fought under leaders termed naiks, Pottinger and Briggs sought out these naiks and tried to sign treaties with them, in order to extend the system of treaties and landed nobility from the plains into the hills. The pattern was familiar. The leader would guarantee the conduct of himself, his family, and his followers and in return, the British would recognize rights, payable by the Bombay government from general revenue.9 ' As this policy failed, it occupied much time when the Bombay Council sat in Political and Secret sessions. See, for example, Political and Secret, vol. 51,1 December 1819; vol. 62, 10 May 1820.

156 / Marathas, Marauders, and State Formation The plan was probably doomed from the start. Neither the'British nor the Brahmin subordinates had lived among the Bhils, and they had no idea of the Bhil social structure. At Pottinger reported: ‘It is very difficult to define what the precise powers of the Bheel naiques are, because the shortsightedness of the Maratha authorities never led them to consider the subject as deserving a thought, and consequently we have no experience to draw conclusions from.’10Nevertheless, the officers pushed on, mak­ ing treaties where they could. In some cases it was with several brothers, in others with the wife of a deceased leader, in still others with two gene­ rations of the same family. Even Pottinger conceded that these naiks were only leaders of mercenary bands which had grown up in the preceding twenty years of anarchy. That the naiks could control their nominal fol­ lowers was little morel than a pious hope. By 1823 John Briggs in Khandesh and John Malcolm in Malwa realized that more than military tactics were wrong.11The treaty system which had brought peace to the plains did not work in the hills. The naiks had little long-term authority, losing followers as soon as they stopped raiding. The failure of both the initial military forays and the treaty system was frustrating as Pottinger’s remarks to a less than sympathetic official Bombay audience in 1822 reveals: I really doubt we had reasons to hope for greater success than has been attained: for while there probably is nowhere a people with less inducement to lead peaceable lives, and in every respect so devoid of principle, and of local or personal ties and attachments as the Bheels, neither is there in all a country of equal extent, and in such jungly and barren state, and so overrun with marauding race, where on the whole fewer excesses, many as there are, are now committed.12

It was in this atmosphere of failure and frustration that some very basic questions were asked. This questioning formed the third stage in the development of the ‘criminal tribe’ ideology. Again to cite Pottinger: [I]t makes all the difference of placing the Bheels in this discussion, either in the 9The first such treaty was in late 1819. Political and Secret, vol. 51,1 December 1819, p. 6446; Political and Secret, 31 May 1820, p. 4666. See also the essays by Robinson and Brandstadter in this volume on the use of similar tactics by the British to pacify other ‘hostile’ groups (chs. 2 and 4). 10‘Replies to Queries, 1822.’ 11Large numbers of prisoners were a continuing problem. It was expensive to hold Bhils in prison and even more expensive to transport them. Often they were released or escaped. See, Political and Secret, vol. 63,16 August 1820, pp. 6360-3; vol. 10,29 August 1821, pp. 5421-2. 12‘Replies to Queries, 1922.’

Bhils and the Idea of a Criminal Tribe / 157 situation of a deeply injured and initiated race or in that of a tribe who had by long licence to the most lawless habits rendered themselves obnoxious to the severest penalties. In the one case it would follow that we had not long enough tried to sooth their passions excited by tyranny and oppression---- In the other it may follow that we forbore too much---- 13 to settle this issue the Bombay government kept returning to four central questions: (1) Were Bhils a corporate, identifiable entity? (2) Were they, and had they always been, separate from Hindu plains society? (3) Did robbery and plunder form a regular, expected part of their yearly income? (4) Did their beliefs and myths glorify and support this robbery? All four were dubious propositions, at best. For example, in his first report on Khandesh, in 1820, Briggs doubted their separation from Hindu society. He pointed out that the Bhils were often employed as village watchmen on the plains and that many had taken to robbery only when their villages became deserted during recent wars. He also noted that Bhils apparently had no language of their own, but spoke some version of the plains languages, whether Gujerati, Hindi, or Rajasthani.14 The Bhils, however, played to some primal British fears because they lived in the forest, beyond the bounds of the law. The Bombay adminis­ tration could not fail to draw parallels to the inhabitants of the great British forests: outlaws, escaped tenants, bandits, poachers, or in older days, knights in trouble. To flee to a forest meant forfeiting all rights and property, all contracts; to be caught in the forest, without good reasons, often meant death.15As for the Bhils, their entire population lived in the forest. The fears underlie John Malcolm’s description of the Bhils. Existing as they have hitherto done, under despotic governments, which placed them beyond the pale of civil society, and which not only gave them no encouragement, or protection, but authorized the lowest of the fiscal officers to take their lives without trial, considering themselves a proscribed and con­ demned race, ignorant to a deplorable degree, believing in witchcraft, blindly obedient to the orders of their chiefs, subject to extraordinary privations, and 13Ibid. 14Captain Briggs, ‘Report on Candeish’, quoted in full in A. H. A. Simcox,i4Memo/ro/ the Khandesh Bhil Corps, 1825-1891 (Bombay, 1912). 15J.J. Jusserand, English Wayfaring Life in the Middle Ages, 2nd ed. (London, 1920), pp. 141-5.

1581Marathas, Marauders, and State Formation constantly exposed to danger from their fellow creatures, and from the ferocity of wild beasts, with whom they share the forests, the Bhils have, in consequence, became the enemies of order and peace. They have cherished predatory habits as a means of subsistence; and receiving no mercy or consideration, they have sought, from natural and simple impulse, to revenge the wrongs they have sustained. Time has interwoven their habits of life, and feelings, with their superstitions, until they actually believe that they were created to prey upon their neighbors.16

In the minds of the British administrators, if the Bhils continued to resist and obstruct the civilization and benefits of British rule, then something was seriously wrong with them. Their thinking followed a process which sociologists term labelling. ‘The question is raised. “What kind of a person would break such an important rule”. And the answer is given, “One who is different from the rest of us, who cannot or will not act as a moral human being and therefore might break other important rules.” M7 The Bhils’ recalcitrance, refusal to sign or honor treaties and reluctance to stop raids on vulnerable plains villages was generalized to imply total moral degradation. Some remarkable flights of conjectural history infused the building of ‘criminal tribe’ ideology. Even the normally sober John Malcolm con­ cluded: [W]hen the Brahmans had established their superior rights and privileges, numbers of inhabitants of India, who adhered to their own superstitious practices, would be deemed outcastes; and as such, would be compelled to fly to the woods and mountains, in order to find refuge from persecution and oppres­ sion, in a life of poverty and privation. Similarity of condition and care of providing for their security, would produce union among men thus situated, and they would divide, according to circumstance, into separate families and tribes, who though condemned by the higher classes of Hindus, might in the course of time be expected to amalgamate, in some degree, with those impure tribes, who, though sprung from the four privileged classes have been subsequently de­ graded, on account of their spurious birth.18

Why, then, was the Bombay government, on so little evidence, anxious to establish that Bhils were a corporate group, cut off from Hindu society, with a mythology that supported regular, organized robbery? The reason “ John Malcolm, ‘Essay on Bhils’, Transactions of the Royal Asiatic Society c f Great Britain and Ireland, 1 (1827):89. 17 Howard S. Becker, Qutsiders: Studies in the Sociology of Deviance (New York, 1963), p. 33. "Malcolm, ‘Bhils’, p. 65.

Bhils and the Idea of a Criminal Tribe /159 is quite straightforward: being sure that the Bhils were habitual, profes­ sional robbers simplified the problem for the entire concerned bureau­ cracy. Early nineteenth-century notions of crime in England defined serious crime as crime ‘committed on a regular basis by professional criminals who have no other occupation’.19 This was the crime that concerned magistrates, prison chaplains and others who gave evidence before numerous commissions and wrote in newspapers. Criminals were widely thought to be a class, even a race apart from normal society. The ferocity of punishments reflected the perceived threat to society. In the opening of the nineteenth century, there were forty offenses for which the death penalty could be invoked; transportation lengthened the list consi­ derably. The final stage in the making of a ‘criminal tribe’ was evolving a practical strategy to contain them. The British first tried well-worn tactics. With Bhils safely outside society as a ‘criminal tribe’ there could be no moral restraint on attacking and burning Bhil villages, holding a clan responsible for individual robberies and frequently sentencing them to death and transportation. But even these draconian measures failed. For ‘notwithstanding the loss by death or confinement of many of the more prominent insurgents and that great and reasonable inducement to accept terms had been held out, little sensible progress was attained’.20 In desperation, Elphinstone, the Governor of Bombay, argued for the formation of a local light infantry composed of Bhils; set a Bhil to catch a Bhil. It could not have been an easy shift in policy for Elphinstone to make because of his long-standing commitment to the conservation of the Maratha administrative system. The immediate pressure for a new plan was economic. There had been a severe famine in 1824, lowering reve­ nue; even in non-famine areas, the initial settlements proved far too high and had to be scaled down by one-third or more. With the continuing high administrative costs, Bombay could not afford a large number of regular troops in Khandesh even if they had been relatively effective against the Bhils, which, to date, they had not been.21 19David Philips, Crime and Criminality in Victorian England: The Black Country, 18351860 (London, 1977), p. 21. “ Quoted in F. J. Goldsmid, James Outram: A Biography, vol. 1 (London, 1880), p. 57. The Bombay government at this time seriously considered cutting or burning down the forest on whole hilly tracts. But tfiey ultimately rejected the idea as expensive, difficult and likely to produce irreparable damage. They also considered fiee-fire zones, marked by boundaries and announced by proclamation. Political and Secret, 30 June 1824, vol. 47, p. 2913; vol. 49, pp. 3565-6. 21There were, however, precedents for this solution, though none of my sources mention

160/Marathas, Marauders, and State Formation John Briggs, Commissioner of Khandesh, vigorously opposed the new plan. For him, Bhils were a proven criminal tribe. To train and arm them, he believed, was only to encourage more cattle-lifting and robbery when the inevitable desertions would happen. John Malcolm, the Commis­ sioner of Malwa, also expressed ambivalence about the new approach: ‘The common answer of a Bheel, when charged with theft or robbery as ‘1 am not to blame. I am Mahadeo’s thief’. In other words, my destiny as a thief has been fixed by God. It is this superstitious impression which offers a great, but not insurmountable obstacle (as it has been too rashly termed) to their reforms.’22 REDEEMABELITY

In the deliberations over the Bhil question, writ small, was one-of the major social debates of Victorian England about the redeemability of the lower classes in general and the criminal classes in particular. It was a debate on fundamental human nature and ran the gamut of Poor Law reform, the extension of the vote, prison conditions, even corporal punishment in the navy. During the first half of the nineteenth century, public opinion generally moved in the direction of redeemability. By 1850, there had been, for example, major reforms in the Poor Law and two extensions on the vote. Pamphleteers and politicians argued that with the right moral education, steady work, less drink, and firm, compassionate leadership by the upper class, anyone—no matter how lowly of birth— could be a reliable citizen. For the Bhils, too, redeemability won out. Over Briggs’s objections, Elphinstone sanctioned the Khandesh Bhil Corps. Thus, the arguments of the moral uplift for the recruits of steady work, education and contact with firm, compassionate superiors meshed nicely with the immediate British need for cheap, tough jungle trackers. Much to everyone’s surprise (probably including Elphinstone’s) the system worked. James Outram, the twenty-two year old commanding officer, first convinced five young outlaws to join him as personal bodyguards. Over the next six months, he recruited one hundred more Bhils, conquered rumors that the men were to be transported, and gradually introduced uniforms, drill and finally weapons. Within a few it. The French began using Native Americans against other Native Americans as early as 1750, a practice the British adopted as they regularly used them us guides and troops prior to the American Revolution. “ John Malcolm, A Memoirof Central India (London, 1832), p. 526.

Bhils and the Idea of a Criminal Tribe / 161 years, the Khandesh Bhil Corps proved so effective against hill groups who raided either travellers or villages that the plan was copied in the Malwa Bhil Corps and Rajasthan Bhil Corps. THE EFFECTS OF ‘CRIMINAL TRIBE’ IDEOLOGY

Consider the effects of the ‘Criminal Tribe’ ideology and some wider implications. To begin, what of the effects on the Bhils themselves? Once identified as a criminal tribe, Bhils were not generally rehired as village watchmen and fort guards when peace returned to Khandesh. Those remaining on the plains were strongly encouraged to settle and become cultivating peasants. These Bhils were settled is their own villages and subsidized loans offered for beginning agriculture. George Gibeme, then making settlements in Khandesh, visited the hamlets in 1828, . . . which I trust will add considerably to the wealth and prosperity of the country. I allude to the settlement of the Bheels under Captain Ovans, the agent. I have visited nearly all the colonies under him during my circuit, and the interest is deeply excited in how rapid an advance to civilized life is observed. A kind of village pride appears to have arisen in the minds of those longer established, the longest term embraces a period of 2 1/2 years... ,23

These were termed plains Bhils and not considered a threat. In fact, settling in agriculture on the plains was the only way a Bhil family could ‘redeem’ itself, other than by taking up service in the Bhil Corps. This pressure accentuated differences between plains Bhils and those remain­ ing in the hills. A century later in the 1960s, plains Bhils were described as indistinguishable in dress and speech from lower-caste Hindus and no longer intermarrying with hill Bhils. For those remaining in the hills, there is a cultural pride in their separation from Hindu caste; a part of their own cultural heritage includes a large measure of inter-clan violence, both cattle-lifting and wife-abducting.24 March is the season for abduct­ ing a wife if a Bhil cannot raise the necessary dowry; May through August when crops are growing and food scarce is the time of house-breaking. Many Baujis or deified folk heroes are ‘men who died in an act that is unlawful for those who do not belong to the community’.25One suspects that, like Scottish clans, these are long-standing elements of the culture, 23Bombay Revenue, 8 October 1828, ‘Jamabundy Report #4’, IOL. 24T. B. Naik, ‘The Turbulent Bhils of Alirajpur’, Journal ofthe Anthropological Society of Bombay, 12,2 (1966-7):21. 25 J. Jain, ‘Religion and “Crime” Among Bhil-meena of Rajasthan,’ WeinerVolkerkunde Mittelungen, 1971,N.F., 13:61.

162 / Marathas, Marauders, and State Formation but it is equally plausible that the Bhils have internalized the idea of being a ‘criminal tribe’. Bhils, like so many others, were affected by being defined as a group, in this case a tribe. Over the two centuries from 1600 to 1800, there had been a slow, hypergamous drift of Bhils becoming Rajputs. The effect of categorizing Bhils as a criminal tribe appears to have slowed hypergamous marriages, meanwhile encouraging Rajput origin myths. What of the effects of ‘criminal tribe’ ideology on administrative thinking? Much of the Victorian thinking on crime has been subtly incorporated into thinking about hill peoples in India, right down to the present. They are, at bottom, seen as a problem. Themes of moral uplift, technical education and prohibition of alcohol run strong in the literature. In addition, when issues of how hill peoples relate to plains peoples are discussed the terms of debate strongly recall Victorian ideas of a criminal underclass. C onclusion

The development of the idea of the ‘criminal tribe’ was the merging of two intellectual traditions, both with deep roots. First, there was the tradition of the Brahmin subordinates of the new British rule. Theirs was the plainsmen’s fear of the forest, the cultivators’ fear of hunting-andgathering peoples, the high-castes’ fear of people without caste, the Hindus’ fear of non-Hindus and the bureaucrats’ fear of an uncontrollable population. To this was added the strangely parallel British tradition; it included a long legal association of migrating with ‘vagabondage’ and the association of forests with crime and outlaws. To this mix were added the more recent ideas of criminals as a race apart, and finally, in line with nineteenth-century ideas of progress, the idea of redeemability. This entire heritage became crystallized and ‘institutionalized’ in the criminal tribe laws of nineteenth-century India.

Chapter Eight

Burhanpur: Entrepot and Hinterland, 1650-1750

The consideration of the pre-colonial city and its hinterland have important consequences for several ongoing debates about the nature of the pre-colonial state, the changes brought by colonialism, and the kinds of structures that survived from the pre-colonial to the colonial eras. There is, for example, the important debate over the amount of central­ ization and authority in pre-colonial states. In India, recent research has stressed the weak central organization of pre-colonial states, with many competing factions and multiple power centres. If this was true, how did the supply network of major cities cut across these fragmented jurisdic­ tions? What were the expected roles played by government, in trade? How city-based were governments? A second lively debate of the last decade surrounds the dependency theory. One core assumption in the dependency theory is that pre-colonial agriculture was overwhelmingly subsistence and that cash cropping was an isolated phenomenon until the advent of world markets and European colonialism. The debate will focus our attention on the cash crops and manufactured items of the hinterland. A third debate concerns pre-colonial entrepreneurship, both its nature and the effects of transition to colonial rule. This debate will direct our atten­ tion to who were the traders, and what were their networks. Let us ground these broad questions in the specifics of Burhanpur, the great pre-colonial entrepot of central India. The early history of Burhanpur is relatively simple to summarize. It was founded around ad 1400, on the banks of the Tapti River (in current-day Madhya Pradesh), by Nasir Khan, founder of the Faruqi dynasty of Khandesh. It served as the capital of this compact Islamic kingdom for the next two hundred years, until conquered by Akbar, in 1600. Thereafter, Burhanpur was the headquarters of all Mughal operations in the South. It was, throughout the seventeenth century, headed by either a major member of the royal Mughal house or a noble closely related by marriage. Let us begin in the centre of Burhanpur, in the late seventeenth century

1641Marathas, Marauders, and State Formation and look through the eyes of Thevenot, the French traveller. Though this quote is long, it gives an extraordinarily vivid picture of the city: Brampour 4s a great town standing upon very uneven ground; there are some streets very high, and others again so low, that they look like ditches when one is in the higher streets___The houses are not at all handsome because most of them are only built of earth; however, they are covered with varnished tiles, and the various colors of the roofs, mingling with the verdure of a great many trees of different kinds, planted on all hands, makes the prospect pleasant enough. There are two caravanserais in it, one appointed for lodging strangers and the other for keeping the kings money, which the treasurers receive from the pro­ vince; that for the strangers is far more spacious than the other, it is square and both of them front towards the meidan. That is a very large place, for it is at least five hundred paces long, and three hundred and fifty broad; but it is not pleasant, because it is full of ugly huts, where the fruiterers sell their fruit and herbs. The entry into the castle is from the meidan, and the chief gate is betwixst two large towers; the walls of it are six or seven fathom high; they have battlements all round, and at certain intervals there are large round towers which jut a great way out, and are about thirty paces diametre. The casde contains the kings palace, and there is no entering it without permission; the Tapty running by the east side of that town, there is one whole front of the castle upon the riverside, and in that part of it the walls are a full eight fathoms high, because there are pretty near galleries on the top, where the king (when he is at Brampour) comes to look about him, and to see the fighting of elephants, which is commonly in the middle of the river---They drink not commonly the Tapty water at Brampour, because it is very brackish: but they are supplied from a large square bason (that is in the meidan) the water thereof comes from a distant spring, and before it fill that bason, passes by the carvansera for strangers which it furnishes.1

We can supplement Thevenot’s observations of the inner city of Burhanpur with a number of other sources. Nicolo Manucci, an Italian traveller of the same period, estimated the permanent garrison at 6000 men.2 The city contained a number of famous shrines, such as the tomb of Adil-Shah al-Faruki, the founder of the Adil Shahi dynasty. Though there were undoubtedly many others, the main mosque was the large Jama Masjid (pictured in a charming nineteen-century drawing by Meadows Taylor).3 'Thevenot, Travels in the Indies, in, London, 1687, p. 71. Reprinted from Indian Economic and Social History Review, 25,4 (1988). 1Nicolao Manueei, Storia Do Mogor, Calcutta, 1966, n, p. 389. 3An etching of which is in possession of the author.

Burhanpur: Entrepot and Hinterland /165 The normative documents which the Mughals surrendered to the Marathas in the 1750s also supplement Tavemier’s observations. They detail, for example, the activities of the Mughal mint in Burhanpur.4 These documents also give rates of taxation for the other markets of the city. There was the horse and cattle market: the large Mughal armies needed animals. There was a vegetable market, and a separate pan and spices market. Also, there was a jewellery market, where the products sold included brocaded cloth, and a tobacco bazaar. There was a separate area for leather workers and dealers, and a money lenders market.3 If Burhanpur was like Delhi or Agra of the period, and it sounds similar in every way, the spatial layout of the city would have included several large households of nobles, complete with kitchens, stables and work­ shops. The strong, regular city wall enclosed an area of perhaps three square miles, stretching nearly two miles along theTapti’s northern edge; there were nine large gateways leading to extensive pleasure gardens beyond the city walls, which also grew food for the noble households.6In the eighteenth century, when the city was, if anything, smaller than it was in the seventeenth century, a British observer estimated that the gardens and suburbs stretched ten miles along the river, mainly to the west of the city.7 Three roads led out of Burhanpur, each defining and passing through a hinterland. First, there was the road to the north that connected the city with Agra and Delhi. If we were to follow this northern road, much of the first day’s march would be up toward the base of Asir fort, which dominates the northern view from the city. Asir hill is a steep, truncated cone which rises 900 feet from the valley floor and dominates the surrounding hills. The rocky scarp rises sharply 100 to 120 feet, virtually all around the fort. History bears out its strength. In over 700 years as a fortified place, Asir was taken militarily only once before the British assault in 1818. The fort and the city were intimately associated. The fort garrison forayed out to put down local disturbances, and perhaps escorted important travellers. One early European traveller, John Ogilvy, asserted 4Cf. Pune Daftar, Peshwa Khandesh Rumals, 203, an account ofthe mahals ofBurhanpur, 1764. 5Pune Daftar, Khandesh Rumals, 196, an account in the name of Naro Krishna, subadar. 6Cf. the early nineteenth-century map, India Office Library, Map G, vn, 3, ‘Aseergurh and surrounding country’. ''British Museum. Add. 29213, ‘A Journal of the road travelled by Colonel Upton from Kalpi through Narwar and Bhopaul and Boorhaunpore . , r bigari 103 bigha 48 Bihar 184,191

218 ! Index Bijai Ram 73-4 Bijapur 172, 183,193-5,197,198, 204, 206-7 Bikaner 191 Bodvad 127 Bohra 50,175-7 Bombay 24,61,178 Bomar 137 Briggs, John 115,117-18,125,152, 156-7,160 Bruce, George 6 Bundelas 64 Bundelkhand 11,101 Bundi 11,13 Burling, Robbins 71-2 Bythell, Lt. Col. R. 88 cakar 186 calicuts 166 canals 132-3

dastur-amal 139-40,142 dasum 123 Daulatabad 92,96,166,193 De Boigne, Comte 14,18 dehezada 48 Delhi 24,165 Deolia 36 Deshasta Brahmin 80 deshmukh 104-5, 135,139,142,1445,149,195-6,198 deshpande 149 Devipura 42 Dewas 36,62 Dhar 32, 36,52, 53,62 Dhangar 193 Dharangaon 171 Dhulet 26 Dhulia 109 Digthan 36 diwan 72,125

cauveri 199

diwani 175

chah 133 Chakradhara 193 Chambal 11,13,75,101,168 Chanderi 36,41-2,48,50-1,53,61, 166, 168 Chandrabhan Rajput 26 Chandsur 137 Chattarpur 177 chaudhri 26, 35,44, 50, 52,60, 109, 128,135 Chaudhuri, K. N. 167 chauth 23, 28,32, 35, 88, 102, 136, 173 Cheros 184 Chitpavan Brahmin 43, 80, 177 Chopda 145 Chota Khan 14 chotrijatra 93 Chotta Khan 78, 81

Doab, Ganges-Jumna 171 DoreGujar 145 Dravidians 151 Duraha 42,66

Dabhabe, Khanderao 136,172 dakli 128 Dakni 183,193 DaraSikoh 187 darbar 70 Das Gupta, Ashin 175,178 Dasahra 10

Eliphinstone, Mountstuart 112-14, 117,154,159 Erandol 102, 107, 109 Faiz Muhammad 73-4 Famavis.Nana 14 Faruki dynasty 133,163 Faruki, Musnad Khan 110 Faruki, Nasir Khan 89 Fategarh 67 faujdar 29,48, 50, 89,128 Fox, Richard 123,146 Gaekwad, Damaji 110-11,138 gaidis 20 Garratt, G. T. 23 gavgana 42 Giberne, George 146,161 Giradur Bahadur 30 Goa 201—2 Godavari 199 Gohad 75

219 / Index Golconda 172, 206 Gond 78 Gondwana 11,42,184,208 Gosains 177 Govind, Rago 39 Gujcrat 23, 36,101,117, 151,170, 178,183-4 Gujars 183 gur 10 Gwalior 61,75 haji 168 Hakaiat Hindustan 123 Handia 111 Hariyana 5 havaldar 85,86,91,95 Hervey, Charles 7, 8 Holkar, Ahilyabai 13-14,19,69, 74-81,179-80 Holkar, Jaswant Rao 22, 114, 117,153 Holkar, Malhar Rao 13,15-16,36-7, 39, 54,62,75-6, 179 Holkar, Tukoji 13-14,77, 81 holi 93 horse grading 187 Hoshangabad 66 hundi 176-7 Hunter, William 175-6,179 Husein Ali Khan 136,172 Hyderabad 10, 23,64, 80 inam 44,50 Indore 30, 32,36,61,167,178,181 Innayat Jung collection 125 Islamnagar 66,68 istawa makta 52-4,101 izafa jagir 44, 50, 73, 81,89,96,135,143 Jahangir 26, 189-91 jain 177 jama 54,128,142 jamabandi 48 Jamav Daftar 44 Jamner 145 Janardan, Trimbuk 39 Jaora 36 jauhar 192

Jat 13,64,79,180 Jhabua 26, 30,36,191 Jhala, Zalim Singh 13 Jhansi 177 Jilkat 45 Jodhpur 191 jowar 105 Jumna 11 Kabul 191 Kakatiya 199,202 Kalabag 36,39,40 Kaladhon 196 Kallars 203 Kalyan [Kaliani] 206 kaul 146 kamavisdar 30,39,40 Kanada 193 Kanauj 166,186 Kantul 11 Karkhanis 86 Kamatak 135,143 karzapati 41 Kesari Brahmi 50 Ketelaars, J. J. 170,172 khalsa 44 khandani 38-9,41-2,44,54 Khilji, Ala-ud-din 88 kharif crop 12,48-9 Kichi Rajputs 51, 170 kiledar 89 kistbundi 103 Kohaj 94 Konkan 43,51 Kotah 11, 13,15, 117 Kotwal 48, 50 Kshatriyas 180, 203 Kulkami 196 Kunbi 145, 149,193 Kurwai 42 Kusumbe 102 Kutavad 42 LakhaNaik 26 Leling 107,138 Lohar 193 Lohara 137 Lucknow 167,178

220 / Index lun 187 McLeod, Lt. John 148 mada-i-maash 26 Madhya Pradesh 88 Madras 28 Madurai 202-3 Mahadev, Ramchandra 105 Mahadev, Vishnu 50 Mahadurga famine 196 mahal mazkur 56 Mahalaxmi 93 mahalwari 99 Mahanadi 199 Mahanavami 201,203 Mahar 148 Mahi River 5 Mahuli 94 Malcolm, John 22, 115, 117, 157-8, 160 Malet, Charles 103, 105-7 Malgav 109,197 Malhar, Trimbuk Rao 52 Malhotra, O. P. 67 Mali 145 Malsure, Tanaji 82 Malwa Bhil Corps 161 Mane 196 Manucci, Nicolo 164 Mamola Bibi 72, 74,78, 80 Mandsur 36 Mandu 166—8 Mangalvede 196 Manikpunj 108,139 mansabdar 52,66,69,189, 195, 200, 204 Maratha succession rules 71 Maravars 203 markets in Burhanpur 165 MarurBaji 82 Marwar 13 mazuri 128 Mehwati 5,180 Meos 5,184 Metcalfe, Charles 99,119-20 Mewar 11,13 Mhasvad 196 Minas 184

Mints 174 mirbakshi 184 mirasi 116,118 Mirza Nathan 192 Mirzapur 177 Mocha 168 Modi script 90 Mohurram 45 mokassa 36,197 More, Santaji 30 Moreshwar, Bapuji 39 Muhammad Shah 136, 173 Munda 151 muqqrarhasil 128 Murshid Quli Khan 64 Mushira 147 Muzumdar 39,43,46, 137 Mysore 10, 23,150 nad 123 nadidev Nadir Shah 13 nadu 203 Nagar Brahmin 177 Nagpur 150,178 naib subahdari 13, 37 Naik, Mahadev 142 Naik, Vasudev 142 Nandlal Mandloi 32 Nandurbar 107 Narmada River 20,94, 101, 105, 114, 166, 193 Narwar 51,72 NasirKhan 163 nawab 66 nayak [naik] 155-6,183,199-205, 207 nazar 41, 103, 189 Neale, Walter 5 Nilkanth, Ramchandra 83, 96 Nimar 111 Nimbait 102,109 Niyogi Brahmins 202 Nizam (of Hyderabad) 14,19,29, 32, 37, 64,66-9, 71-2, 80, 88,96-7, 108, 110,117, 138 Nuiz 201 Ogilvy, John 165

221 ! Index opium 167-73 Ormuz 201 omis 167 Ovans, Captain 161 Outram, James 160 Pachore 137 Paduvedu 200 paikusta 116 Pali 104 Panipat 13, 20 Pathan 98-99,104 patil [potail] 22,46,48, 103,195-6 patkari 133 Patna 175 patwari 48,128,135 Pawar 13, 15,32, 36-7, 54,62, 153 Peshwa Daftar [Poona Daftar] 49, 51, 61, 88, 90, 95,125, 179 peth 179 pindari 13, 20, 22, 79, 98, 106, 114-15 Piploda 36 pir 97 poljatra 93 Poona [Pune] 13-14,24,53,61,92, 144, 147, 153, 166,181 Poor Laws 160 Portuguese 188,202 Pottinger, Henry 154,156 Pudokatti 208 Punjab 23, 64 Purandar 82 Puma 127-8,132 pyade 39 qanungo 36,48-50, 52-3, 60 qasbah 125, 141, 189 qazi 96 QutbShah 143 rabicrop 12,48-9 RaghobaRao 13 Ragugarh 106-7 Rahatgarh 73 Raipur 94 Raisen 166 rajamandal 44,49 Raja Ram 28,83, 135, 172

rajaputra 184 Rajasthan Bhil Corps 161 rakhwala 39-40 Ram Das 82 Ramgarh 26 Rampur 105 raqaba 128-9 rasad 40,44-6,53^, 56,62 RathBagur 11 Ratlam 36 Raver 145 Reddi 203 Reinhard, Walter 76 Reven Kunbi 145 Rewa 112 Reynolds, Captain Charles 105-7 Richards, J. F. 69 Roab 77 Rohilkhand 13 Rohillas 14 Roy, S. N. 14 Rozkird 95,103 Russell, R. V. 2 ryots 46 sabnis 86,90-1,95 SadaRam 50 Sailana 30, 191 SalabatJung 140 Salbai, treaty of 14 Sambhaji 83 Samru, Begum 74, 76-7, 80 Samugarh 187 sanad 28, 37,46, 50,59-60,66-7, 69, 72, 78-t80, 191 sankrant 93 saranjam 36 sardeshmukhi 28, 35, 138 sarkango 112 sarkar 27-8, 30,35,50, 50,103,166 sarmandloi 112 Sarungpur 36,61-2, 166 Satara 196 Saugor 66 Sehore 41-2,66 Sen, S. N. 9 Sendume 109 serai 168

222/ Index Seth,Jagat 64 Sevappa 200 ShahAlam 76 ShahJahan 27 Shahabad 42 Shahu 29,32 Shaista Khan 20S Shajapur 61,66,166,178 Shankar, Moro 39 Shelke, Sultanji 32 Sherwood, Dr 6 Shinde, Bikaji 32 Shinde, Mahadji 13-15,18-19, 178-80 Shinde, Nimaji 135,172 Shinde, Ranoji 37, 39,54,62,70,153 Shivaji 28, 84,88, 95,194,205-7 Sikhs 64,99 silledar 20 Singh, Balwant 106 Singh, Hari 180 Singh, Jai 13 Singh, Jaswant 26 Singh, Karan 190 Singh, Keshodas 26-7 Singh, Maha [Shekhawat] 206 Singh, Pratap Singh, Raja Man 191 Singh, Ram 26, 206 Singh, Tej 206 Singh, Zalim 13, 19 Singhad 82,92,166 Sinh, Raghubir 12 Sironj 44,48,50,53,61,168,170, 177-8,180 sirpau 104,194 Sitamau 65,191 Skinner, Col. 76 Sleeman, James 2,4 Sleeman, William 6,8,22,76,78 Songir 110 Spear, Percival 1, 3 suba 123 subahdar 35,60 Sujalpur 42,51 sukwasi 116 Sultanpur 107 Sunel 36

Surat 24,61,168, 170-1,178, 205 swar 39,66 Sykes, William 118 tacavi 51,142,149 tahsil 146 Tal 36 taluka 26,44,52 Tamil 204 Tamilnadu 203 tampera 46 tantrism 10 tappas 122 Tapti River 88,116,127,132,163, 164-5, 170-1, 205 Tara Bai 74, 80,83 Tavemier, Jean Baptiste 166,177 Tellegu 193,199,202, 204 teriz 142 Thagi and Dacoity Department 7-9 Thakar 193 Thai 118 Thana Naik of Thandla 26 Thanjur 200, 203 Thevenot 164 Thomas, George 70,79 Thompson, Edward 23 Thornton, Edward 8 Tipu Sultan 19 titimmah 128 Trimbuk, Hari 46 Trimbuk, Krishnaji 50 Tugluqs 122 Tulsipuja 93 Thungabhadra River 183,199 Udgir 90 uftada 128,130 Ujjain 29-30, 32, 36,166,176, 178, 180, 181 Upton, Col. 88 Utran 108,139 Vaishya 203 Vakil 18,35 Vajragad 82 Varangav 145 Vijayadasmi 93

223 / Index Vijayanagar 199-204 Vindhyas 11. 24. 166, 170 Warangal 202 wazir 73 watan 105, 111 ‘weak candidate’ strategy 72 well irrigation 135 Wright, Caleb 8

Yar Muhammad Khan 39, 41, 66, 71-3, 80 Yaval 137, 145 zabt 129 zakat 46, 179 zaminzada 146 zhadti 49, 90 Ziegler, Norman 145, 189