The past few days saw new fighting in Arakan, whereby today I like to revisit my earlier comment that Myanmar’s inability to become a functioning national entity is the fact that the country, with its present borders, is a colonial creation bringing together peoples and ethnic groups with little in common, and even centuries of conflicts with the Burman kings that predate British rule. And declassified records from British archives reveal that the origins of the current plight of the Rohingya people can be traced to their participation in World War II, notably the communal violence that broke out between the Muslim Rohingya and the Buddhist Rakhine communities during and after the war.

During the 1940's 22,000 of what now is called Rohingya are believed to have spilled into Bangladesh, then British-controlled Bengal, after Japanese forces invaded Burma in 1942 during the Second World War.

Since the 19th century, northern Arakan (or Rakhine state) had been connected to British India through the imperial networks of migration, transport, and governance. The British ruled Burma as a part of British India until 1937 when Burma became a separate British colony. During most of the 19th and early 20th centuries, imperial administrators encouraged extensive migration of “loyal” Indians to offset the influence of the “less reliable” Burmese to maintain stable imperial governance.

During World War II, these divided loyalties determined battlefield strategies of the British in the Allied Burma Campaign. A large number of Rohingya Muslims, who were mostly uneducated indentured laborers in the rice plantations of present-day Rakhine, were recruited to fight on the side of the British-led “Fourteenth Army” against the Japanese forces. The Burmese National Army led by Aung San (the father of the present-day Myanmar leader, Aung San Suu Kyi) fought on the side of the Japanese, who promised them independence from British rule.

Communal violence broke out between pro-British Rohingya Muslims and pro-Japanese Rakhine Buddhists during the war. In 1943, when Rohingya refugees returned to their villages with British troops, the violence continued between the Rohingya Muslims and the Rakhine Buddhists. This time it was about retribution for what happened during the war. It got so bloody that the British military administrators decided to mark the town of Akyab as a “protected area” to stop the return of the Rohingya to their villages to prevent Muslim-Buddhist communal bloodshed.


The Arakan Campaign

The Arakan Campaign of 1942–43 was the first tentative Allied attack into Burma, following the Japanese conquest of Burma earlier in 1942, during the Second World War. The British Army and British Indian Army were not ready for offensive actions in the steep terrain they encountered, nor had the civil government, industry, and transport infrastructure of Eastern India been organized to support the Army on the frontier with Burma. Japanese defenders occupying well-prepared positions repeatedly repulsed the British and Indian forces, who were then forced to retreat when the Japanese received reinforcements and counter-attacked.

Japan’s successful air attacks against Lower Burma, starting in December 1941, and the invasion by its troops in January 1942 led to the evacuation of Rangoon (now Yangon) in early March and, ultimately, the British retreat in late April 1942. Arakan, stretching along the Bay of Bengal, was not, in this initial phase of the war, a major theatre of action, neither for the half-million Indians who fled home via Northeast India nor the British troops which failed to halt Japan’s military offensive in the Irrawaddy Valley. However, Arakan became immediately and remained for two years the scene of ferocious violence between parts of the Buddhist and Muslim communities in the margins of the war between Japan and the Allied forces. Japan’s local allies, the nationalist Burma Independence Army (‘BIA’), entered Arakan ahead of regular troops. The deep-seated anti-colonial and implicit anti-Indian sentiment of this newly formed Burmese armed force hit an explosive context in central Arakan, where rapid demographic growth, a diminishing pool of agricultural lands, and the ongoing immigration of Chittagonians from Bengal had led to the rise of unmediated communal tensions.

Explanations of where and which actions first triggered mass violence on 28 March 1942 vary. Accounts seem to agree that in the townships of Minbya, Myebon, and Pauktaw, Muslims, locally a minority of 10 to 13 percent, were attacked, killed, and driven out of their villages in April 1942. In June, violence escalated further north and hit Kyauktaw township with its 35 percent Muslim population. Houses and mosques were burnt to the ground. Thousands of people fled into the towns of Maungdaw and Buthidaung on the border with Chittagong, where Muslims formed a majority population (56 percent in Buthidaung and 79 percent in Maungdaw). A retaliation campaign against the Arakanese was prompted and lasted until the month of May 1942, emptying Maungdaw and much of Buthidaung township of its Buddhist population: “all Buddhist buildings, pagodas, and monasteries, were razed or burnt, and all Magh [that is, Arakanese] villages burnt and all Magh property (mainly cattle) seized”. 1

Figures of people killed, dispossessed and villages destroyed, as indicated in later memoirs, need to be assessed against the plausible number of the population which can be derived from the Provincial Tables of the 1931 Burma census. It is certainly impossible to know the exact amount of people killed or dying from drowning, famine or disease.2  The number of Buddhists evicted and fleeing south was probably higher than the number of Muslims expelled in the center, because the number of Buddhist Arakanese in Maungdaw and Buthidaung in 1942 (estimated at 78,000) was sensibly higher than the total Muslim population of the townships of Minbya, Pauktaw, Myebon and Kyauktaw (likely 47,000). Still, the impact of the violence on people in neighboring towns, such as Buddhists from Rathedaung in the north or Muslims from Myohaung, is rarely revealed.

Yet, A.F.K. Jilani’s allegation that “more than 100,000 Muslims were massacred and 80,000 fled to Chittagong and Rangpur refugee camps” does not stand up to the evidence.3 The earliest estimate appears in a petition of the Jamiat ul-Ulama of Maungdaw (24 February 1947) stating that 40,000 “innocent people” were killed when the “communal riots spread”.4 If it was meant to include both Muslims and Buddhists, this is surely a figure that warrants serious consideration. The rescue of individual groups from the massacres needs to be included in such tentative arithmetic of victimhood. An Arakanese civil servant, U Kyaw Min led 9,000 Arakanese out of Buthidaung to safe the ground in Bengal in early May 1942; around 5,000 Arakanese were brought by the British to a camp in Dinajpur, and 20,000 Muslims to another refugee camp in Shubirnagar; in April 1943, pockets of Muslims were allegedly taken out of Kyauktaw by Major Aung Tha Gyaw during the First Arakan Campaign.

Several issues raised by this sequence of mutually destructive violence will be addressed underneath like underlying demographic and economic factors; how the Japanese and the British militaries confronted the persecution, how the ethnoreligious hostilities became entangled with the ongoing warfare. The last section underscores that with the denial of an investigative record, a poisonous legacy has lived on.

Any attempt to recount the thick background of the communal the violence of 1942 must include a consideration of the distant and the nearer past of the region to understand how historical patterns overlap and where they do not. Arakan had been an independent Buddhist kingdom from 1430 to 1784, fighting off Mughal attempts at conquest while a large part of its population was Muslims.5 Bengali Muslims, deported and settled by the kings in the Kaladan Valley were the pre-colonial ancestors of Arakan’s indigenous acculturated Muslims, described in many colonial reports.6 Situated at the ethnic and cultural frontier of what are South and Southeast Asia today, Arakan’s political development was determined by coastal migrations, maritime trade networks, and multi-layered configurations of regional rivalries. After a century of political weakness, it was conquered in 1784 by the Burmese king and lost its elites and traditional social order.

Annexed by the East India Company in 1826, domestic peace favored population growth. Still, Arakan remained an impoverished backwater until 1862 when the British created a single territorial unit, British Burma, out of three successively annexed parts (Arakan, Pegu, and Tenasserim). Arakan would lack infrastructure and industrial development projects during its 122 years of colonial rule, but its rice export was developed with an extension of its cultivable lands and the stimuli of taxation and immigration policies. The British, keen to attract Bengali settlers to Burma in the 1870s, were only moderately successful. In the case of Arakan, they tried to draw Chittagonians from across the border, the Naf River, to Akyab District, the most populous of three districts (with Kyauk Phyu and Sandoway). Yet settlers did not come in high numbers, as land to till was still abundant in Southeast Bengal. The situation changed radically in the late 1870s and throughout the 1890s until Chittagonian migrants formed the majority population in the north of Akyab. In Maungdaw township, their percentage increased from 66 percent in 1891 to 79 percent in 1931. The British saw the Chittagonians, unlike the Arakanese, as “thrifty and hardworking,” explaining that “the Arakanese […] gradually being pushed out of Arakan before the steady wave of Chittagonian immigration from the west is only too well known”.7

As the agricultural development went hand in hand with demographic growth (both Buddhist and Muslim) and expanding rural settlements, it did not lead to clashes.8 The reasons may be threefold. First, North Arakan was not densely populated. Until the turn of the century, there was enough waste of land to prepare for cultivation. The area could accommodate the flow of migrants. “Pressure on the land became acute” approximately before 1910, and the influx of new immigrants moved then further down south (creation of Buthidaung township after 1901).9 Second, Arakanese left places where Chittagonians dominated, and Chittagonians kept to themselves, not mixing with the Arakanese.10 Third, the production, harvesting, transport, and milling of rice functioned as an annual cycle of interdependent work where rice-growing Arakanese land-owners depended on Chittagonian seasonal labor. Seasonal laborers (‘coolies’), hired in high numbers for harvesting, were considered as indispensable and counted in the tens of thousands since the 1870s.11 None of the developments before World War I challenged a symbiotic pattern that had existed in previous historical configurations.

However, in the 1920s and 1930s, tensions soared.12 The anti-Indian mood in Burma increased in conjuncture with the anti-colonial movement and economic distress. Another factor was the scarcity of land in combination with the coeval demographic growth of both Buddhists and Muslims, notably in central Arakan (such as Minbya and Myebon townships), where Chittagonians had not much settled before. The numbers of Muslims there were relatively low, but local Muslim growth was sometimes remarkably fast. Unlike in Maungdaw, the dominating presence of single male Chittagonian, temporary laborers may also have become a factor of social frictions. Finally, the administrative separation of Burma and India following the Government of Burma Act (1935) 13 generated new pressures and perspectives for greater autonomy may have led to novel imaginings among the Buddhist and the Muslim elites in Arakan, incrementally redefining their notions of social and political belonging.


The Muslim Area of North Arakan

When the British administration broke down in late March 1942, local Muslim leaders created ‘Peace Committees’ and divided the control of the land, abandoned by the British and not yet occupied by Japanese forces, among themselves. When the Japanese troops occupied most of Arakan and the British returned to the North (September 1942), both faced a situation where persecution in parts of central Arakan, of Maungdaw, and a large part of Buthidaung had effectively taken place. Both did not launch investigations and condoned de facto circumstances of indiscriminate violence and territorial dispossession because they needed the co-operation of the civilian population to pursue their war plans.

Japanese sources on policies in Arakan are scarce, and archival documents on the communal situation reportedly not extant. British sources, often later memoirs, are few, limited to the Muslim area and making contextualization an uneasy task.14 Any attempt to piece together the facts and make a critical interpretation can only try to do justice to a succession of horrific crimes overlaid and ultimately obscured by military priorities. Japan's military was crucial of the anti-Indian the bias of Burma's nationalists, but Thakin Tun Oke's former members of the Burma Independence Army were called – likely including people who had been instrumental in triggering anti-Muslim violence – were playing an ongoing role in the local administration of Arakan until 1944. The Japanese abolished the BIA in August 1942. As a British Burma expert noted in 1945, the BIA "fought, bravely enough, alongside the Japanese, and […] also took over the administration of the districts which successively fell into enemy hands", but with such disastrous results "that the Japanese themselves had to intervene [and] suppress" it.15

As the British intended to occupy North Arakan given an offensive, the former Defence Secretary to the Burma Government, Denis Phelips, was commissioned in the Indian Army and directed to build up a military administration in Maungdaw and Buthidaung. Implementing his mandate depended on negotiations with local bigwigs and their interests. Demonstrating British power and inspiring confidence were difficult as Maungdaw, the main town, was first lost to the Japanese in October 1942 and once again in May 1943. The formal creation of the Military Administration (by order of 31 December 1942) acknowledged local power arrangements, gratified Muslim chiefs who had led the murderous persecution in North Arakan with official positions, and tolerated the re-allocation of Buddhist lands to new occupants who paid taxes to the peace committees.16 On the other hand, local Muslim leaders had few alternatives but to be seen as loyal supporters of the British. Military sources of this period refer to North Arakan Muslims as ‘Chittagonians’ and do not differentiate between their society, which had emerged over the preceding decades and older strata of Muslims present in Arakan. War memoirs testify profuse gratitude towards the Chittagonians who played a crucial role in ‘V Force,’ a reconnaissance and intelligence-gathering formation active along the war front with Japan.17 It is within this context that alleged British promises for the creation of a separate Muslim state could be understood, though they must have been voiced orally as no documents provide any proof. Lt. Col. Phelips’ “Muslim Area of North Arakan” was referenced after the war by the Jamiat ul-Ulama of Maungdaw, a leading group of Muslim dignitaries, as a step towards the creation of an “autonomous State of the Frontier Area”.


Revenge Killings Shrouded by Military Offensives?

Intending to reconquer Burma, British military rationale also dictated the inducement of co-operative relations with Arakan’s Buddhist majority living under Japanese rule, even though anti-colonial nationalists had sided with the Japanese. Other Arakanese, nonetheless, had been part of the former colonial establishment, and members of the Arakanese elite were still advising the exiled Burma Government in Simla.

Yet, the First Arakan Campaign (1942–43) was going to worsen relations with the Arakanese while its aftermath created further untold misery for thousands of civilians who remained in Japanese-occupied territory. British troops advanced from mid-December 1942 crossing the Mayu Range into the Mayu River Valley, but, unable to overcome Japanese resistance on the southern Mayu Peninsula and at Rathedaung (east of the Mayu River), they were stalled.

As they advanced, a vengeance campaign took place at their rear: “large numbers of Mohamedans bent on loot and revenge followed in the wake of our troops as they re-entered Buddhist territory”.18 Major Aung Tha Gyaw, the Military Administrator’s liaison officer reported to the Burma Government “harrowing tales of cruelty and suffering inflicted on the Arakanese villages in [the] Rathedaung area”: “Most of the villages on the West bank of the Mayu river have been burnt and destroyed by the Chittagonian V force”. Hundreds of Buddhist villagers reportedly crossed the Mayu to take refuge on its eastern side. Pearce, the Army’s Chief Civil Affairs Officer raised the alarm with the Burma Government’s Delhi representative: “If the Arakanese form the opinion that our policy is to oust them through the medium of Indian (including Chittagonian) troops, then the Japs will probably find more allies than they even expect”.19 Interestingly, this issue had been raised earlier in a conversation of November 1942 which included Major General Lloyd who, a month later, led the 14th Indian Division to the battlefield. According to a review for the Governor of Burma, one of the participants had summed up the situation in Buthidaung:

Pagodas and Pongyi Kyaungs [that is, Buddhist monasteries] have been razed to the ground. Proclamations and so on are made in Urdu and people have abandoned Burmese and Arakanese languages in everyday speech. This has given disturbances [the] character of Jehad which will add complications if many Musalman troops employed in re-invasion in the rest of Arakan.

To this, Lloyd had replied that “he could not promise that there would be no murders of Arakanese by Muslims in retaliation for earlier outrages”.20

The campaign failed. After the defeat at Donbaik (18 March 1943), the British soon faced a Japanese counterattack and were forced to retreat, losing even Maungdaw.

In Japanese-held territory, the Muslims paid a heavy price. According to a British source, the Japanese had first tolerated Muslim reprisals against “several Buddhist villages” before suppressing the first wave of communal violence. Still, they later prevented allegedly 30,000 Muslims from fleeing to the British occupied territory. Remarking that “conditions among them are […] appalling”, the Military Administrator – speaking in mid-November 1942 – thought that many had “died of starvation and that few of the others [were] able to make the journey if the frontier […] opened”.21


Burma’s independence

After Burma’s independence from British rule in January 1948, violence erupted again in Rakhine. According to the documents from the British archives, a large number of Rohingya rebels were “ex-army men” who attacked “the flanks of regular troops” of the Burmese military before retiring to the hills and forests.” In response, the Burmese government armed onslaught of one minority (Rohingya Muslims) with another minority (Rakhine Buddhists). Thus, one of the earliest blueprints of government support for violence against the Rohingya was crafted.

The partition of South Asia and the independence of Burma offered two models for the Rohingya leaders to integrate themselves into the neighboring nation-states. Their strategy? To make language key to how nationhood would be defined. When the Rohingya Arakan Mujahed Party campaigned to join the partitioned Muslim-majority nation-state of Pakistan, it pushed to use the Urdu script. When the Rohingya Arakanese Muslim Autonomy movement desired autonomy within independent Burma, it pushed for the adoption of the Burmese script for its language.

Those language battles mapped onto real-life violence. In 1952, Pakistan engaged in a violent crackdown on Bengali-speaking students in Dhaka, the capital of East Pakistan. This transformed the demands for Urdu as the Rohingya language into near treachery. The spoken form of the Rohingya language, after all, is similar to the Chittagongian dialect of Bengali. So, how could these separatist rebels adopt Urdu, the language used by Pakistan, when Urdu-speaking Pakistan was trying to repress their ethnic compatriots in East Pakistan?

The multiple Rohingya language scripts, Burmese, Urdu, Nagori and Hanifi, testify to the Rohingya’ struggle to preserve their identity in increasingly violent frontiers of multiple nation-states. It serves as evidence of their efforts to seek recognition and refuge across borders and cultures. These ambiguities in language and territoriality produced in the battlefields of World War II and reproduced in the borderlands of Myanmar, Pakistan and later Bangladesh continue to haunt their “stateless” identity. And it enables the Myanmar state to perpetuate the myth that the Rohingya are merely “Bengali-speaking migrants” and foreigners in their own land.

Not to mention that if the Rohingya are “really” Chittagonian, then Chittagonians are really indigenous to Arakan.


1. British Library, India Office Records (BL IOR), Mss Eur E 390.

2. A journalist visiting Maungdaw in 1943 reports that the number of those killed in early 1942 was estimated at 15,000; see Gordon Waterfield, Morning Will Come, John Murray, London, 1944, p. 104.

3. A.F.K. Jilani, “The Muslim Massacre of 1942”, 2006; History of Arakan (Burma), World Muslim Congress, Karachi,1978, p. 36.

4. Memorandum from the President of Jamiatul-ulama, North Arakan to Arthur G. Bottomley (the British Under-Secretary for Dominions), 24 February 1947, para. 4.

5. Western and Persian sources suggest a complex, cosmopolitan population in the kingdom’s seventeenth-century capital Mrauk U.

6. More than half of Kyauktaw’s Muslims identified as “Arakan Mahomedans” at the 1931 census. The occurrence of the term “Rooinga” (denoting ‘Arakan’ in East Bengali dialects) in Francis Buchanan’s 1798 mention of the language of Arakanese Muslim deportees in Amarapura, confirms this connection. The semantic equivalence of the use of this term with the ethnic-communal meaning implied by modern Rohingya writers has led to heated debates. See Francis Buchanan, “A Comparative Vocabulary of Some of the Languages Spoken in the Burman Empire”, in Asiatick Researches, 1798, vol. 5, pp. 219–40.

7. R.B. Smart, On the Revision Settlement Operations in the Akyab District Season 1913-17, British Burma Press, Rangoon, 1918, p. 14.

8. There is no conflictual violence referred to in British administrative sources; this does not mean or imply that there were no tensions or incidents of violence. The shock caused by the anti-Indian riots in 1938 is an illustration of the aloofness of British authorities regarding social frictions and the risks of communal violence.

9. Smart, 1918, p. 15, see above note 10.

10. This is visible in the 1901 census data for Maungdaw. See Smart’s brief comment on the Arakanese: “his only reply is to move on.” (ibid.). Buddhist migrants from South Arakan settling in Buthidaung were said to have shown greater resilience. As for Chittagonians, they “live in their villages and do not mix in any way with the Arakanese population,” W.E. Lowry, Report on the Revision Settlement Operations in the Akyab District Season 1901-02, British Burma Press, Rangoon, 1903, p. 5.

11. For an early description of ‘Chittagong coolies,’ see Reports on the Revenue Settlement Operations of British Burma for the year 1867-68, vol. 1, Office of Superintendent of Government Printing, Calcutta, 1869, p. 42. Occasional estimations vary considerably in the annual reports (1866–1936). A range of 35,000–50,000 seasonal migrants apply for the interwar period.

12. Such tensions are poorly documented. Arguments presented here are primarily derived from a micro-analysis of decennial census data and invite further review.

13. See

14. Unhelpfully, three relevant files have been missing at the British Library since 1993: India Office Records (hence IOR)/L/PS/12/2271 (“Arakan top secret”), IOR/L/PS/12/2265 and IOR/L/PS/12/2268.

15. B.R. Pearn, “Burma Since the Invasion,” in Journal of the Royal Society of Arts, 1945, vol. 93, no. 4686, p. 156.

16. Associating civil officers appointed by the Military Administrator (31 December 1945, with crimes against humanity hinges on the discussion of the context and individuals in the memoirs of Peter Murray (BL IOR, Mss Eur E 390) and George L. Merrells (Mss Eur F 180/38). One example is Omra Meah, a young schoolmaster, and leader of the Maungdaw Central Peace Committee, who was responsible for driving out the last Arakanese from Buthidaung town in May 1942. He was appointed as township officer of Maungdaw Central Circle while his associates Munif Khan and Nur Ahmed became additional township officers.

17. An often-quoted example is Anthony Irwin’s Burmese Outpost, Collins, London, 1946. On V Force, see the entry in Wikipedia.

18. U.K. National Archives, WO 203/309, f 106A, D.C.C.A.O. (B) Eastern Army to Col. K.J.H. Lindop D.C.C.A.O. (A) G.H.Q. Delhi, 16 July 1943.

19. BL IOR R/8/9, [Chief Civil Affairs Officer (Burma) C.F.B.] Pearce to F.S.V. Donnison [Representative of the Burma Government in Delhi],14 February 1943.

20. BL IOR R/8/9, Note by [Raibeart MacDougall] Counsellor [to the Governor of Burma], 10 November 1942.

21. Ibid.


For updates click homepage here