Imperial competition for the Dai territories

At the time of British annexation in 1886, the Dai polities had been pressured from all sides for more than a century. The Sino-Burma wars in the late eighteenth century saw waves of Burmese soldiers moving across the Dai polities to reach the Mekong, followed by Qing troops pushing them back to the Irrawaddy. Immigration of Han Chinese populations into Yunnan formed an elite in the merchant towns in the eastern Dai polities, thus causing friction and rebellion. The Du Wenxiu or Panthay rebellion (1856–1873) took place as Qing state control in Yunnan weakened.Footnote ¹¹ In 1885, the Burmese court was unable to control any part of the Dai territories, apart from the town of Lashio and the road getting there. Eventually, the arrival of British forces brought down the kingdom of Burma. The historian Sai Aung Tun shows how this period of uncertainty also opened up divisions between different groups of allied saohpas. They fought the Burmese king, the British, or other saohpas in succession wars. British officers described this long period as a time of ‘chaos’, blaming Burma’s Kunbaung rulers and claiming that British annexation would bring law and order to the country.Footnote ¹²

The major reason for reaching China’s market on the overland routes changed in the course of the nineteenth century. In the 1860s, the purpose was still to gain entry to the communication routes and to reach the western market towns. Advocates for sending missions across Dai territories into Yunnan pointed to the shorter distance compared to using the sea routes and to less competition at the overland ports of entry. A complicated and partly successful expedition in 1869, from Bhamo in north Burma to Tengyue (Tengchong) and back again, was led by Major Edward B. Sladen (Figure 1). It was followed by a disastrous second attempt in 1875, when the British diplomate Augustus Margary was killed. The Rangoon Chamber of Commerce and Sladen himself had miscalculated the consequences of the Du Wenxiu rebellion. In 1869, Wenxiu’s governor at Tengyue protected Sladen against Han merchants who otherwise would have fought British competition on this most valuable of all routes. Sladen’s success can in fact be limited to him and his men returning to Bhamo alive. With Wenxiu’s government crushed in 1873, Margary was killed. Once British forces took control of the western Dai polities, they aimed to extract hardwood and minerals, and legislated that forests and mines were government property. A British consul was placed in Tengyue in 1897 to oversee trade with Burma and Yunnan. But the earlier enthusiasm for Yunnan’s markets dampened in favour of extracting natural resources in the Dai territories.Footnote ¹³

Figure 1.

Map of the Northern and Southern Shan States, including administrative borders.

Source: Drawn by Laurie Whiddon.

Competition for territory, natural resources, and commercial routes and markets soon intensified. In the late 1880s, French forces advanced via northern Annam towards Yunnan from their recently annexed lands in Indochina. Coastal Annam and Tonkin promised minerals and hardwood and, like the British, the French too planned for a railway line to Kunming. British and French encroachments and conquests of territories, as they sought plants and minerals for their value as commodities on global markets, can certainly be analysed as ‘commodity frontiers’ in Jason Moore’s revision of the concept. Even though beyond the scope of this article, Moore’s critical remark about the ‘slippery category’ of frontier is worth pointing out. The frontier is simultaneously a particular kind of socio-spatial movement at the edge of capital expansion (a space-of-flows) and a kind of place where uncommodified land still remains (the space-of-places). The Dai polities, with long histories of channelling high-value flows, much of which for the purpose of accumulating capital long before British and French annexation, will require yet another revision of the concept ‘commodity frontier’ to be applied to the study of this region. Here, strong Han merchant trading houses channelled tea, textiles, and other valuable goods on the mule caravan transportation routes for the international markets before the onslaught of European conquests. Such capital-driven trade also had a transformative impact on the places of production. The two European powers finally came face to face at the powerful Dai polity of Kengtung and the geopolitical border between them was agreed in 1896. The border commissions’ work was concluded along most of Yunnan’s western borders in 1899.Footnote ¹⁴

To address their lack of information about the Dai lands and polities, well-armed survey expeditions were sent to meet the saohpas in the Dai polities and to demand their submission to British rule. Borrowing from Bérénice Guyot-Réchard’s terminology, British rule in the Dai territories was that of a fair-weather affair, as they were capable of reaching these territories only in the dry season.Footnote ¹⁵ The first British headquarters were established as Fort Stedman near Yawnghwe at Lake Inle in 1887. They later moved to Taunggyi, 17 kilometres north of Fort Stedman as the bird flies.Footnote ¹⁶ When the Shan States province was bifurcated into northern and southern divisions in 1888, a northern headquarters at Lashio was established on the River Yaw 200 kilometres northeast of Mandalay (Figure 1).Footnote ¹⁷ Arthur Hedding Hildebrand was appointed superintendent of the Shan States; James George Scott was made his assistant and was promoted to superintendent of the Northern Shan States in 1891.Footnote ¹⁸ From their reports, it appears as if their methods of working were very different. They read as if Hildebrand preferred peaceful and persuasive negotiations whereas Scott spoke by the force of arms. However, it was primarily their mode of writing reports that differed. Scott, arriving in Burma as a journalist who had published short stories, expressed the drama and gallantry of his mission, while Hildebrand conveyed success, order, and the peaceful signing of treaties in agreement with the saohpas. Mistakenly, we may understand their texts as if Scott was the army officer and Hildebrand the administrator. However, despite their different narratives, they both used armed and bureaucratic violence when British governance was put in place.

To discuss encounters involving rifles and guns as armed violence is to state the obvious. A bureaucracy, in contrast, is at times described as a switchboard run by officials, working according to rules, busy redistributing revenue for the benefit of the state and people. Such a deceptively neutral representation may divert attention away from a bureaucracy’s capacity to forcibly control people and transform societies into the image of the intentions of those that organized and oversaw administrative work. In the Dai territories, bureaucratic practice often took judicial forms, which served to hide armed violence. References to laws and rules played a central part in reports from the field when officers explained the necessity of firing on villagers when carrying out administrative tasks.

There are important studies of colonial bureaucracies and authoritarian regimes that rest on the premise of coercive instruments forcefully applied by the ruling power. Thus, bureaucracy is analysed as such a tool, which makes violence work through bureaucratic practice. Step by step, an underlying structure is put in place. This infrastructure changes the conditions for the polities so fundamentally that, eventually, its effects cannot be undone. Borrowing from Tania Murray Li, it conquers and repurposes the landscape and subordinates people and law to new goals. Violence materializes, as in the case of Indonesian palm oil plantations, which Li conceptualizes as ‘infrastructural violence’. Violence is in-built and has material fixity.Footnote ¹⁹ The argument resembles Foucault’s frequently referenced study on the isolated prison world, which exercises indiscriminate discursive control through the hidden surveillance of a panopticon as a forceful instrument of disciplining power. Colonial bureaucracies are among those that at times are analysed as such authoritarian disciplining regimes. However, in our case, the conception has the weakness of presuming a fully developed institution whereas, in the Dai territories, there was a bureaucratic institution in the making and a regime that was far from being in control. Li’s point is also reminiscent of David Graeber’s work on the rule-based state that exercises power and violence through its bureaucracy. It is then seen as a form of structural violence, as can occur in the forms of race, sex, gender, class, or privilege. For Graeber, violence need not be acted out, but simply be present as a threat, so that it embodies the ‘social relations [that] the pervasive threat of violence makes possible’. For the individual, it arrives as psychological violence. His argument is helpful to our discussion. In this way, violence can be distinguished from aggression. John and Jean Comaroff’s concept of ‘lawfare’ may be borrowed from its post-colonial context to understand British colonial bureaucracy. Lawfare uses ‘its own rules … to impose a sense of order upon its subordinates by means of violence rendered legible, legal and legitimate by its own sovereign word’. As will be argued, bureaucratic practice as exercised in the Dai polities rested on ‘the violence inherent in the law, to commit acts of political coercion, even erasure’. The emphasis on practice is crucial for understanding the working of bureaucratic order in these polities. It was acted out in particular places and did not only refer to rules, it also made the rules. It is about how things were done.Footnote ²⁰

Officers like Hildebrand and Scott were vested with armed and bureaucratic force, and yet they depended on trust. They had to come to speaking terms with the leadership of the Dai polities, such as the stubbornly resisting influential saohpas of Kengtung and Mong Nai (Figure 1); if not, the bureaucratic grid being laid out would break. The colonial state rarely had the capacity simply to keep a colonized population under the boot. These were not white settler territories or regions long under colonial rule such as the Madras Presidency or parts of Bengal. The intention was that they would become directly ruled imperial provinces, and they were usually heavily understaffed. In these situations, the assumed peremptory power of the colonial state tends to be exaggerated and therefore misleading.

In British Burma, as in British India’s North-Eastern Frontier, merchants and their representatives preceded the armed and civil forces. From the Brahmaputra valley to the borders of Yunnan, commercial operations and alliances prepared the way by integrating into markets and a business society at large. In Burma, the merchants and European chambers of commerce were impatiently pushing a reluctant British Government of India to support commercial missions and improve the transport infrastructure between Burma and Yunnan.Footnote ²¹ Treaties and agreements left paw prints for administrators and army generals to follow. However, at times of military advance and conquest, even when supported by a native allied force, mutual trust was ephemeral. There were many competing interests, even within the colonial apparatus.

The request for uniformity was a constantly recurring complication for British administrators. Surveyors and revenue officers in India’s North-Eastern Frontier, who settled territorial control and assessed the value of land and produce, struggled throughout the early decades of the nineteenth century with how to apply bureaucratic control in a uniform way across enormously varied territories. The colonial administration’s conception of land revenue caused conflicts in the accounts as it was based on fixed ideas of landscapes and productive tasks, as in classifications of ownership and land types. This led to dramatic confrontations with natural realities as well as conflicts with existing regional political hierarchies and land and revenue relations.Footnote ²² The Dai region was similarly characteristic of varied natural conditions, multi-layered social and political landscapes, mobile populations, and unstable and permeable borders. Competing for control against two other empires added extra strain and urgency. This colonial frontier was a moving target.Footnote ²³

In the formation of governance, armed and bureaucratic violence were part of the same force of subjugation and operated by officers who held a gun in one hand and a pen in the other. Missions comprising mostly military officers and troops, who were ordered to fix the amount of annual tribute, could end by burning the village that had been targeted for the tribute. In this way, British operations fuelled the turbulent situation in the years following British annexation, when violent conflicts escalated within the Dai territories, involving troops under European, Chinese, and Dai rulers. Yet interest, ambition, and purpose separated the many actors, and between groups the odds were extremely uneven. With such a pattern, it is unproductive to disconnect the two forms of operation as one armed and violent, and the other civil, bureaucratic, and peaceful. Here, bureaucracy was a violent force per se.

Armed violence

Open conflicts followed the death of King Mindon and the succession of Thibaw in 1878. There were those in the Burmese elite who wanted to reform the state and those who wanted the kingdom to secure the old order with its many privileges—what the historian Thant Myint-U calls a conflict between reformists and royalists.Footnote ²⁴ Throughout Thibaw’s reign, the European merchant community exerted external pressures on the kingdom’s economy and the British Resident in Burma acted in domestic disputes to secure British interests. All in all, the economy staggered under aggressive British competition within the external and inner markets. Poor cotton harvests and falling world prices pulled down cotton and silk manufactures, and cheap British products outrivalled Burmese piece-goods. The oil industry declined, and financial speculation affected trade and credit. Myint-U concludes that the situation reflected the weakness of the Burmese state. While gaining immediate access to the Dai territories after the fall of Thibaw, the British also faced social upheaval and civil war in the region—an anarchic warlike situation west of the Salween River in Aung Tun’s words. Strong groups of allied saohpas formed. One grew out of the Mong Nai saohpa’s conflict with Burma’s royal court. Having been King Mindon’s favourite, the saohpa turned into King Thibaw’s enemy. Heavy taxation was one of the reasons for the saohpa and his allies to revolt against Burmese rule. The saohpa took refuge in the equally strong yet more inaccessible Kengtung polity in the east. Several more among the minor saohpas allied in what became a confederacy, which proposed the Burmese Prince Limbin to act as a suzerain and replace Burma’s rulers—first the Burmese, then the British. Similar allied groups operated elsewhere in the Dai territories, whereas some of the saohpas closer to Burma’s lowlands sided with the British and asked for protection. Warfare caused flows of refugees who escaped into Laos, Siam, and Lower Burma. In Christian Daniel’s words, it was a fluid political situation that spread into the Thai areas. This was the situation into which British officers moved from 1887 onwards.Footnote ²⁵

The handful of British officials who were selected for settling the administration of the Shan States were a mix of a few hardened senior officers and younger inexperienced men. Of the future superintendents, Hildebrand was a senior officer and Scott a junior official without officer’s rank.Footnote ²⁶ Each British mission consisted of a few British officers and between 50 and 400 soldiers selected from different regiments. They included Ghurkhas, Pathans, Punjabis, Manipuris, Dai, and Burmese. The soldiers were chosen on the basis of their availability, food habits, adaptability, and knowledge of the region.Footnote ²⁷ People who faced the British troops were either regular villagers who had raised stockades in front of the village entrance or, if the saohpas and his men had prepared and gathered support, some 500–600 fairly well-armed men from different polities would attack strategic targets. When there were conflicts, internal to the Dai territories, such armed forces moved between villages to gain support or find refuge.

There is much silence in the general reports about the aggression waged against villagers outside combat. Battle grounds were uneven. British troops had better arms whereas troops under the saohpas mostly had locally manufactured guns, bamboo spikes, and bows and arrows. More importantly, the British were often outnumbered and the saohpas’ men had the advantage of knowing the terrain. When reports from encounters landed on the chief commissioner of Burma’s desk or were submitted by him to the British Government of India in Calcutta, they included few remarks on actual fights or casualties, unless British officers had been killed. Such reports were the endpoint of a chain of messages from the place of the encounter. But as is shown by the 1893 event discussed below, even the first letter after an event that was sent to the nearest superior officer would mostly not state more than that the men fought bravely, that they chased away the aggressive criminals, and, perhaps, two sepoys (Indian soldiers in the British-Indian army) and a horse had been killed. The number of ‘bandits’ were counted as ‘uncertain’ or ‘many’. This may look like an under-reporting of violence, especially when whole villages were burnt to the ground in ‘punitive missions’. However, as will be discussed below, such violence was so common that it had become an accepted practice that did not merit more than one or two descriptive sentences in reports. The burning of a village could just as well be reported orally upon return to headquarters. It was almost always reported as an unwanted and sad consequence of a situation where people had not yielded to the British sovereign or to her representative, the troops. The officer in charge would give as an excuse that, unfortunately, he had been compelled to use a measure of violence. As an act of force, it was a statement. Such warfare and punishment were significant features of the British punitive missions. As Anandaroop Sen shows, coinciding in time with Scott’s tours, the Lushai-Chin expeditions left behind a trail of burning villages, including their grain stores and livestock. Punishments also included prohibiting people who had lost their homes to construct villages elsewhere. The Bengal government wanted ‘absolute destruction’.Footnote ²⁸

Scott both burnt villages and occasionally reported on the events. The trail of letters left behind followed a hierarchy. A subordinate officer reported to his nearest superior. An army lieutenant would report to a captain and so on up the chain of command. But as the British missions were understaffed, according to the officers stationed there, when on tour, most missions ran on a rather informal basis. Occasionally, informal reports could transfer horizontally and there was an interweaving of civilian and military authority. A medical doctor could write a note to a military police officer who, in turn, included what he thought relevant in his own report to the inspector-general of police who endorsed the views and passed it on to the superintendent of the Southern or Northern Shan States. The two superintendents reported to the chief commissioner of Burma and he, in turn, reported to the relevant department or board of the government in Calcutta. A severe individual event from a mission would thereby be diminished as it moved up the hierarchy when included as one of many items in a report to the highest authority.

The gazetteers comprise another genre of reporting. There is an important difference between gazetteers in Yunnan and in British Burma. Whereas the former were written by local scholars who compiled local sources and Qing provincial imperial records, the latter were part of the British Imperial Gazetteers, authored by senior British colonial officials with work experience from the region in focus. Both the British and Qing gazetteers provided works of reference for newly arrived officials. But in contrast to the local scholars of Yunnan, the majority of the authors of the British gazetteers were born and bred in Europe and their texts were purposed to carry an imperial narrative of progress and sovereignty in order to instruct the officers on how to secure the British empire in Asia. Scott wrote the Gazetteer of Upper Burma and the Shan States in five volumes to be published in 1900. Sentences and paragraphs here and there in the text can be traced to longer reports and correspondence between him and other officers. Considering the enormously rich Scott archive collections, one in Cambridge University Library and the other in the British Library, and all the documents authored by Scott, addressed to him, or that mentioned him in the British Burma administrative records and in private collections, it is quite remarkable to see studies on the Dai polities that only cite Scott’s gazetteers.Footnote ²⁹

Among the cases where armed violence, ending in the burning of a village, is reported, one set of reports stands in a different light. The incident reported reveals the importance of place and the officer who was in charge. It occurred in the small village of Bangda (Pangtap, Figure 1), at the far north-eastern corner of the Dai territories. At the time, these were disputed lands through which the future Sino-Burmese border would run. Villagers here faced George Scott and his men on 4 April 1893. Scott had been warned to move carefully in the border tracts adjacent to China. Nevertheless, he forced his way into the village with 80 soldiers drawn from the Indian army (53 Gurkha infantry men and 27 Punjabis on horseback) and their native officers. Present were also Lieutenant ffrench-Mullen, officer in the military police battalion, Mr Martini, district superintendent of police, the civil surgeon at the Lashio headquarters, Mr Bradley, and Scott’s wife Dora. The village was located right on the border that was negotiated between Chinese and British representatives. Bangda, in fact, belonged to the Zhefang (Chefang) polity of the China side of the disputed border. While on his way there, Scott had been asked not to enter such sensitive tracts. This is why the events that followed needed an explanation beyond Scott’s ordinary routine.Footnote ³⁰

A migration of people from further north, into the Northern Shan States, also played into this situation. Across a long period of time, and as part of a larger migration eastwards of Jinghpaw people mostly of the Lahpai lineage, the Jinghpaw extended their territories into the border zone of Yunnan and south into the Dai polities. Mandy Sadan explains how, in the lower hills, many saohpas of smaller and poorer polities were Jinghpaw who had adopted the Dai polities’ social organization.Footnote ³¹ Their villages were very small in this region, ten houses or fewer according to Scott. All the Northern Shan States had Jinghpaw settlements (Figure 2). Closer to the border that was being negotiated with China, British officers on tour had difficulties knowing how to place them. The position of the paddy fields did not follow the suggested borderline, which caused uncertainty if Bangda was inside British territory (the North Mubang polity)Footnote ³² or in Zhefang (Figure 1). The British mostly referred to Jinghpaw men as ‘raiding tribes’, ‘criminals’, and ‘dacoits’. Some Jinghpaw groups were simply claimed to be ‘alien’.Footnote ³³

Figure 2.

Map of the Kachin Circles of North Hsenwi, that is, the Jinghpaw settlements in North Mubang, 1895–1896.

Source: From the Report on the Administration of the Southern Shan States (ARSSS), 1895–1896.

A rebellion among the Jinghpaw in North Mubang against the saohpa, who was Dai, was the main reason for Scott and his party to travel as far as Bangda. Within a year of his succession, the saohpa began discriminating against the Jinghpaw. British officers noted that there had been heavy fines and even death sentences administered for minor offences. As a result, Jinghpaw attacks quickly spread all over the Mubang polity between December 1892 and February 1893, beginning with an attack on Mubang town. Failing to understand the complexity of the conflict, British officers added fuel to the fire by including Dai troops in their ‘punitive missions’ that chased the Jinghpaw assailants. These troops burnt virtually all the Jinghpaw villages that they passed. However, the British officers soon realized that the rebellion was not directed against themselves but was aimed at removing the Dai saohpa. The Jinghpaw, in fact, were anxious to avoid a conflict with the British.Footnote ³⁴

‘Dacoit’ was a useful word that carried judicial weight. The Indian Penal Code stated that dacoity was when five or more persons committed robbery, or they attempted to commit the crime, or aided others in committing robbery. The sentence for such a crime was life imprisonment or rigorous imprisonment up to ten years. However, the term was used liberally, as evidenced in reports from British India’s North-Eastern Frontier.Footnote ³⁵ When on tour to prevent the conflicts from escalating further, Scott reported on dacoits almost every second day, always referring to Jinghpaw men. About two weeks after the events at Bangda, he wrote to his friend Symes that since he would be on tour and away from Lashio for a longer period of time, it seemed necessary that he explained why he had marched on Bangda and why it was burnt.Footnote ³⁶

Scott loved writing immediate reports with an adventurous touch and his superiors occasionally complained of him using too much poetry. This time, Scott waited two weeks and then took a defensive position. His narrative was crisp and unusually short. He claimed he had marched towards the suggested frontier line to wipe out Jinghpaw attacks on Dai villages, which he assumed related to the uprising in North Mubang and to the killing of an English officer, Williams, at a place called Mang Hang on 6 February.Footnote ³⁷ At Bangda, he wrote, to avoid creating alarm, he first tried to advance with a few men and asked for passage through the village. But they were fired upon and were forced to defend themselves while their interpreters ran into hiding. Scott emphasized that they never crossed into China. However, the real drama is in the narrative’s silences. He wrote:

In the report, the destruction of Bangda disappeared at the immediate moment between the full stop after the words ‘burn the village’ and the capital letter of the following sentence ‘As soon as this was done’. Scott emphasized that burning the village was part of putting an end to the Jinghpaw uprising in North Mubang. The logic of his argument was a bit strained. He wrote that razing the village to the ground ‘was necessary to come to an understanding’ with the villagers. He pointed to the sensitive and still unsettled permanent borderline with China and claimed that people in Bangda therefore should have listened to his orders. It was their fault that he burnt their village. He pleaded with Symes, ‘I did everything possible … the burning of the village was deserved and was moreover a military necessity under the circumstances.’ Finally, he repeated his defence in a list of four points: (1) There was no evidence that Bangda belongs to China, (2) He had to go to Bangda to put an end to the Jinghpaw disturbances; they should have listened to him, (3) He did everything possible to avoid a fight, even at the risk of the life of himself and his men, (4) The village deserved to burn, and it was a military necessity. In fact, these four points seem delivered both as the rationale of his acts and as an instruction for the chief commissioner of Burma on how to write his own report to the Government of India.Footnote ³⁹

Already two days after the event, Scott’s companion, Lieutenant ffrench-Mullen, who led the advance guard of Gurkhas, submitted his own report to the deputy inspector-general of police in Burma in a demi-official letter. The police officer passed it on to Chief Commissioner Edward Symes, acknowledging that the officers in Bangda ‘appear to have done very well indeed’, and Symes endorsed the report on the same day. As with texts that are written soon after an incident that has deeply agitated its author, Lieutenant ffrench-Mullen’s report is longer and more detailed. Omitted from Scott’s report is the kind of trouble he had expected to meet in three Jinghpaw villages on the China frontier. The day before reaching Bangda, ffrench-Mullen received this information from Scott and prepared the Gurkha advanced guard. Arriving at Bangda, Scott ordered ffrench-Mullen to charge the stockades, which was quickly done, and both Scott and Bradley advanced thereafter. The Jinghpaw fled and the Gurkhas went from house to house in search of people as they ‘cleared the enemy out of the houses’. The three Europeans then proceeded about a kilometre beyond the village and found defences but no people, and they returned to the village when they heard gunshots. The Jinghpaw had reappeared and now fired into the mission’s baggage train. After the Gurkha soldiers had chased away all enemies, the outlying portions of the village were burnt and, after moving the troops and baggage to a suitable camp site, Scott, Martini and Bradley returned to see the entire village destroyed. The next day, they found all villages along the road ahead empty. ffrench-Mullen concluded that neighbours were happy to be rid of Bangda, ‘the bully of the entire district’. Whereas Scott chose to report that none of the Jinghpaws had been hit, ffrench Mullen had written that he could not tell how many had been hit, ‘but there were a good many, I have reason to believe’. Soon after, ffrench-Mullen reported having recently received intelligence which stated that seven Jinghpaw men had been killed and three had been wounded.Footnote ⁴⁰

Outside the line of command in the British bureaucratic hierarchy, Scott kept a private diary. This text was intended only for himself. It has an abbreviated and a less polished story:

The chief commissioner would not have read Scott’s private diary and the descriptions of how the company of men, after completely razing the village, feasted on the villagers’ fowls and pigs while speaking jokingly about the battle that had just ended. The officers’ reports, projecting their courageous action and mission success, made the commissioner immediately assure the Government of India that ‘no blame can be attached to Mr. Scott’. In the biannual report from the Northern Shan States, published in September the same year, Scott claims that his earlier assessment was all a mistake. Williams’ death in Mang Hang on 6 February had no relationship to the Jinghpaw uprising in North Mubang. The killers were probably Chinese men from Zhefang on the China side of the border and Jinghpaw men from the neighbourhood. Scott’s Gazetteer, published seven years after the burning of Bangda, stands in an even starker contrast to the narratives in his diary. Here his report on ‘the Kachin disturbances’ has countless dramatic accounts of how Jinghpaw dacoits terrorized, burnt, and looted villages all over the Northern Shan States, and how British officers bravely and tirelessly put down rebellions. Only a handful of events where British officers set fire to a village are mentioned in passing. What happened in Bangda was summarized in two sentences. The sparseness of information has resulted in a fabrication: ‘The solitary exception was at Pang Tap, near the Chinese frontier, where he [Scott] was fired on. The village was burnt and several of those who were engaged in the attack at Mang Hang were killed.’Footnote ⁴²

The years following the capture of the king and annexation of Upper Burma can be characterized by similar violent solutions to armed attacks in the eastern hill tracts. These regional conflicts, which were entangled in the unruly transfer of power from Burma’s king to the British government, were deemed ‘disturbances’, to use British terminology. To burn villages was a practice of conquest that came under the label ‘pacification’. Destruction was also reported from Jinghpaw areas in north Burma and from Khasi, Jaintia, Lushai, Chin and Naga Hills, west of the Indo-Burma range. In the treaties that followed, the heads of polities stated that their villages had been burnt by British troops or were under threat of extinction and, therefore, they submitted to the British government.Footnote ⁴³

The balance of power in the western Dai territories shifted once the British armed presence increased. In the late 1880s, the British were often outnumbered and faced many casualities since they depended on military police garrisons when army battallions were being used further north in the Jinghpaw territories. Bhamo, the large port on the Irrawaddy in north Burma, was attacked both under late Burmese and early British rule from the east and south by the Ponkan saohpa and his allies. Two British military missions against the saohpa and his men had failed already in 1886. In 1889, Burma’s chief commissioner ordered systematic punishment of the saohpa and all who followed him. ‘Any village from or near which any opposition is offered will be treated as hostile and will be destroyed. The force will inflict as much injury as possible on the houses and property of the Ponkan Kachins … If the saohpa does not submit, his village [is] to be destroyed.’ Two months later, the deputy commissioner of Bhamo, Mr G. W. Shaw, reported in meticulous detail that 197 houses had been burnt, ‘large well-built barracks, with their household utensils, such as wooden paddy-grinders, looms, spinning-wheels, &c., had been destroyed. Some 10 head of cattle had been shot, a large number of pigs and fowls killed, fruit-trees had been cut down and some 74,000 pounds of paddy had been taken or destroyed. Thirty-five bins, containing 50 maunds each of paddy …’. By the time the column returned to Bhamo, another 5,000 pounds of paddy and 10–12 head of cattle had been destroyed—probably more, assumed Shaw, ‘as a careful record was not kept’. Chief Commissioner Edward Symes praised Shaw for exercising ‘a wise discretion in not demanding too large a sum’ as a fine. The expedition produced ‘encouraging effects’.Footnote ⁴⁴

The act of burning villages was not unique to the British. It was part of warfare in south-western Yunnan. The Qing army, or people in the polities who were loyal to them, attacked and burnt villages. They particularly searched for and destroyed the grain stores, which forced people to leave. Christian Daniels and Jianxiong Ma explain that there was a general problem with food supply in the hills. This was caused by the increased population from the eighteenth century onwards. Burma’s kings also ordered villages and stores to be burnt when they wanted to depopulate a particular area and thereby prevent the Qing army from advancing against Burma. To control a basin in the hills meant to control the grain supplies, and people who lived here secretly secured food in caves. This was why agrarian lands were more important than buildings, which were soon rebuilt.Footnote ⁴⁵

To the British, destroying every building in a village was an act of collective punishment, causing destruction to effect submission, as well as revenging killed British officers. Even after the Jinghpaw attacks on Scott’s party at Bangda had ceased and the villagers had fled, Scott and his men returned to destroy whatever remained after the fire. Complete destruction was meant to teach people a lesson and force them to submit to British rule.

In later reports, with the benefit of hindsight, British officers concluded that it was a mistake to be reactive and use ‘punitive missions’ after Jinghpaw attacks had taken place in the Jinghpaw region. Instead, they should have been proactive. In a 1912 gazetteer that reported on the northernmost Bhamo district, the author writes with regret that a reactive mode of operation had resulted in ‘comparatively hurried excursion[s] to the hills and back again’.Footnote ⁴⁶ It was reasoned that what guns and torches could not achieve, the payment of tribute or of land revenue would. Submitting payments to the government was not primarily to raise government revenue but to be ‘a convincing testimony of [the people’s] positions of political subjection’, as the author writes. Parallel to the armed encounters, the two superintendents Scott and Hildebrand constructed and applied a bureaucratic grid across the polities. The intention was to replace the existing organization with one tailored to the British administration. However, the administrative setup was argued to be sensitive to the existing culture and to introduce efficient administration where none had existed earlier. Scott, having trained under Hildebrand as a junior man in his operations in the hills, represents a close-knit group of British officers and their shared goals. Thus, a closer look at Hildebrand’s operations in the Southern Shan States will provide a key to understanding the impact of bureaucratic violence.

Bureaucratic violence

In the 1890s, Hildebrand put in place a scheme of institutional changes that altered the political system of governance in the southern Dai polities. The model had been endorsed earlier by Lord Dufferin, viceroy of India (1884–1888), as a simple form of governance, authored by Charles Crosthwaite during his long commissions in the North-Western Provinces. Appointed chief commissioner of Burma (1883–1884 and 1887–1890), Crosthwaite conveyed to Hildebrand that he thought it efficient to appoint a ‘headman’ in every village who owed his authority to the colonial ruler, not to the community. This person was to be responsible for enforcing law and order, as well as collecting revenues. Crosthwaite’s ideas were adjusted by Hildebrand to the conditions in the Dai territories, and made to apply to the saohpas. As we shall see, such a simplistic system proved impossible to introduce. Replacing saohpas with headmen was one piece of Crosthwaite’s three-part policy for the transformation of Upper Burma. The other two were, first, to kill and capture all rebel leaders and uproot all villages that protected insurgents and, second, to recruit a police force from regiments of the military police in India so that the army could subsequently be withdrawn.Footnote ⁴⁷

The restructuring of the administration did not transpire instantly. It is important to remember that this British colonial administrative setup was to be applied to polities with an elaborate political, cultural, economic, and religious organization, many of which were included in the Qing state bureaucracy and hierarchy. As a persistent and forceful policy of remaking the polities’ bureaucratic organization, Hildebrand’s operations were much more detailed than Crosthwaite’s simplistic ideas. He planned to fit the needs of the British administration into operational practice while adjusting the important principle of respecting the ‘native custom’ of a conquered state by imitating something called ‘native’ and ‘customary’ Shan (Dai). He complained of a lack of staff to implement this plan. Parallel to these operations, the same officers were often sent on ‘special missions’ to locate and negotiate the geopolitical border between British Burma and the advancing strong powers of China and Siam. The latter also included negotiating with French officers in their advances from the southeast.Footnote ⁴⁸

Both the British and Qing empires were bureaucratic entities that required some fixity, predictability, and uniformity. In addition to the geographical surveys that mapped a particular territorial space and introduced social and economic information, the boundary commissions and their many preparatory survey missions turned regional mapping into a matter of global politics. The Anglo-Siamese boundary commission’s work concluded in 1893 whereas the Anglo-Chinese boundary commission took longer. The hard border from north to south, which separated eastern and western Dai polities, was still not fixed but de facto came into operation in 1900. However, the border resulting from the commission cut across natural and socioeconomic realities which made it dysfunctional for a long time to come.Footnote ⁴⁹

As evidence shows, Hildebrand’s initiative was not a radical change from the violent and repressive missions as was claimed in later reports. This policy was not an alternative to the systematic destruction of villages even though the latter was said to have been replaced by strictly administrative relations between the ruler and the governed. In fact, armed and bureaucratic violence operated side by side as a continued practice in the grid laid out across the region, the one fuelling the other and often being exercised by the same officers. Keeping in mind that polities were interconnected in various ways, by marriage, political alliances, and trade, armed violence in one section of the Dai territories was immediately communicated and made a concern in polities far away from a particular, individual event. In this way, armed violence ‘embodied the social relations’ and was ever present across the entire region.Footnote ⁵⁰ Scott’s operations in Bangda in 1893, seven years after the annexation, was not the last time such practices were used. On the contrary, they were routine and reported in a somewhat off-hand manner. However, Scott’s reports from his long tour in 1896–1897 among the Wa, in the most central parts of the region, resemble reports regarding his ‘mistake’ in Bangda. The Wa tour was another instance when Scott went against orders and improvised far beyond intended plans. Half a year earlier, on a mission to subdue resistance against British rule, in the midst of conflicts internal to the Wa, soldiers under Captain Elliot’s command had destroyed almost every second village they passed in the central Wa territories east of the Salween. Scott followed in the footsteps of Elliot's and earlier missions, with equally violent results (Figure 3). The secretary to the Government of India remarked that if Scott had observed the Government’s orders, he would not have become involved in any violent encounters with the Wa. This time, Edward Symes recast and sanitized events: ‘… the attitude of certain of the Wa tribes unfortunately compelled him [Scott] on several occasions to assume the offensive. Several villages were burnt and other losses inflicted on the Was, while Mr Scott’s party also suffered some loss in killed and wounded.’ Symes concluded by thanking Mr Hildebrand for his efficient administration and Mr Scott ‘for the tact and courage which he displayed under very trying circumstances during his tour in the Wa country’.Footnote ⁵¹

Figure 3.

‘The last of Daokhoma’, sketch from a tour.

Source: Sketch filed in IOR, BBPP, January–July 1892, Matters regarding the Shan State of Manglün [Wa polity], before the reports from J. G. Scott’s first tour in the Wa territories, 1892.

In the early years of British administration, when expeditions went ‘there and back again’ into the Dai hills from the headquarters in the western lowlands, the British tried to put out the flames of revolt and extract information about the saohpas, polities, and their alliances and conflicts. More importantly, many of the measures taken in the interim period between conquest and settled conditions repeated those where exceptional conditions provided an argument for legitimizing a particular judicial situation. This allowed for solutions that set aside certain laws and legal principles in the long term. Along similar lines, Anandaroop Sen argues that the ‘punitive expeditions’ in the Lushai-Chin Hills, west of the Indo-Burma range, were always transitory in temporal and seasonal structure, which explains the generation of an exceptional legal apparatus: ‘The historical mutability of violence in these frontiers is key to the persistent immutability of its exceptional structure.’Footnote ⁵²

British administrative initiatives were introduced into a situation of agitated relations between king and subjects. The stronger saohpas made their positions clear, as evidenced in the correspondence from 1886 between the western saohpas of different alliances and in letters exchanged by the Mubang saohpa Sawbwagyi Kambawzrata, Colonel Edward Sladen, and the chief commissioner. Without the Burmese overlord—the king—in place, existing conflicts grew and multiplied. The saohpa described how neither Mindon Min nor Thibaw had respected the Dai polities’ autonomy and had requested contributions from tributes and the thathameda tax (a form of property-based tax). In fact, no such contributions were asked of the saohpas during the Konbaung dynasty until Mindon Min’s rule. After conflicts with the king and his loyal saohpas, and three years spent as a refugee in British territory from 1881–1884, the Mubang saohpa eventually submitted to the British government.Footnote ⁵³ Simultaneously, beginning well before the British annexation, a group of saohpas east of the SalweenFootnote ⁵⁴ had gathered in support of the Burmese Prince Limbin against King Thibaw. Promoting two contrasting ways to address the unruly situation in the Dai region, these two groups of saohpas nevertheless had similar ambitions. In letters from four of the saohpas under Prince Limbin to the Hsibaw saohpa, they reported on the movements of their troops and on the British occupation of Mandalay and the deportation of the king. Responding to the Hsibaw saohpa’s suggestion to gather all Dai polities into a collection of independent states ‘like Germany’, they sent a resolution. They wished to see a suzerain king, preferably Prince Limbin, who would protect the Dai polities both from external and internal confrontations. There would be no taxation and obeisance would be made only once in three years.Footnote ⁵⁵ All saohpas agreed to paying tribute and pledging obeisance, but not to paying taxes. In a letter to the chief commissioner, the Hsibaw saohpa made use of the principle that the British claimed to hold in high esteem—to respect ‘native custom’ when settling newly conquered states. The saohpa explained: ‘the levying of tribute from the Shan States being not according to custom, I hope the British Government will see that the Saohpas are exempted from the payment of such tribute, and that under its protection these Saohpas are enabled to govern their respective States’.Footnote ⁵⁶

The British government’s translator Taw Sein Kho explained in a note that before Mindon Min’s rule, the saohpas and myozas paid homage by annually giving ‘royal presents’ and between eight annas to a rupee. The last two kings had requested presents twice a year and introduced the thathameda tax. Now, the saohpas understood the British communications to mean that their former customs and practices—those in place before Mindon Min’s reign—were to be respected, unchanged. We may also note that the Chinese emperors’ requirements from the polities had been channelled via the saohpas. They had full control of all the lands of the polity, however, while themselves being subsumed under the emperor of China. This information was a problem for the British government in India as they wanted to continue to receive revenues from the Shan States. To respect ‘native custom’, the principle of governing India according to the laws of the country when settling land and revenue in a recently conquered territory, was a long-established legal principle. It had to be handled by way of negotiation or rhetorical arguments. Already during British East India Company rule, the balance between acknowledging aristocratic privileges as part of native custom and as compelling as proprietary rights; and acknowledging spaces free from government interference as a violation of the sovereign power’s authority tilted towards the powers of the British sovereign. When these two opposing principles were legally adjudicated, the question of custom was often argued to be irrelevant or damaging to the government in each specific case. Half a century before the question of the legal nature of native custom entered the administration of the Shan States in the 1830s, Mughal privileges were removed on the British North-Eastern Frontier as a result of decades of judicial deliberations. There had been sharp divisions within the colonial administration. There were discussions for or against respecting the long-standing privileges and in the end, the previous varied land and legal conditions made uniform revenue settlement impossible to implement.Footnote ⁵⁷

The tributary relations formed a multicentred web that reveals the extent of the suzerain’s power. However, links to the centre were made manifest only at times of making contributions. Such links were part of a larger complex of relations, which could include protection of communication routes or sending soldiers to the king when so required. The circumstance of such requests rested on the strength of Burma’s king. The relations between court and saohpas were occasionally reconfirmed by privileges, which impacted on relations between the individual saohpas. From the 1750s to the early decades of the nineteenth century, the Konbaung dynasty had expanded through warfare into Ayutthaya and Lanna (seized from Burma by Siam in 1776), Arakan, Manipur, and Assam (conquered between 1784–1817), and Kengtung and Mubang in the early nineteenth century. From the first Anglo-Burmese war (1824–1826), Burma’s reign shrank when authority over these rulers and territories were lost, and the British war indemnity and commercial treaties delivered a heavy blow to the country’s economy.Footnote ⁵⁸ In this way, the reach of a saohpa’s reign changed over time. Seen as spatial relations, boundaries internal to the Dai territories fluctuated. We also need to place these ties to Burma’s king in view of the much older ties with Chinese imperial administration. They were deeper, stronger and, during their greatest extent, reached as far as the Irrawaddy River. There were saohpas who could rely on support of imperial troops when in need. Saohpas whose polities were in both Burma’s and China’s reach could pay tribute to both.Footnote ⁵⁹

At first, the British did not interfere with the old order. In 1886, the chief commissioner agreed to move slowly on altering the way the polities were governed. He argued that, after a couple of years, the saohpas’ ‘independence’ could begin to be circumscribed and the British ‘would govern through the Shans’. The British should ‘as far as possible, [keep] aloof from the direct administration of any Shan State’. However, this situation soon changed.Footnote ⁶⁰ While largely keeping the existing administrative terminology, the contents were altered step by step. Within a few years, the British introduced a form of indirect rule in the Dai territories (sometimes assigning the role of a suzerain to the British government) that had been tested across a century in British India. Both the former kingdoms on the Indian subcontinent (the ‘princely states’) and many small polities in forest and hill tracts were governed indirectly. Some historians emphasize ethnic or racial classifications as the organizing principle and rulers indiscriminately called ‘chiefs’—a ‘universal archetype’ in a ‘standard appearance in state administration around the world’, in Benjamin Hopkins’ interpretation. True, Crosthwaite brought his template for indirect rule from Punjab, with native rulers picked by British officials and called ‘headmen’ to function as mediators between the British government and the people. However, at the level of encounter, imperial peripheries were not global. Practices were shaped according to particular circumstances, which is evident in the case of the Dai. Conceptual universality permeated the high-level reports but dissolved in the encounters between mission officers en route and non-complying people.Footnote ⁶¹

This form of rule had compelling control mechanisms that allowed the native ruler to remain in place and in charge of all domestic affairs, while being subordinate to the British overlord. All foreign affairs were surrendered to the British colonial government and the native ruler was expected to cater to the needs of the British army. It was not uncommon to replace the local judicial court with a British-appointed magistrate, and a ‘resident’ officer could at times have significant powers over the native ruler. These administrative protocols were guaranteed through signed treaties or, as in the Dai polities, a sanad, and was a flexible form of governing that allowed for adjustments. In fact, such a flexible form of rule allowed the governing principles to survive for more than a century, spanning from when the French East India Company established such alliances with the rulers of Hyderabad and the Carnatic in the mid- to late eighteenth century. The British East India Company soon followed the same model, for example, in treaties with the Mughal, Mysore, and Maratha rulers and, in the 1830s, with the heads of many of the small polities in the Khasi Hills on the North-Eastern Frontier, which is discussed below. The balance of power between the British and native rulers could vary according to the situation but the British government often dictated the conditions. In the late nineteenth century, with the mercantile corporation long gone, there was not much leeway for native rulers to influence the conditions. Already by the turn of the nineteenth century, this form of governing was known as ‘subsidiary alliance’.Footnote ⁶²

The administrative grid to be applied was far removed from the realities of 1886. In contrast to territorial influence based on relations of allegiance and rank in the Qing state, the British administration operated through a universal logic. All saohpas were to be equal in rank and the polities were treated as equal entities, forming ‘states’ (to use British terminology) within two divisions, the Northern and Southern Shan States. The administration had a hierarchy of sub-divisions, states, and circles. The political situation in the northern and southern Dai territories differed and Hildebrand explained that the violent conflicts in the north and the more peaceful development in the south were the reason why there were fewer and larger northern polities compared to the many southern polities. As observed by Jianxiong Ma, more important for the fewer and larger entities in the north was the impact of the old and highly developed transportation system between Yunnan and northern Burma, which worked against the more fragmented situation in the south. In the British administrative organization, the northern polities concealed the many small, dependent, subordinated polities held within them. In fact, the most important middle section of the communication routes connected Tengyueh and Baoshan in Yunnan, via Mubang and Hsibaw, with Bhamo and with Mandalay on the Irrawaddy. Here, the violent conflicts before and during the annexation also related to the wealth that was concentrated in the valleys through which valuable goods were transported. The Wenxiu government had wished to take control of these routes but failed to do so. In the 1880s, the Han merchant families returned to a strong position. In the large southern part of the Dai territories, with its more than 30 ‘states’, similarly important communication routes were not absent, but the road network was poor. Much of Hildebrand’s time as superintendent, in fact, was taken up by applying pressure on the southern saohpas to rapidly build roads that could carry wheeled carriages and connect the British headquarters and the ports on the Irrawaddy with markets and mines in the interior hill tracts. This was also made a condition of several subsidiary alliance treaties (the sanads).Footnote ⁶³

The subsidiary form of governance was now expected to bundle the many polities and materialize them as a uniform administrative entity. However, the integrity of the native polity, as assumed in the subsidiary alliance’s basic principles, was undermined when British expectations collided with the polity’s rules and customs of government. For example, when villages were surveyed and assessed to settle the amount of the thathameda revenue or the level of tribute, based on the productive capacity of households and land, the status of particular socio-political elites would not be reconfirmed. The idea of setting all polities ‘on equal footing’ fitted ill with powerful hierarchies in the Dai society. This was especially evident when the new administration’s conception of the saohpas’ status came into conflict with their old official titles. During the Ming dynasty, nine heads of polities on the southwestern frontiers were granted titles on the third level in the state. These titles were of high rank, next to the provincial governor, which was the second level, and higher than the fourth level of the provincial magistrate. Over time, the Qing rulers lost control of most of these saohpas, with only the Mubang and Kengtung polities remaining under their immediate influence. After the Sino-Burmese war, the Qing court also lost direct control of these two. However, within the Dai territories, the old titles and their influence still held some sway.Footnote ⁶⁴

The British superintendents now introduced an administrative structure that would raise small and dependent myosashipsFootnote ⁶⁵ to the level of saohpas and vastly reduce the privileged strong polities. At first, the importance of the existing hierarchies went unobserved by the British. But soon, as in the case of Kengtung, the principles laid down in the new bureaucracy began to be used to put a powerful polity in place and limit the saohpa’s and his ministers’ powers. Ma points out how Kokang gained in status as a polity after it was made part of the territory demarcated for Britain as compensation for the Qing government granting land to France in the southern territories in 1897. As one of Mubang’s 49 districts, Kokang was a low-ranked polity. However, it gained political and economic status from its important geographical location on China’s southwestern frontiers as a portal for British trade, not least in opium, and the Yang chieftain was appointed to Myosa. Thus, whereas the strong polities like Kengtung were brought down, small ones like Kokang were elevated and put on equal footing with the other saohpas.Footnote ⁶⁶

At this point, I want to introduce a small note on land revenues for readers who are at home in British Indian historiography. British administrative usage of land settlements and revenue for the purpose of making people with itinerant livelihoods become sedentary is well known. In his study of colonial Bengal, especially the forest woodlands of the East India Company’s South Western Frontier, K. Sivaramakrishnan concludes that, in the early years, British administrators strove to achieve security by sedentarizing people with mobile livelihoods and forming regimes of control and, thereafter, used administrative means and property regimes, primarily by creating reservations, to gain control over forest resources. In the process of this transformation followed the making of ‘tribal places’, contrasting ‘tribe’ to ‘caste’ as sociological categories. In Bengal, Sivaramakrishnan argues, the two phases of ‘statemaking’ comprised East India Company rule followed by British Crown rule. Likewise, in south India’s hill tracts, as elsewhere within the Indian subcontinent, grazing and shifting cultivation was successively curbed—often by introducing land revenue settlements that privileged sedentary cultivation and conditioned clauses in land law with racial categories, especially those of ‘primitive tribe’ and itinerant livelihoods. In a recent study of British rule in agrarian Assam, Bodhisattva Kar argues that the sedentary cultivators were privileged before the shifting cultivators and any other nomadic livelihood. The ambition was to universalize the administration of a largely varied social organization in an equally varied territory. In the process, the generic category of ryot (approx. peasant) was given centre stage. These cases and many more show that although sedentary cultivation was favoured, the application of administrative forms to achieve control was adjusted to place and time.Footnote ⁶⁷

However, these studies on how cultivation was sedentarized are problematic to use for a comparison with early British operations in the Dai territories. Having been partly subsumed under the Chinese emperors and their bureaucracies for 500 years, the Dai territories thereafter passed through an unstable period during the Hui rebellion and, a decade-and-a-half later, fell under formalized British administration from the 1890s. The main reason why the British wanted to move the control of revenue collection from the saohpa to the British revenue administration was not to sedentarize the population but to undo the powers of the elite. When the territory was divided up between the British and the Chinese rulers, when under threat, the western Dai saohpas and their ministers were effectively cut off from the support they could earlier count on from Chinese troops. In view of Sivaramakrishnan’s study, we may conclude that the two successive phases of securing control by means of arms and establishing formal administration to control natural resources were not introduced in the Dai territories in two consecutive periods but were implemented simultaneously across two to three decades.

Kengtung was located a long way from British headquarters, and it took time before a British officer first visited the polity. Scott arrived (somewhat by mistake) in 1890, and Hildebrand made his first visit in 1892 as part of the Anglo-Siamese boundary commission.Footnote ⁶⁸ Hildebrand’s report included a proposal for the assessment of tribute for the period 1893–1897 which applied to the Central and Eastern divisions of the Southern Shan States. Kengtung was not included in these lists. It was not until 1896, with the succession of the saohpa, that a sanad was put in place,Footnote ⁶⁹ which altered Kengtung’s status. Borrowing the term from the Mughal empire, the sanad was a legal document, used for decrees, grants, treaties, and proclamations throughout British India. The significance of sanads can be seen in the discussion of the ‘homage presents’ of gold and silver flowers that were given to the king of Burma. Such gifts were submitted together with the tribute to a king who represented the powers of the Buddha. Thus, the gifts carried a strong symbolic value, while their monetary value was irrelevant.Footnote ⁷⁰ The secretary to the Government of India, Foreign Department, wrote to the chief commissioner of Burma:

The sanad was granted in 1897; it granted the saohpa his authority on certain conditions. The saohpas had no problems understanding the importance of a sanad. The Qing state manifested the saohpa’s hereditary power with five signs of imperial recognition: a title paper, an official seal, the saohpa’s fingerprint, his genealogy, and a testimonial. The new saohpa was empowered by these documents and by the seal. Now, the sanad replaced the five imperial signs and worked to proclaim the new overlord’s authority.Footnote ⁷²

The Kengtung sanad had 11 conditions and an introduction, which stated ‘… to permit you to administer the territory of Keng Tung in all matters, whether civil, criminal or revenue, and at any time to nominate, for the approval of the Chief Commissioner, a fit person according to Shan usage to be your successor in the Sawbwaship’. The conditions that followed the introduction largely kept to the principles of a subsidiary alliance. The saohpa was to administer the territory according to the custom of the country, and in all matters would be subject to the guidance of the superintendent of the Southern Shan States (para. 4). The British stated that they would not interfere with Kengtung’s domestic politics and matters of justice as long as they followed advice given by the superintendent. The saohpa also had to abstain from communication with states outside British India (para. 2). Foreign relations were in the hands of British authority. Disputes connected with any other part of the Shan states were to be submitted to the superintendent, and his decision had to be applied (para. 7). Thereby, paragraph 7 marked a limit to Kengtung’s exercise of justice. The saohpa could do as he wished when handling cases of subjects of Kengtung’s polity, but not when subjects of other Dai polities were involved (para. 11). Thus far, no conditions in the 1897 sanad were outside of the ordinary.

In addition to these paragraphs were conditions specific to Kengtung. These paragraphs were also used in sanads for other polities; however, the implications would differ between the polities. The sanad stated that the government reserved proprietary rights in all forests, mines, and minerals in Kengtung. When the saohpa wanted to work the forests or mines, he had to lease them from the government and pay royalties on all metals, precious stones, and other minerals (para. 3). He was also to give land for free at any time if the government wanted to build a railway through Kengtung territory. Former landowners were then to be compensated by the government and, from then on, the government would resume all jurisdiction of the lands required for railway purposes (para. 8). The superintendent could at any time appoint an agent to keep him informed about conditions in the state (para. 6). This agent was equivalent to a resident officer and was a euphemism for a British intelligence gatherer. Finally, the saohpa had to maintain order and keep the trade routes open (para. 5). Paragraphs 9 and 10 regulated duties on opium, spirits, and fermented liquor, and requested that the saohpa turn over criminals who had taken refuge within the state. Paragraph 1 was the only one that benefitted Kengtung. The government promised an exemption of five years of tribute payment ‘as a mark of favour’. The British understood that they had to interact gently with a polity as large, strong, and influential as Kengtung.Footnote ⁷³

As can be seen in this discussion, the devil is in the details. The simple separation of authority over domestic and foreign issues, which signified a subsidiary alliance, disappeared in the special conditions. Two sets of conflicting paragraphs deserve special attention. First, administering the polity of Kengtung ‘according to the custom of the country’ excluded most of the country’s territory from the saohpa’s authority, including the valuable natural resources as well as lands where a railway might be built. These terms were to be repeatedly breached in all polities whose sanads included the condition. Forest regulations in particular were often violated. Second, the condition prohibiting the saohpa from interfering with states outside British India, since this was a foreign and therefore a British government domain, was also made to apply to other polities within the Dai territories. This brought the foreign domain right up to Kengtung’s borders. As should be noted, the British government chose to call these polities ‘states’. In 1869, Edward Sladen, while staying as a guest of the governor to the Du Wenxiu government in Tengyue, made a point of complimenting the governor for allowing the Dai to govern themselves, ‘instead of imposing on them chiefs and governors of an alien race’. In contrast, when the Burmese banned the Dai, who lived further south, from holding any such autonomy, Sladen explained, ‘the states were subject to chronic disturbance, and industry had yielded to indolence and poverty’.Footnote ⁷⁴ In 1869, when discussing with the governor the reopening of the passage and trade route from Bhamo to Tengyue, it suited the British to emphasize that these polities were autonomous ‘states’ and not dependents. By the 1890s, with the territories now under British control, the word ‘state’ remained in British India’s vocabulary. However, the word’s semantic contents had been hollowed out. In the long view, we may note that these polities had never been sovereign in the absolute meaning of the term. They had paid tributes to both their neighbours and been granted certain autonomy by the emperor of China for guarding the frontier, yet they remained firmly subsumed under his rule. Autonomy never corresponded to sovereignty.Footnote ⁷⁵

Changes did not take place overnight in Kengtung. In fact, it was one of the last polities in the Southern Shan States to adjust to the new administrative principles. Without the support of the Qing troops to fight back the British, Kengtung’s strong position was coming to an end. The stubborn resistance from the saohpa and his ministers delayed the full impact of the sanad. In 1898, Symes wrote:

Even though the British administration of the western and central districts of the Southern Shan States was stabilizing in the late 1890s, British officials were still sensitive to negative reactions from the saohpas’ administration and kin. The selection of a new saohpa, myoza, or a high-ranking monk was particularly difficult, which called for a public event to put a lid on ‘heart-burning’ feelings. Such events—durbars—had occasionally been held throughout the 1890s. They were organized to gather the saohpas at an occasion where the superintendent could distribute information and orders, settle disputes, install office holders, carry out punishments, and hold social events. At the earliest durbars, held in 1891–1892, Hildebrand had optimistically called for budgets to be submitted from each polity only to find that the figures provided were out of touch with reality and the saohpas had begun to over-assess people. Hildebrand’s ambition to reorganize the polities to suit a centralized revenue authority with predictable revenue incomes (termed ‘tribute’) failed badly. Already in 1893, an idea was launched to bring the saohpas’ sons to a school in Rangoon to be taught the basic knowledge necessary for running the administration of a state. But the parents refused to send their sons so far from home and in 1895–1896, it was agreed that the British southern headquarters of Taunggyi would be the location of the school. However, it took several years simply to get the building in place.Footnote ⁷⁷

In the late 1890s, when Kengtung was also made a region under British administration, Hildebrand instilled sufficient fear to invoke submission to British rule simply by his own and his officials’ presence. The officials and officers were therefore constantly en route through the area. In May 1898,16 saohpas with their accompanying officials, retainers, and monks were invited to a durbar at Taunggyi. The large crowd was balanced by a powerful representation of British attendees comprising all political and civil officers stationed at the British headquarters, the colonel commanding the Southern Shan States, and military officers from Fort Stedman. Gifts for good service and sanads were presented to the Yawnghwe saohpa, the Wanyin Myoza, and the ngwegunhmu of Ywangan.Footnote ⁷⁸ These gifts, which had a carrot-and-stick incentive effect, had to be handed over in such public forums, as explained by B. S. Carey. For example, ‘in Ywangan feelings ran high’ when the next heads were acknowledged and, in fact, decided by the British. Carey, officiating in place of Hildebrand, related how he had clarified the government’s policy of rights to succession to the audience of Dai saohpas and their accompanying officials. The men who succeed the former office holders were disputed between different groups in the polities. Carey reproached the Ywangan officials for insubordination and threats. If they got their way in selecting another son of the former head, the boy—a minor—would have de facto given the ministers control of Ywangan. Carey warned them about ‘speedy and dire punishment’. The message was forcefully echoed at the following week-long deliberations with all saohpas. Half a year later, the same procedure was repeated at the durbar at Loikaw with the Western Karenni chiefs.Footnote ⁷⁹

Conversely, durbars could also serve as an opportunity for honouring saohpas, granting them special distinctions. For example, the visit of the British viceroy and his durbar in Mandalay in 1893 was made a special occasion for the saohpas of Monè (Mong Nai), Mong Pan, and Mong Pawn, the myoza of Samka, and the ngwegunhmu of Pwehla who were granted chains, swords, and medals of honour by the viceroy for their services to the British sovereign. Loyalties mitigated divisions between saohpas.Footnote ⁸⁰

The long-term challenge for the bureaucracy before it could take root was mostly explained in terms of the difficulty in making people visible and thereby accountable in administrative reports. Noting the tributes as ‘revenue’ and making them part of the revenue department’s administration under the Government of India was a step in this direction. As G. W. Dawson summed it up in the Bhamo district gazetteer, the payment of tribute would be ‘a convincing testimony of [the people’s] positions of political subjection’. Tributes in the Southern Shan States were fixed in five-years cycles: 1888–1892, 1893–1897, and 1898–1902. Collections were based on the principles of house tax. The results up to 1898 were meagre. The territory was far too large for the officials to cover in their attempts to collect a reliable enumeration of houses, even in years when one of them worked throughout the rains. Therefore, the inventories were based on native sources and local knowledge in both the Central and Eastern divisions. Hildebrand had to trust the saohpas. Yet by 1897–1898, only 55 per cent of all the houses counted had been taxed in the Southern Shan States. Carey explained to the displeased Government of India that, where tributes could be raised, it was only due to the return of people who had fled their villages during the volatile disturbances in Thibaw’s final years and at the time of annexation. Migration now increased the number of houses that could be taxed in the future. The figures on which tributes were calculated and the collections so far were reliable only in the administrative westernmost Myelat division. In the Central division, only Yawnghwe had been surveyed. But this had taken place as early as 1892 and the figures were therefore unreliable because of migrations. Finally, Kengtung was exempted from tribute until 1902.Footnote ⁸¹ Carey wrote:

Carey’s frank honesty was not well received by the Government of India. Hildebrand, who was in the habit of reporting in a more diplomatic tone, defended his assistant’s work. The Government of Burma, that is, the chief commissioner, had acknowledged Hildebrand’s modifications for refining the assessment. But the Government of India reprimanded Hildebrand for favouring the saohpas at the expense of the interests of ‘paramount power’. Hildebrand received complaints about his failure to discourage people from migrating between states as a way to avoid taxation. Newly arrived persons did not have to pay taxes in the first five years, which resulted in people moving to the neighbouring polity and receiving another five years’ exemption. The government also complained about the insignificant tax burden, and the liberal granting of exemptions from taxation overall. In response and in a patriarchal tone, Hildebrand pointed out the difficulties ‘to wrestle with these ignorant and proud, but timid and loyal, chieftains … to strike the golden mean between good policy and good revenue administration’. A defensive accusation against the superior authority’s inability to comprehend the scale of the operation in view of the few assistants at his disposal can be sensed between the lines in Hildebrand’s report. Travelling with staff and guards in the mountain terrain, from village to village, counting houses and assessing productive resources in order to calculate a correct level of tribute, had to be done in the dry season when other missions were also ordered to be carried out.Footnote ⁸³ Hildebrand lacked staff and, more importantly, the bureaucratic grid could not be put in place by means of a sledgehammer policy. There were times when the mere threat of armed violence worked better than obliterating villages to convey a lesson. The bureaucracy had to be pieced together by unplugging the supporting structure of the existing organization in the Dai polities—the privileges, the saohpas’ unconditioned access to forest and mineral resources, the control mechanisms of the long-distance transportation system, and the social and symbolic manifestations of relationships between the saohpas, and between the saohpas and the former superior power. Bureaucratic violence thereby had a severe impact on the very existence of Dai society.

From a broader perspective, the decade leading up to the turn of the twentieth century saw a bureaucratic grid coming into place through a skeleton of intended administrative functions. The grid remained incomplete with many leaks. Revenues increased, yet budgets failed. Timber leases added to income while inspections identified large areas of girdled teak and green teak trees cut on the saohpas’ orders, which were against the Upper Burma Forest Regulations. Newly constructed roads were few and the work progressed slowly. The violent force embedded in such a bureaucracy materialized in the continuously travelling missions with many informal meetings and formal durbars. The reports make them appear as peacefully conducted within a civil administration, whereas they were almost entirely carried out by army officers who also had civil duties. They were the very same officers who occasionally joined the punitive missions or the tours with civil purposes, which turned violent on the mission’s route. Armed violence was obvious through its deliberate execution, whereas violence within bureaucratic practice operated in exceptional judicial arrangements, political coercion, blackmail, threat, and institutionally initiated eradications. These two methods of enforcement were different in intent and institutional form but overlapped and reinforced each other in practice.

These bureaucratic operations represented the proactive measures to remake the Dai polities and locate them in their positions of political subjection under the British Government of India and within its institutions. Both army and administrative operations extended the bureaucratic grid. During its setup, the saohpas were made the key to secure law, order, and revenues. The presence of the British army and the threat of punishment, including the destruction of villages, was the incentivizing stick built into the grid. Such pervasive threats of violence were transmitted in the veins of social relations, with the saohpas embodying them within the Dai polities. Troops could not always be present even if garrisons were placed at strategic locations, and the military police force was not strong enough to face hundreds of armed men under the saohpas. The superintendents nevertheless had to make it work, which explains the negotiations and judicial exceptions. The saohpas had to be won over, but more through the use of sticks than by carrots.Footnote ⁸⁴

In this way, armed and bureaucratic violence materialized through the practice and the procedures of the officers on the ground and in everyday encounters in which saohpas and their ministers, and local conflicts and immediate needs also played a part in British efforts to govern this region. The very same British officers acted simultaneously on civil and military mandates; however, they were well hidden in reports addressing questions raised by the different departments in the British Government of India. The importance of realities on the ground is particularly evident in situations when superintendents or army officers had gone a step too far and acted on their own judgement, sometimes against orders. As discussed earlier, there were several occasions when the chief commissioner of Burma had to endorse and validate such operations retroactively while burying excessive armed violence in administrative prose.

In fact, it was never a question of following the long-established principle of respecting native custom or acknowledging existing rights after annexation. The exceptional conditions in the interim period between conquest and settled administration, which aimed to flatten Dai social relations and its political hierarchies, took the sting out of the old elites. The new political border, which was already partially at work in the 1890s, frustrated the dynamism that had pitched the Qing and Burmese courts against each other and made the Chinese emperors send armed support to the saohpas. This was an empire soon to fall and the emperor’s promise to stand by the saohpas in return for protection of the frontier tracts had already withered.