Some Thoughts on the Representation and Misrepresentation of the Colonial South Asian Labour Diaspora

This paper explores the representation and misrepresentation of indentured Indian labour migration in the colonial archive and subsequent historiography, and offers some suggestions for alternate ways of viewing the origins of the South Asian labour diaspora. Drawing on ongoing research, it emphasises the need to move away from well-worn dichotomies of ‘slavery’ and ‘free labour’ and instead to re-centre the role of human agency and creativity, even in the face of structural opposition. By reflecting on the limitations of the colonial archive, and emphasising the role of pre-existing networks, returnees and recruiters in facilitating and informing patterns of migration, it suggests ways to move beyond the persistent popular misrepresentations of this process, and find alternative, subaltern perspectives on the indentured labour experience.

Criticised by British abolitionists, Indian nationalists and subsequent historians as a 'new system of slavery' that signified the 'international shame of the Indian', the migration of up to two million Indians as indentured labourers, or 'coolies', from the 1830s to 1920s remains a controversial and emotive theme in histories of labour mobility, diaspora and globalisation. Whilst the study of race and migration is a vibrant and fast developing field, existing works on the Indian Diaspora have tended to homogenise the lives of migrant labourers, ignoring the vast diversity in experiences over space and time. They have also tended to ignore the existence of different 'categories' of nineteenth-century Indian migrant, the communities and networks they established, and the interaction of indentured labourers with them. Beginning with Tinker in 1974, historians and social scientists have instead assumed that indenture was essentially a new form of slavery. 1 Such assumptions often reflect an uncritical reproduction of nineteenth-and early twentieth-century colonial and nationalist discourse in a contemporary form, however, rather than providing a careful examination of the many different experiences of migrants from a subaltern perspective. Conversely, those historians who suggest that emigration was a result of rational, informed and deliberate choices on the part of migrant Indian labourers, which led to considerable improvement in their living conditions and could be equated with free white labour, have perhaps been too dismissive of the coercive elements of indenture, leaving them open to accusations of imperial apologia, or failing to acknowledge the limitations that Indian conditions placed on 'free' labour migration. 2 Recent studies of the colonial discourse on slavery within India have highlighted how problematic post-Enlightenment ideas of 'freedom' and 'unfreedom' were in the Indian context. 3 Misrepresentations produced as part of a polarised nineteenth-century debate for and against indenture have persisted, however, despite the fact that they arguably tell us more about the subjectivity of the colonial observer than of the subaltern migrant. These representations are heavily influenced by past prejudices, by present day methodological approaches, and by biases and omissions within the colonial archive. This can often lead to approaches that focus primarily on the systemic functioning of indenture and its more coercive elements. Yet as a humanitarian discipline historical research should also explore how individual ingenuity can overcome structural opposition. It is partly for these reasons that the linguistic turn in historiography has given way to a 'cultural' or 'practical' turn in recent years. 4 We do not need to deny the importance of economic and material structures in order to assert the role of culture and of the individual. Theoretically they can co-exist at different levels of analysis. This has come to be widely accepted in the aftermath of the subaltern project, and the role of ideology, culture, emotion, and of the individual has assumed renewed importance in contemporary historical research. These approaches are vital for historians seeking ways to escape from the 'Tinkerian paradigm' that has hitherto dominated the historiography of Indian labour migration and who desire instead to focus on the crucial issue of the subaltern's own motivations, experiences and outcomes.
The study of subaltern networks and individual agency has been crucial within the work of the 'Becoming Coolies' research project on Indian Ocean labour migration. 5 This project contextualises Indian Ocean indentured labour by placing it within a longer history of Indian labour mobility. It explores how indentured migrants of the mid-nineteenth century drew on pre-existing patterns of migration and contributed towards the development of networks that would inform the choices of later migrants. In doing so, it seeks to uncover diverse individual and collective experiences within often complex patterns of migration and remigration. Micro-histories and prosopographies lie among the available tools in this endeavour, as well as the use of oral histories and local archives that lie as close as possible to migrant destinations in and around the Indian Ocean.
A critical and properly contextual approach towards the use of colonial archives is key to understanding the insights they offer about indenture. Colonialism and the 'colonial state' are sometimes represented as unitary entities, but there were marked differences in the policies towards and treatment of migrants in each of the colonies, as well as fierce competition for labour between different colonial governments. These endeavoured to direct Indian overseas labour migration towards specific plantations, whether they were in Malaysia, or Sri Lanka, or South Africa, for sugar, tea, coffee, or rubber production. There was endless discussion of the merits and demerits of the various indenture schemes (usually in overtly racist terms), and a desperate attempt to exercise control both in response to the demands of planters and the various critics of these schemes. However, there is also clear evidence of individuals using their wit and skills to evade those controls and to pursue their lives outwith the system of colonial regulation. Unfortunately, these are often obscured by a focus on administrative priorities, management, and rules and regulationswhen they worked and when they did not work. Over-reliance on metropolitan archives and administrative priorities means that we commonly lose sight of the exceptional and creative groups and individuals who found the means to resist structural violence and contribute towards the building of the new societies that emerged in the multiple destinations of Indian overseas migrants. Such perspectives can be challenge by a focus on the migrants' own perspectives and experiences. Letters written by both male and female migrants, or on their behalf (in exchange for a small payment), which have been preserved in several colonial archives are one of the several resources available in this endeavour. These letters were often undelivered because they were misaddressed, whereupon they were held by the Protector of Immigrants in the colony or would be sent to the dead letter department of the Calcutta Post Office in India. From here they could be collected upon payment of a fee. These letters would be kept in case of collection and some are still preserved to this day. They form one of the most important sources for learning about the affective or emotional relationships between migrants, their families, and their peers, providing insights into the networks they used in order to sustain themselves. 6 Human agency is paramount in the new approach to the history of Indian overseas labour migration, which explores how individuals made new lives for themselves by adapting and creating relationships beyond the frameworks provided by religious and governmental institutions and laws. Studies of convict labour migration have already shown how Indians even in the most constrained of circumstances could be far more than mere 'passive victims'. 7 It is similarly becoming apparent that, contrary its representation in popular histories, indentured labour migration too may have allowed greater opportunities for agency than has previously been acknowledged. This was even more so for the millions of contract labourers and free 'passenger' Indians who migrated overseas in the colonial period. The challenge is not simply to re-tell this history of migration as a succession of stories of individual heroisms (although such heroism existed in abundance), but to seek out the examples of the creative practice of the subaltern that conflicted with the structure, expectations, and agenda of the imperial labour regime. 8 In most attempts to generalise about the indenture system, it is depicted as a purely exploitative political economic structure: designed to take the maximum amount of work for the minimum amount of wages from the lowest sections of society. Indian migrants are portrayed as cruelly manipulated and the ending of indentured migration by legislation in the imperial legislative council, at the instigation of Gandhi and Gokhale, is touted as a simple unqualified good. This simple narrative is widely upheld in ex-sugar colonies, where Hindu revivalism and the contemporary politics of identity formation have impinged powerfully upon the writing of history. However, the legislation of 1917 only concerned itself with Indians migrating overseas as indentured workers. Those migrating within India were not considered, and Indians continued to migrate overseas as contract labourers even after the abolition of indenture in 1920, just as they had before the regime of indentured migration was invented. The traditional narrative of indenture is clearly therefore not the whole story. This paper provides an overview of some of the main contours of these complex and diverse patterns of migration, before going on to look in more detail at some of the myths and misrepresentations that have emerged regarding them. The final section explores some alternative ways in which we might think about indentured labour recruitment.

Contours of the Indian Ocean labour diaspora
Although indentured labour migration has garnered the most attention from historians, throughout the colonial period many more Indians migrated without ever signing indentured contracts. They could thus benefit from a relatively greater freedom of movement and degree of choice amongst their employers. Colonial Malaya, for example, saw more 'free' migrants and contract labourers arrive from India, than any other location. It has been estimated that those Indian labourers recruited to work in Malaya by kanganis, trusted workers sent by their employers back to India to engage workers from their home villages, alone totalled some 1,186,717 between 1865 and 1938. 9 Most settled in Penang and Singapore and in the rubber zone along the western coastal plain and foothills between southern Kedah and the southern tip of Johore. The largest inflow of Indian migrants of all sorts was between 1911 and 1930, when an average of 90,000 arrived each year. In Penang in 1891 Indians constituted more than 10% of the population. By 1911 Perak, Negri Sembilan and Selangor within the rubber zone had similarly large Indian populations, with the number of Indians reaching a peak of 33% of the population of Selangor in 1931. In Singapore, where large numbers of Indians were employed in diverse occupations, they consistently averaged at least one in ten of the population, rising to a total of 50,860 residents by 1931. 10 Since there were no labourers recruited at all on contracts of indenture after 1910 (who were in any case a minority of the workforce), the total numbers working in rubber plantations cannot be known for certain. However, by 1940 over two million acres were under rubber production, making Malaysia by far the world's largest producer. 11 By 1941 it was estimated that a total of 1,910,820 had come to Malaya on assisted migration schemes, whilst 811,598 had arrived entirely by their own efforts. 12 Overall, between 1860 and 1957 it has been calculated that probably as many four million Indians migrated to Malaya, although 70% of this number returned home in the same period. 13 Most of these migrants were bound by onerous debt obligations to the Indian kanganis who recruited them (as most had previously been indebted in their natal villages), but there are also numerous stories of those who were able to avoid or subvert themtheir migration and re-migration being often the first steps in this direction.
In the case of Sri Lanka, tens of thousands of Indian labourers were recruited each year by kanganis. Many were seasonal migrants from Southern India who did not stay for long. However, in 1871 the census reported a total of 123,565 labourers working on 996 (mostly coffee) plantations, of whom 115,092 were Tamil and 1,336 were described as 'Moors'. 14 This was prior to the boom in tea and rubber production that supplanted coffee (following the rust epidemic of the early 1880s). This boom quadrupled the Sri Lankan plantation acreage between 1881 and 1940, reaching a total of nearly 1,200,000 acres. It has been suggested by Patrick Peebles that commonly at least one worker was needed per acre under cultivation. However, the total numbers of the 'free' migrant labourers employed in plantation work, for which the government assumed little or no responsibility, will never be known for certain.
Since it shared a common border with India, Burma/ Myanmar experienced a large amount of circular, seasonal migration for work in paddy fields, paddy mills, and fisheries, as well as in the docks and factories of Rangoon. Some were recruited by kanganis, and were given advances, which they repaid at work. However, most signed no contracts at all, but made their own way by land, or by coastal ferries, or took the longer journey by boat across the Bay of Bengal from Tamil Nadu. In the absence of passports or visas in the nineteenth century, people could cross from one British imperial territory to another with relative ease. This was especially the case in Burma, which was part of the Indian empire. The 1931 Census, quoted in the 'Baxter' Report on Indian Immigration, commissioned by the Government of Burma, gave a figure of 1,017,825 for the total number of people of Indian origin in Burma, of whom 62% were born in India. 15 In Rangoon alone there were 212,000 Indians resident in 1931, and 280,000 by 1941, accounting for 56% of the city's population. 16 For those who did sign contracts of indenture, the historical archives reveal very little about their lives after they had served out their five-year contracts, despite the fact it appears that that many 'time-expired' workers (as they were called) chose to re-indenture. Thus the estimates vary hugely, but by the end of the colonial period, probably close to four million Indians had become permanent settlers in the colonial societies of Fiji, Mauritius, the Caribbean, South and East Africa, Malaysia, Singapore, and Burma/Myanmar. There they established new lives for themselves and integrated into their host societies to a greater or lesser extent. Outcomes could be markedly different depending on time and place. Some diasporic Indian communities became among the most prosperous sections of society, as is the case in South Africa, Guyana or Trinidad, where they contributed enormously to the building of modern economies. However, in other places, they were significantly marginalised and struggled to establish a foothold outside of the plantation, for example in Sri Lanka, Assam, and Myanmar. In Myanmar the impact of the depression of the 1930s led to a wave of antiimmigrant riots and many Indians were sent home and were not welcome to return. There were also powerful anti-Indian political movements in Malaysia and Sri Lanka following independence, which led to massive out-migrations from these countries as well. The same happened in East Africa in the 1970s and Fiji in the 1980s. 17 Despite this, Hindu migrants from India remain an important part of Myanmar society today. In Yangon, there are almost as many Hindu temples as there are Buddhist temples and local people visit both. The two religions have merged, and in the present day it is difficult to distinguish the religious and social practices of citizens of Burmese origin from those of Hindu Indian origin. Muslim migrants though still experience difficulties, especially the indigenous Rohingya community, who are felt to be culturally different and are frequently persecuted. It is clear therefore that the story of migration depends greatly upon its geographical context. The policies of the state have a major role to play, but so too does the creativity and ingenuity of those who migrate.

Modern myths of migration
Rethinking the nature of Indian labour migration in the colonial period raises questions about the supposed dichotomy between indenture and the so-called 'second wave' of Indian overseas migration. Ronald Takaki quotes a post-1965 Indian migrant to the USA as explaining: It is a class thing. They came from the farming, the lower class. We came from the educated middle class. We spoke English. We went to college. We were already assimilated in India, before we came here. 18 Such interpretations attempt to set modern, skilled migrants apart from their nineteenth-and early twentieth-century forebears, who are seen as being of low class and low caste statusdalits and adivasi labourers coerced or tricked by British recruiters and their Indian subordinates into migrating overseas. Contemporary migrants by contrast are described as middle-class and highly educated. Yet such clear-cut distinctions do not stand up to thorough investigation.
India has a long history of migration by people from a range of social backgrounds. People may now work in IT companies rather than sugar mills, but if examined closely there are many similarities to be found between contemporary overseas migrants and their colonial and pre-colonial counterparts.
The representation of 'first wave' migrants as primarily uneducated dalits and adivasis does not reflect the real demographics of indenture. Rather it stems from early objections to overseas migration. In the early-nineteenth century the British government in India, and most of the provincial governments, were strongly opposed to Indian overseas migration. There was a powerful British business interest in industry and trade within India, embodied in British-owned coalmines, manganese mines, jute mills, cotton mills, plantations for indigo and raw cotton production, and tea plantations in Assam. Those who profited were keen to retain the cheap labour force that drove these industries. Indeed, with the first migration of privately recruited contract workers from southern India to French colonies in the 1830s, there were protests from local British businessmen. The shipment of Indians to Mauritius and Guyana led to further criticism both in Britain and India and migration was suspended in 1838. Supporters of indenture successfully countered by insisting that they would only take unskilled and impoverished labourers overseas. 19 When indentured migration was resumed under closer supervision in 1843, following intense pressure from British planters, it thereafter became a common belief that only the poorest and least useful were being recruited for labour overseas. However (as discussed later), although migrants from the forest highlands of Chotanagpur may have predominated in the early years, within a few decades people were migrating from all sections of society.
Other common myths include the notion that most migrants were men, either forced overseas or driven by personal greed. Contrary to these assumptions, from 1842 onwards colonial regulations stated that 40% of migrants should be women. Although this figure was not often reached, usually at least 20-30% of the migrants were women and in the case of Mauritius, the quota of 40% was often exceeded. 20 By the later nineteenth-century migrants frequently travelled overseas with their entire families, determined to settle and build new lives for themselves, and the number of single female migrants increased progressively after 1900. Thus women over time made up a significant portion of the overall migrant population. The dichotomous interpretations that Indian migrants were either kidnapped or duped, or that they were profit maximising individuals can also be challenged. The reality was more complex, and often reflected creative survival strategies in the context of difficult local conditions in India. Similarly, the assumption that women who migrated were usually sex workers, or 'single, broken creatures' of 'low morality' has been strongly contested by historians in recent years. 21 One of the most common catalysts was widowhood. Others might have chosen to migrate to escape from violent or abusive relationships. A typical example of the latter phenomenon is seen in the case of Singaria, a 17-year-old woman from Mandla district in Madhya Pradesh (the former Central Provinces of colonial India), who went to Assam in 1905. She stated that she had done so in order to escape from a violent husband, who used to beat her and tie her to a charpoy and, on one occasion, 'tied her to a horse's heels'. 22 While fraud was clearly present in some cases, women were not passive victims and indeed many of them went on to become recruiters themselves. Unlike in Assam, where women were especially valued as tea pickers, they did not always work in the fields overseas. Instead, they performed ancillary work around sugar plantations and fulfilled the vital role of providing unpaid support for their labouring sons and husbands. As women were not labouring in the fields, other opportunities could open up for them. They could become entrepreneurs and even overseers. The latter were known as sirdars in Mauritius and one woman, Sukoneea, was in charge of sixteen men on the Clemencia estate in the 1870s. 23 A further assumption, common in anti-indenture literature, was that indentured migrants were plucked unawares from static rural communities. However, the Bhojpuris have a very long history of labour migration, dating back to the beginning of the Delhi sultanate, which can be traced in the poetry and the folklore of the region. 24 Dirk Kolff has shown that for many centuries migrants from UP and Bihar provided most of the soldiers for the army of the Mughals who then later on were the backbone of the Bengal army, which served the East India Company up until the revolt of 1857. 25 Aside from military service, Biharis regularly migrated across north India in search of work. A striking description of the migration out of Chota Nagpur is provided by F.B. Bradley-Birt, an assistant magistrate and later Collector assigned to Khulna, Midnapore, Hooghly and Calcutta in the late 1890s and early 1900s.
Purulia is best known as the gateway of Chota Nagpore for the traveller, the push-push and the coolie [. . .] At Tulin is the half-way rest-house between Ranchi and Purulia, where, a few miles further on, the road leaves Manbhum and enters the Ranchi district. It is down this road that many of the coolies came in the old days on their way to the tea gardens of Assam and Bengal, induced to emigrate by the hope of better wages, or by difficulties or scarcity at home, or perhaps the victims of the threats, promises, or persuasions of a recruiter who was not to be denied. Chota Nagpore is one of the great recruiting-grounds for the labour districts, and emigration has been active ever since the first occupation by the British and the increased knowledge among the coolies of the world beyond. So long ago as 1827, the magistrate of Ramgarh speaks in one of his reports of a large number of Kols annually leaving the district to work on the indigo plantations of Behar and Bengal. Ever since then a steady stream of emigrants has been flowing out of the district, mostly to the tea gardens of Sylhet, Cachar, and Assam on the one side, and Darjeeling and the Duars on the other. 26 The 1896 Indian Labour Commission describes five major 'free' migration streams: first to the Duarsarising from the Santhal Paragans, and the Ranchi and Singhbum districts in Chota Nagpur. Second, there was migration from the Eastern districts of the North-West Provinces (NWP) and Western Bihar districts to Eastern Bengal and Assam, for harvesting, road making, and railway construction: 'These expeditions are made at the commencement of the cold weather, and the men return prior to the settling in of the rains [. . .] A few men [. . .] stop in Assam throughout the rains.' Third, there was emigration from Ranchi and other parts to Calcutta to dig drains, work on gardens and other miscellaneous work. Fourth, there was emigration from the NWP and Bihar to Calcutta to work in cotton and jute mills. Fifth, there were a number of people who migrated to work on tea gardens in Sylhet of their own accord. 27 Finally, there was also migration to work in the coal and manganese mines of Chota Nagpur, which were developing by the end of the end of the nineteenth century; seasonal migration for agricultural labour at harvest time in many parts of northern and central India, and migration to work in the mammoth task of Indian railway construction. 28 All of these forms of migration, whether to work in the coal and manganese mines of Chota Nagpur, tea plantations in Upper Assam, or to work in sugar plantations overseas, were assisted in some manner with money and the help of formal and informal recruiters.
Another recurrent theme in anti-indenture literature was the mistreatment of labourers on their arrival in the destination colonies. In the early years of indenture there is little doubt that labour was hard and the treatment of workers by planters used to managing slave labour was harsh. A drawing of slaves clearing a forest in Mauritius in the early 1800s indicates the sort of work indentured migrants would have had to do upon their first arrival in Trinidad, Reunion, Martinique, Mauritius, or Guyana (see Figure 1).
In every one of the sugar colonies, vast areas of forest had to be cleared in order to make way for the plantations. Similar kinds of labour would have been carried out by a mobile labour force within India, as workers recruited from the same areas as the early indentured migrants were involved in clearing forest frontiers in the Sundarbans and elsewhere. 29 Accommodation was initially very simple, but after a few decades substantial and well-built Indian settlements began to grow. Numerous Indian shops were established too. More research is needed to provide a clear understanding of just how many ex-migrant sugar workers went on to become farmers, shopkeepers, charcoal burners, fishermen, and (in Mauritius) even plantation owners in their own right. However we know they did from family histories that are recalled in the present day. 30 In Mauritius, as the sugar industry slumped, many plantations were divided up and sold to small planters in the 'grand morcellement' of the 1880s. By 1900, if not before, Indians may have accounted for as much as one-third of the value of all property changing hands. 31 Hundreds of old migrants were also applying for free licenses to become sellers of fruit, vegetables, milk, roasted gram, pistachios, and numerous other commodities. 32 In South Africa, farming was especially crucial to exindentured labourers. There were thus 2,000 Indian market gardeners in and around Durban by 1885, and by 1910 Indians owned 10,000 acres of arable land in Natal and leased a further 42,000 acres. 33 Photographs of Indian houses and settlements and houses from the end of the nineteenth century show that those who had moved off the plantations had by this date often acquired sufficient capital to build themselves substantial properties (Figure 2). Indentured labour migrants were sometimes referred to as girmitiyas because they signed a 'girmit', which was a fixed five-year contract or agreement. The indentured labour contract is not particularly exceptional or strange. It was common in seventeenth-century Europe and North America, and it is alive and well in the present day, including in the British and Indian armies. 34 Possession of this 'girmit' set them apart from other Indian labourers, who worked purely under agreements made by word of mouth. This was to have increasing significance over time, as the indenture system became more regularised. It amounted to the first modern form of labour contract to be introduced to the Indian subcontinent, anticipated only by the contracts signed by sepoys in the Indian army (who were often recruited from the same localities). Under the various 'Masters and Servants' regulations found in the different colonies, the rights of employers and employees were far from being equal. 35 Nonetheless, the girmit evolved into a contract that involved a formal, legal, reciprocal agreement between worker and employer, actionable before a magistrate: something hitherto unseen (and still uncommon) in the Indian subcontinent. However minimal the rights it bestowed, for many this was still an advance on the customary agreements that governed their relationships with their peers and masters in the traditional rural economy.
One of the most interesting aspects of Indian overseas migration is how the complex rules and regulations established to manage the system of indentured labour, the constant criticism to which they were subjected, their frequent revision, and the resistance of workers, contributed to the evolution of the concepts of workers and human rights. In 1920, the League of Nations set up the International Labour Organization and from its advent it promoted the rights and health and safety of workers, arguing that they should be accepted internationally. For inspiration they looked to the regulations governing overseas Indian migration. The management of Indian labour migration to overseas plantations thus played a crucial role in the development of the modern concept of workers' legal rights and of human rights in general. 36 One conclusion that can be drawn from this is that Indians have not only been global citizens for far longer than we usually imagine, but have played a crucial role in the international regulation of global trade. The global-connectedness of India over a very long period, before, during, and after the colonial era, is arguably one of the reasons why India's globalisation was so successful in the 1990s. Many Indians already hadsometimes from generations backrelatives who had migrated overseas, giving them regular exchange with and knowledge of the wider world.

Rethinking patterns of labour migration
Statistical analysis can often produce insights that go against popular understandings of the past. This applies particularly in the boom years of the sugar trade in the 1850s and 1860s. By looking at the total immigration statistics immediately after 1857 (Figure 3) we can see that there is a huge leap between 1858 and 1859 and another increase in 1861 and 1865.
The conventional argument is that in this period Indians migrated overseas in order to meet the demands of the plantation owners during a period of exceptionally high sugar prices and that migration was controlled and directed by the British planters and colonial governments, but these figures suggest that Indians were also making many of the decisions. The migration of 1858-60 followed the British conquest of Awadh in Hundreds of thousands of people lost their lives and there were serious food shortages in the years that followed. In addition, a hundred thousand mutinous soldiers of the Bengal army were demobilised in 1857. This followed the disbanding of the 30,000-strong army of the Nawab of Awadh in 1855. The British ceased recruiting Biharis and people from eastern UP for the army and instead recruited Sikhs and Ghurkhas amongst others, who they believed would be more loyal. Ex-sepoys and their families found themselves destitute and looking for work, undoubtedly leading many of them to migrate overseas.
It is possible to deduce where migrants originated by examining the Emigration Passes and Ships Lists preserved in the archives of most of the destination countries. Marina Carter and I have collated the registers for a significant proportion of the ships carrying indentured labour to the sugar colonies that left Kolkata between 1858 and 1869, covering a total of about 4,500 migrants. For this period it is possible to see who every migrant was, where they came from, and their destination. With a few exceptions, they were all leaving the areas most heavily affected by the uprising of 1857. Recruiting depots were based in Chhapra and Arrah; both of these were large districts at the heart of the uprising. The most important information revealed is who these migrants were. It has often been said that indentured workers were all dalits and adivasisunskilled, illiterate people. However, after analysing the statistics of the migrants bound for Fiji, Brij V. Lal argued that those migrating came from a much wider cross-section of society, including a great many from high status and 'middling' castes. 37 We have completed a similar analysis with the registers of migrant ships leaving Calcutta. Those travelling can be grouped by social status, labelling them Brahmins and Kshatriya, Vaishya and artisan castes, 'Middling Castes', Muslims, Dalits, and Adivasis (according to definitions of the time). Interestingly, it immediately became apparent that the distribution of these different sections among the migrants very closely resembles that within the population of the NWP, Oudh, and Bengal as a whole, according to the 1881 Census of India. An almost identical cross-section of society is seen among migrants as is seen within society itself. Only 30% of the migrants on the boats were from dalit and other backward castes, which is roughly equivalent to their proportion within the population of UP and Bengal in the second half of the nineteenth century. Any variations are very small, such as 11% Brahmin and Kshahtriya among migrants compared with 12% in NWP, Oudh, and Bengal. This finding tallies with Brij V. Lal's study of the origins of Indian migrants in Fiji and the brief analysis of George A. Grierson published in 1883. 38 It is clear that the migrant population was diverse and their reasons for leaving India likely were too. The Brahmins and Kshatriya amongst their number will without a doubt have included ex-sepoys, given the time period and the places where they originated.
Mauritius was one of the most popular destinations for migrants, particularly in the first few decades of the government-regulated indentured system. This may be related to proximity as well as to the prior settlement of Indians. In addition to the eighteenth-century migration of skilled workers from southern India, some sepoys who had participated in the 1810 British conquest of the island, settled there. Even more significant, is the early recourse to returnee recruiting by Mauritius employers. People chose to go where their kinsmen were, forming migrant networks. These networks are one of the most important themes in our research and they are well documented as they were utilised by planters. As the inefficiency of using hired agents to recruit Indian labourers was discovered, plantation owners came to rely upon former indentured migrants acting as returnee recruiters. Labourers who had served out their contracts were paid to return to their home villages and bring more people back with them. Throughout the Indian Ocean this became the dominant form of recruitment. These recruiters would often return with members of their own families, sometimes bringing their entire family back to the colony at the planter's expense. Occasionally they would choose to go elsewhere. There was a very interesting amount of re-migration that went on; migrants could return from one place and then travel onwards, having heard of better opportunities in another colony. 39 The Indian kanganis, maistries, duffadars, arkatis, and sirdars who acted as intermediaries in Indian overseas labour migration, many of them former labour migrants themselves, are a key element in the story that needs to be deconstructed. Often demonised in colonial and contemporary economic and social thought, these middlemen do not fit with the Lockean, Enlightement ideal of sovereign individuals directly interacting with both employers and the state in a free economy without feudal hierarchies. British colonial reports frequently offered the opinion that the Indian peasantry would be prosperous and happy but for the moneylender. They conveniently evade the fact that the moneylenders' presence was a response to the peasants' inability to pay British taxes. In actuality, the moneylenders and other intermediaries kept society together by allowing people to meet their obligations. However, moneylenders are universally reviled in colonial sources, as are intermediaries in relation to indentured labour migration. The returnee recruiters and those labourers who migrated in gangs were considered problematic as they appeared outwith the control of the government. They had their own methods and mode of operation separate to colonial authority, but they were in fact vital.
Within the labour economy, formal regulations coexist with informal organisation. Formal regulations involve government laws and contracts but informal intermediaries play a very important role in every society. However, the jobber, or the intermediary, is commonly despised because they are seen as having suspect loyalties. This arises from their concurrent allegiance both to their employer and their responsibilities towards the labourers that they recruit. This position makes them an intrinsic part of the economy while leaving them open to criticism from all sides. The band of workers recruited by an intermediary, or sirdar, often continue to work together once they enter active employment. This is a highly typical way in which the Indian economy managed itself, and continues to do so. Arjan de Haan, writing about industry in contemporary Kolkata noted that, 'the process of recruitment has been predominantly personalistic: recruitment was and still is mainly through personal relations.' 40 This is a point worth stressing. According to the system of indentured contracts, individuals ideally should have presented themselves for employment overseas, but right from the beginning gangs of workers were organising themselves under intermediaries, applying collectively and continuing to function as groups once overseas. Networks among migrants were thus crucial to the mobilisation of labour and its management upon arrival.
Contemporary British accounts of intermediaries were invariably negative, and much subsequent historiography has followed the same path. However, Patrick Peebles in his book The Plantation Tamils of Ceylon (2001) has dismissed the idea that the intermediary was some sort of patriarch. According to Peebles the accounts 'do not reflect actual conditions, but rather planters' distaste for the independence and prosperity of the kanganis. . .'. 41 In other words, they resented their divided loyalties and their lack of control over them, even though they could not dispense with their services. In 1846, for example, the Protector of Immigrants in Mauritius commented, no immigrant ever forms an engagement for himself or even communicates with the planter, who is standing before him for the purpose of obtaining his services. They invariably and implicitly follow the will and the directions of the sirdar to whom they have given their confidence. 42 Statistics give us an insight into how central the sirdars or gang leaders were to recruiting migrants for Mauritius. In 1858, there were 810 'old emigrants' who were sent back to India to recruit, bringing 14,722 new recruits to Mauritius. In 1860, 135 went and brought back 1,811. 43 The first batch followed the uprising of 1857. Migrants were doubtless returning and pointing out the hardship that people were experiencingincluding famine and land lossand reasoning that moving to Mauritius provided an attractive escape. The organisation of recruits into gangs meant that they could compete for better wages. This was less the case upon arrival, but when they re-indentured after five years they had some bargaining power. The sirdar would take advantage of the planter's desire to retain his workforce and negotiate better hours or other benefits for his gang. Throughout the nineteenth century, British and French planters continue to complain about the practices of the intermediary (whether labelled sirdars, kanganis, or arkatis). A petition from 1847 signed by sixteen Mauritian planters claimed: 'The dominant influence of the sirdars over the bands of the labourers on the sugar estates has become [. . .] outrageous and offensive.' 44 Yet, in 1875, when a Royal Commission inquired into the treatment of Immigrants in Mauritius they concluded that, Notwithstanding the many objections that there are, both theoretically and in practice, to the Sirdar system, it is one so entirely constant with the habits and customs of the natives of India, that we feel it there would be great difficulty in breaking through it. 45 This concession is a recurring excuse for the British; they could not admit that the intermediaries were in any way ideal or beneficial, and therefore they accepted them as a part of 'Indian custom' instead.
In the case of Assam, the Assam Tea Company relied heavily for labour recruitment upon European contractors until 1915, when Act VIII made garden sirdars the only recruiting agents in the recruiting districts. Up until that time, licensed garden sirdars, equipped with loans and advances, worked alongside the European contractors. The latter were disliked by the planters (who described them as 'unscrupulous') for the very high fees they would charge per recruitsometimes as much as Rs 120-150 per labourer at the height of the tea boom: a very considerable sum at the time. 46 The European contractors themselves relied themselves heavily upon local agents or arkatis, who in turn relied upon others at village level to recruit the labourers. Contemporary British records depicted the arkatis as 'the scum of the country, ex-convicts, burglars, thieves, dacoits and notorious badmashes' and as 'a heartless scoundrel who would boast that he could by ill treatment make anyone "willing" in a few minutes to emigrate [. . .] and who was feared as much as much as a man-eating tiger'. 47 The duffadars involved in recruitment via Dhubri in Goalpura were similarly described as 'an excrescence on the "free" emigration system, who [. . .] only exists on account of the unlimited possibilities of profit in the passage of coolies from one middleman to another'. 48 Collectively, these intermediaries were described as 'blood-suckers' and blamed for any and all irregularities and deception in the process of recruitment (neatly absolving the European contractors and planters). This characterisation is repeated in later Indian historical writing. However, other sources from the same period describe the arkatis as 'belonging to the very class of people who emigrate as labourers and live and move among them'. 49 The garden sirdars who later on dominated the trade after 1915, were certainly recruited entirely from amongst the workers on the tea plantation. Recruiting was regarded as an important opportunity to earn extra money and to travel home at the expense of their employers, for which tea workers keenly competed. Once there, they used their own networks, contacts, and knowledge of the locality, in the same way as the arkatis, to bring in recruits. 50 Intermediaries occur in every society. Intermediaries, 'friends of friends' or 'go-betweens' are essential to the functioning of any economy, and often reveal the limitations of modern law. In the absence of an entirely free, transparent and accessible legal system, an intermediary is often the only way matters can progress and the parties involved can get paid. A crucial element in the relationship is trust, which has emerged recently as an important theme in development economics and economic and social history. 51 Indices of trust have been drawn up for different societies and the conclusion that a lack of trust causes backwardness has been embraced. The commonly accepted solution is to strengthen the courts and legal system, thus better ensuring that contracts are upheld. However, this assumes that the informal economy has no security of contract. It appears more plausible to argue that there are many obstacles to growth in developing economies and in most of them there is no shortage of trust, it is simply perpetuated by informal rather than formal mechanisms. As the case of indentured labour recruitment demonstrates, people relate to each other through a variety of non-state connections: kith and kin, religious links, and as well as arkatis, kanganies, and sirdars.
'Official' methods of indentured labour recruitment were specially geared to engender trust. Government messages announcing the availability of overseas employment opportunities were written in local languages and the schemes were identified as being arranged by the government or 'Sarkar'. The latter was an important factor in encouraging early migrants to place their faith in the indentured system. Dual language permits were issued, summarising what to expect as a migrant labourer. These enabled them to outline what their rights would/should be if they signed the indentured contract. They even had contracts for groups signing up together. The ideal system may have been that people were working as individuals, but the authorities came to acknowledge the reality that people were migrating as groups. All exchanges were carefully receipted and documented. In the Natal state, Mauritius and Fiji archives, the paper trails of these transactions are perfectly preserved. These sorts of records do not survive in all archives; some have been destroyed or discarded. However, in these three archives much remains, including expenses claims from returnee recruiters in order to recoup the money they had spent engaging labourers. The depot stands out as a very important place. If people turned up at the depot as an individual they would often decide which of the sirdars to follow based on their charisma, energy, and reputation. To secure the co-operation of the local authorities sirdars would carry letters identifying who they were, attesting to their good character, and stating the name of the estate owner for whom they were recruiting. An example of one such recruiter is 'Dhibby Deen', a 'Buneeya' by caste who was sent back to India in 1840 after five years, along with 400 rupees in savings, in order to recruit other workers. The document he carried from his employer Mr Barlow stated: 'The bearer of this, Dhibby Deen, is a man of good character who returns to his native country with a sum of money, his intention being to come back to Mauritius [. . .].' 52 So successful were these returnee recruiters, that in the sample year of 1857 over 90% of migrants for Mauritius from Madras were recruited by them -65% in Calcutta, and 41% in Bombay. 53 Quite apart from the official documentation assigned to generate confidence, informal trust was generated amongst the workers themselves. One of the most important services that the sirdar provided was acting as a bank: facilitating a form of saving by retaining a portion of the wages earned by the labourers. He (or she) would then return this when a labourer went home or if he became ill and his pay was docked. The sirdar could also lend money, usually at quite high interest rates, for special occasions. Most notably, the sirdar would pool all the money that was paid to his/her gang and then redistribute it at the end of the month. The cause of this practice was strict plantation regulations. These were especially fierce in Mauritius; if a labourer took one day off work without prior permission they would be docked two day's wages: the notorious 'double-cut'. If they asked for permissionperhaps they were ill or wanted to visit relativesthey would only lose the days they did not work. Therefore, the sirdar would divide the collective wages of his gang equally among the workers, regardless of unexpected time taken off, in order to mitigate or undermine what they saw as being an unjust system imposed by the planter. A letter of 1845 from a plantation owner called Rudelle in Grand Port, Mauritius describes how the sirdar pooled the wages of his workers. 54 Rudelle did not understand why this was happening, strongly disapproved of the practice, and banned it. Interestingly, he himself offered his services as a banker and claimed that he had three to four thousand pounds in Spanish gold in his hands that belonged to old migrants. This shows that migrants were able sometimes to save quite significant sums of money.
The role of the sirdar was one of constant tension. One of the main bones of contention was wages. These were initially fixed at the time of migration, but at the end of the usual five-year indenture contract they could be renegotiated if the workers wished to stay on and the sirdars often drove a hard bargain. The planters complained bitterly about this, although it was observed: 'The sirdars, if they possess an illegitimate influence, are doubtless possessed of a legitimate influence also, and one which is of material assistance to the labourer in making his contract with his employer.' 55 Once they commenced work, it was then sometimes complained that the sirdars would 'badly' advise their gangs of workers. In 1871, Magistrate Pearce cancelled the contract of two sirdars in the service of Mademoiselle Bestel in Mauritius after he heard the advice they had given to their men. They had been telling them not to work so fast, instructing them that if they made thirty cane holes they would still get their five rupees and no more if they made sixty. 56 The two sirdars also arranged a Holi celebration, which was interpreted by the plantation owner as a riot, causing them to lose their jobs. These examples add to the impression that the sirdar/ foreman/intermediary clearly intervened on behalf of the labourers. However, the successful relationships between workers and intermediaries usually escape our gaze as colonial archives mostly record only complaints from planters and workers. Workers' issues with intermediaries were only recorded when they felt the need to appeal to traditional avenues for redress after informal methods of arbitration had broken down. The archives only tell us part of the story; the part regulated by the colonial state, rather than the part where people successfully regulated themselves.

Conclusion
In conclusion, India has never been the static society often depicted in colonial sources, but possesses a long history of migration in North India and elsewhere. Indeed, India has been situated at the centre of global trade for thousands of years. 57 Contacts, trade and migration between East and West began far earlier than most of us imagine, and despite the assertions of colonial officials and planters, there is substantial evidence that high-caste Indians and skilled labourers and artisans were among them. The continuum of nineteenthcentury Indian ocean migration began with convicts and slaves (as detailed by Anderson and Allen), but went on to include many others. 58 Many in the 1860s were probably ex-sepoys. Movement from one category of labour to another would have been common: someone might work as a sepoy, then as a domestic servant, then perhaps they could become a lascar or an indentured labourer, progressing finally to farming or assisting in or keeping a shop. One of the great myths of colonial sociology was that Indians always adhered to a single occupation specific and appropriate to their caste throughout their lives: a custom that was reinforced by British administrative practices. 59 This is evidently not the case in the field of long-distance labour migration. Furthermore, in spite of popular myths, new research has revealed the important contribution of women to the building of migrant societies in far-flung colonies. Most importantly, pyramids of informal affective relationships and networks of trust can be seen to have been crucial for the success of overseas labour migration, even though colonial archives do not very effectively reveal them.
Trust and knowledge went hand in hand, and from the 1820s onwards returnee migrants were bringing back to India regular news about foreign places. The peasants of North India were no doubt informed about overseas opportunities even in this early period. The story of Indian overseas migration thus suggests that we need to decolonise Indian social and political thought and stop thinking of Indians as static, conservative people, reluctant to leave their homes. The kala pani was not so black for many millions, who found sailing to 'tapu' (as it was usually described) to be a passport to opportunity. It would be foolish to maintain that indentured labourers' experiences and agency had much effect on the larger political economy that created and defined the indentured labour system. However, it would be equally foolish to suggest that, despite the harsh conditions they endured, individuals did not succeed in extracting a measure of material benefit and personal improvement from it. Many of them returned to their places of origin with substantial savings or remained voluntarily as free workers in their destination. Some forged new careers as intermediaries of various sorts and played a crucial and complex role in the functioning of the imperial labour system. Others became farmers, shop-keepers, school teachers, and even plantation owners themselves. Their descendants went on to constitute the successful middle-class communities of Indians dispersed throughout the globe in the present day. Hopefully, by discarding colonial stereotypes and biases new research will help engender a more nuanced understanding of India's long history of migration, as has happened with the re-invention of Australian history. 60 The Indians now living in South Africa, Trinidad, Guyana, Fiji, and Malaysia, need a new story: one in which the 'stain' of indenture is expunged and their ancestors no longer depicted as mere victims, but rather as amongst the most enterprising of Indians, who made at least as great a contribution to the development of the modern world as European migrants in the same period.