The Re-ethnicisation of Politics in Myanmar and the Making of the Rohingya Ethnicity Paradox

Introduction

Collective violence in Myanmar has been historic and pervasive. It is a manifestation of deep-seated problems and developments that have deepened with Myanmar’s unstable and incomplete transition to a more democratic, that is, less authoritarian regime. Foreign media and Western governments have framed the problem mainly in terms of diffuse, local communal violence on one side and deliberate genocide on the other side. Both approaches reflect the desire to give meaning and cohesion to the unprecedented magnitude of collective violence. Explanations about the causes and onset of violence have widely differed. The following brief literature review outlines the focus points of scholarly research. While it is true that some scholars refer to more abstract and general theories such as religious identity politics (Gravers, 2015; Helbardt, Hellmann-Rajanayagam, & Korff, 2013); ontological approaches stressing a hierarchic order which privileges the Burmese majority (Rampton, 2016); essentialist approaches claiming that the Burmese have a privileged status simply because of their racial status and awareness (Walton, 2013); a national race approach which deals with the genealogy of race-based inclusion/exclusion criteria (Cheesman, 2017b); a psychological narrative based on symbolic interaction frames (Schissler, Walton, & Thi, 2017; Thawnghmung, 2016); or demographic interpretations related to the alleged Muslim overpopulation issue (Blomquist, 2016), others have pointed to the role of the government and the state in Myanmar’s transition (Bünte, 2014; Croissant & Kamerling, 2013; Dukalskis, 2017; Jones, 2014). What characterises the latter is the analysis of top-down processes seeking to understand the priorities and strategies of the military elites. Whereas Dukalskis stresses the importance of ideologies and how they are adjusted to fit the changing circumstances, Bünte (2014) and Croissant and Kamerling (2013) have interpreted this transition process as controlled, institutionalised elite power sharing. By contrast, Jones (2014) has suggested that the military has succeeded in containing not only ethnic separatist threats but also in safeguarding its crucial economic interests. Finally, there are studies that aim to shed light on the puzzling relations and interactions of national elites with local stakeholders. They stress the role of bottom-up local or regional stakeholders shaping political or social outcomes that have affected the nation as a whole. For instance, Van Klinken and Aung (2017) have pointed out how the anti-Muslim rhetoric of local political and religious stakeholders in Rakhine State has influenced and configured the political discourse at the national elite-dominated level.

To sum up, the mentioned approaches offer valuable and meaningful insights that engage with the onset of collective violence. Nevertheless, other factors beyond demographic trends, elite manipulation or religious sentiments need to be taken into account to get a better understanding of the dynamics of violence. I argue that the tension between alternating de- and re-ethnicisation of the Burmese state and politics has increased with the uncertain and incomplete transition to a less authoritarian regime and that this process has been a catalyst for the formation of a contentious and contested Rohingya ethnicity among the Arakanese Muslims.

The analysis proceeds as follows. After defining ethnicity, a general framework that explores the sequence prompting and leading to collective violence is discussed. Thereafter, four critical junctures of violence are identified. The first phase covers the historic legacies of ethnic classifications left by British colonialism. The second phase encompasses the progressive institutionalisation of ethnicity in the legal systems. In this context, the 1973 census, the 1982 citizenship law and the 1989 changes to national identity cards will be discussed. Beginning with the third phase after 1988, the military began to actively promote Buddhism as nation-building tool, whereby religious identity reinforced the ethnicity narrative. The fourth phase that roughly overlaps with the cautious opening of the political system after 2011 and the ensuing intense military–civilian elite competition for state control reflects a sharp upturn in the confluence of violence. Significant developments in this context are the 2015 race laws, the 2015 expulsion of Rohingya party members from the National League for Democracy (NLD), the 2017 murder of Aung San Suu Kyi’s legal key advisor U Ko Ni and—finally—the 2017 wide-scale anti-Rohingya pogroms. The emerging conclusions and main arguments are restated in the last section.

Definitions: Ethnicity and Ethnicisation

Ethnic and racial identity claims are defined as that part of the person’s self- concept that is related to her membership within a group that includes language, religion, customs and traditions. Ethnicity per se is not a problem as self-definition is usually considered an essential quality of ethnicity, but if ethnicity becomes a function of adversity, antagonism or polarisation, then out-groups must be defined along with the internal group whereby the other typically belongs to the minority group. Ethnic relations are subject to change reflecting the social context, political constellations or economic circumstances (Cho, 2018). The treatment of Muslim minorities in Myanmar has evolved over and time. For example, after independence from British Burma, Arakanese Muslims were allowed to practise their faith, join the army and work in government positions and intermarriage was accepted. Almost 70 years later, the Burmese majority not only rejected the Rohingya ethnicity claim but has labelled Rohingya as Chittagong, Bengali or kala (foreigners) to justify their expulsion. Likewise, the 2015 laws on the prohibition of interfaith marriage and religious conversion reflect the fearful perception that intermarriage with aliens endangers national unity and the nation. On the one hand, the Bamar majority along with some Western scholars such as Jacques Leider, Derek Tonkin or Bertil Lintner has refused to recognise Rohingya as an ethnic minority because they believe that the label ‘Rohingya’ is a modern political label and construct. On the other hand, claims of Rohingya ethnicity have increased with the mounting level of discrimination. As mentioned already, group conflicts should not be framed in terms of static, binary social categories but as sets of historically framed dynamic processes. In other words, conflicts are not about ethnicity per se but about extrinsic processes that may exacerbate or overlap with other conflicts and tensions. In times of transition, crisis and uncertainty, ethnicity claims may become more relevant than in times of peace.

If applying the ethnicity concept to Myanmar, it means that there exist many diverse Muslim groups. Among those in Rakhine State, there are many who speak the Chittagonian dialect of Bengalis. They have been classified along with other Muslims living in the same region as ‘Arakanese Muslims’ but due to political circumstances which are beyond their control many have come to describe themselves as Rohingya. In the next section, a framework that explores the links between ethnicity and violence is discussed.

Analytical Framework

Mainstream ethnicity frameworks reflect either a purely instrumentalist or utilitarian viewpoint (ethnicity is a ‘means’ for elites to manipulate the masses in order to achieve political or economic ends and goals) or a purely constructivist perspective (ethnic ‘identities’ are formed through the interaction of the in-group with the out-group and may change over time). The weakness of a purely constructivist perspective is that it is unable to explain why some group members might not accept the ethnic framings of their leaders or refuse to recast the other as opponent. In other words, it cannot account for the situational choice of violent versus non-violent conflict behaviour. Indeed, Walton has pointed out that in some areas of Rakhine State locals refused to exercise violence against their Rohingya neighbours, which means that the adverse social construction of the other was not working everywhere or effective (Walton, Schissler, & Thi, 2016).

A framework that tries to integrate constructivist and instrumentalist arguments has been proposed by Paul Brass who has studied the ways in which elites in India have rallied political support and mobilised masses along ethnic lines (2003). Brass embraces both constructivism as he posits that ethnic identities are not given but framed and instrumentalism as he believes that elites manipulate the masses by appealing to religious sentiment and ethnic identity. He posits ‘that the existence of so-called Hindu–Muslim conflict has been used as a justification for a strong state in India to retain the unity of the country in the face of its internal conflicts’ (2006, p. 115). It follows that outbursts of violence are rather produced and deliberate than spontaneous and rather endemic and recurrent than isolated events. In line with the constructivist assumption that ethnic identity is a variable that may change and be changed over time, his argument is that the relationships among elites and between elites and the state are of critical importance in the study of the rise and perpetuation of ethnic violence. Another assumption that further underlines the theoretical base of his work is that cultural differences are not central to ethnic conflicts and that these are rather the result of the general political and economic circumstances in which elites operate.

Based on ethnographic and empirical research, Brass has identified a recurrent pattern of violence sequences (2003). His model (Table 1) can be summarised as follows: first, there must be a precipitating incident with the potential to be transformed into a riot; second, there must be elites and stakeholders with an interest in fomenting violence; and third, these elites must be aided by an ‘institutionalised riot system’ (IRS). Its existence suggests that violence is a matter of choice that needs to be planned and organised beforehand. It is divided into three phases: preparation, activation and explanation. During the preparation phase, ‘fire tenders’ spread rumours; during the activation phase, ‘conversion specialists’ lead mobs into committing violence; and during the explanation phase, media fuel hatred with master narratives of alien ethnic or alien religious threats to national identity. These networks continue to simmer even when there is no violence. Fourth, these elites must operate within a discursive framework which makes it possible to frame some precipitating event as an ‘ethnic’ incident. Finally, the state administration must not be resolute or competent in preventing violence. Given these factors, violence should result with high probability.

OPEN IN VIEWER

Table 1.

Synopsis of Brass’s Cycle of Ethnic Violence

Precipitating incident

Elites with interest in fomenting violence

Institutional riot system

Discourse of communalism

Absence of a competent state administration

Full-scale riot

Source: Table based on Brass (2003).

When successful, as it most often is, the principal beneficiaries are the national elites. Brass further mentions that special circumstances favour the emergence of violence: before and during elections, and during movements of mass mobilisations, especially when the political balance between contending forces appears to be changing, that is, when political opportunities are such that violence increases the support for a political movement or decreases the support for another.

Brass puts a focus on the interpretive process which follows violence, rather than on the search for their causes because he aims at rendering the violence cycle transparent and comprehensive. For this reason, the impact of historical legacies on one side and the realities of everyday, institutional and legalised discrimination on the other side are not reflected in Brass’s framework. However, his interest into the dynamics of the violence cycle should neither discourage scholars from exploring the colonial origins of contemporary violence nor investigating the legal environments responsible for the institutionalisation of ethnic identities as objective and absolute reality. Hence, colonialism (first phase) and institutional legalisation (second phase) have been added to extend the scope of Brass’s cycle making it more explicit and coherent.

First Phase: Colonial Narratives of Ethnicity

Up to 1784, Rakhine State was an independent kingdom, called Arakan (official name until 1989). Rakhine State is separated from Central Burma by a mountain range. In the past, Arakan comprised Chittagong in present-day Bangladesh to the north and extends down all the way south to the Irrawaddy Delta. Rice has dominated the livelihood and fortunes of people living in the Irrawaddy Delta since the middle of the nineteenth century. Long accessible only by sea due to the impenetrable range of mountain jungles, the Arakan region is now linked by air and road with the rest of Myanmar. Arakan was established by Indian settlers. Conquered by invading Burmans in the eleventh century, they re-established the state in 1238. Mrauk U, the ancient capital of Arakan, was a reflection of the diversity and relative tolerance of its population, incorporating both Hindu–Buddhist and Islamic practices. The Sandhi Khan Mosque in Mrauk U, demolished by the Burmese military in 1996, used to be seen as a symbol of Muslim integration and presence in Arakan. It was constructed in 1433 ad by Muslim army General Sandhi Khan who came to help rethrone Arakanese King Narameikhla. The latter had taken refuge from a Burmese invasion in Bengal. Arakan gradually succumbed to outside powers. First in 1666, it lost the northern part of its territory to the sultans of Dacca. Then in 1784, the south was conquered by Burma. Many Buddhists (and presumably Muslims too) fled north to the area around Chittagong. The Burmans returned in 1793, later crushing Arakanese rebellions in 1795 and 1811–1815. In 1825, the British annexed the territory and Arakan remained under British control until 1948.

There can be no doubt that historic events and episodes have continued to flame tensions and violence in Myanmar. The first influential episode was the exchange of populations and enforced internal migration within British India. Internal migrations were common and encouraged within the borders of British India as the example of British Burma demonstrates. Colletta et al. show the influx of Indians and Chinese changed the social fabric:

The British encouraged and sometimes subsidized immigration of Indians (meaning all those from the subcontinent) to assist in governance and in staffing some lower and intermediate professional positions and certain occupations. (…) Two groups of Chinese entered: one from Yunnan Province as a result of the Panthay Muslim rebellion and the other by sea from south china. The economy was in British, Indian, and Chinese hands, with the Burmans essentially relegated to rice agriculture, the bazaars, and petty trading. (Colletta, Lim, & Kelles-Viitanen, 2001, p. 267)

Migration flows occurred not only from British India and China to Burma and it was not only Indians and Chinese that were migrating to Lower Burma. Indeed, after the British annexation of Arakan in 1824, the massive migration of Burmese deeply changed the demographic balance of Arakan:

For as well as the unwelcome influx of Muslims recorded in the British Gazetteer of 1917, the number of Burmese in Arakan had risen from 4,632 at the time of the 1871 census to 92,185 in 1911, all drawn by the chance of making money in the thriving port (Cockett, 2015, p. 38).

Today’s fears and bias against Rohingya are not limited to Muslim immigration but must also be seen and interpreted in the light of the historic invasion of Arakan by the Burmese as Kingsbury et al. note:

the Rakhine are reacting from a ‘siege mentality’ created by feelings of religious and territorial encroachment by their Muslim neighbors, ‘acute sense of political, cultural, historical, economic, demographic and religious besiegement from the Bamar (main ethnic group in Myanmar)’ and, more recently, perception that the international community ignore their concerns. (Kingsbury et al., 2015, p. 161)

The second influential legacy was the British colonial obsession with nation and race, whereby race was entangled with nationality. The classification of people on the basis of race and religion under British colonial rule fostered the idea of monolithic races with an imagined majority race—the indigenous races—leading and building the nation.

Tonkin, who has examined the British census practice in Arakan, found that Muslim minority groups in Arakan were either recorded as assimilated into the Burmese majority race or as belonging to the Indian race. He summarises the British racial categorisation as follows:

The British recognised that the Arakan Muslim community included both “old” settlers from the 15th Century onwards, who were quasi-indigenous, and “new” settlers who arrived during British rule. The former were classified from the 1921 Census into an “Indo-Burman” Group, thus recognising both their Indian and Burmese heritage (and indeed most of them spoke or used Rakhine Burmese), while the latter were classed among the broader “Indian” Group. By the 1930s the descendants of the “new” outnumbered descendants of the “old” settlers by at least four to one. The “old” settlers included distinct ethnic groups like the Heins, Myedu, Zerbaidi and Kaman, while the “new” settlers were ethnic Bengalis mostly from the Chittagong Region. (Tonkin, 2014)

For sure, as Tonkin asserts, the word Rohingya was not found in British census records, but why? One explanation could be that the British imposed overlapping categories of ‘Mussulman’- and ‘Indian’-obscured Arakanese self-identities which transcended the Muslim-Buddhist dichotomy. Indeed, as Duncan argues, Arakanese Muslims were thought to be racially very similar to Arakanese:

There is one more race [stressed by author] which has been so long in the country that it may be called indigenous, and that is the Arakanese Mussulman. These are descendants, partly of voluntary immigrants at different periods from the neighbouring province of Chittagong, and partly of captives carried off in the wars between the Burmese and their neighbours. There are some 64,000 of them in Arakan which differ from the Arakanese but little, except in their religion and social customs which their religion directs. (Duncan, 1875, p. 30)

The colonial classification system was challenged in the 1950s by anthropologist Edmund Leach with his study of the connections between the Kachin and Shan groups. As he put it more than 50 years ago, precolonial Burma was a ‘wide imprecisely defined frontier region lying between India and China’ where

the indigenous political systems which existed prior to the phase of European political expansion were not separated from one another by frontiers in the modern sense and they were not sovereign Nation-States. […] The political entities in question had interpenetrating political systems, they were not separate countries inhabited by distinct populations. (cited in: Tambiah, 2002, p. 137)

To sum up, British colonialism imposed population exchanges and a rigid grid of ethnic identities and categories on a shifting world that laid the foundations for the later rise of re-ethnicisation by the Burmese majority.

Phase Two: Legal Institutionalisation of Ethnicity (1962–1988)

Since Burma gained independence in January 1948, the political elites have been searching for a balance between a viable state and the centrifugal tendencies presented by the different identity politics of the country’s many groups. Rakhine State has experienced years of economic and developmental neglect. Also, there has been an ongoing conflict between secular socialist policies, as espoused by the military, and more traditional and racist views based on Buddhism as a state religion. Prime Minister U Nu’s pro-Buddhist and federalist balancing policies, however, triggered a military coup. By 1958, he had already transferred power to a military caretaker government. After he attempted to transform Burma into a federal state, which would have accommodated many of the grievances of Burma’s minorities, the army took over and imposed a largely socialist order. First under a military council (1962–1974) and then under the personal rule of Ne Win (1974–1988), the army imposed a socialist system on the country, which badly mismanaged the economy. A system developed in which lip service was paid to socialism, but corruption and personal gain for a small group of military families and their business associates set the tone. Ne Win was eager to isolate the country and liberate it from foreign influences. The nationalisation of all economic sectors resulted, for instance, in the forceful return of hundreds of thousands of Indian merchants after 1962 and led to anti-Chinese riots in 1967. Next, Muslim organisations were banned, religious schools closed down and the construction of new mosques was stopped. Whereas the expulsion of Chinese and Indians had to do with economic woes and feared competition, the targeting of Muslims had to do with unwanted internal migration patterns. In that regard, the 1973 census data (Table 2) had revealed two things: first, the post-1948 dwindling of the foreign population which had been due to a ‘mass exodus of the non-indigenous groups’ (Maung, 1986, p. 8) and second, economic migrations of Arakanese Muslims to eastern borders and other commercially vital areas such as Mandalay, Pegu, Prome, Maolmein, Bassein and so on. Ne Win did not want that. He used the census as pretext for reversing the 1948 Union spirit of citizenship.

OPEN IN VIEWER

Table 2.

1973 Census Data of Burma’s Race and Ethnic Composition

	Total (%)
Burman	68
Shan	8.9
Karen	6.6
Arakanese	4.4
Other indigenous races	6.7
Mixed Burmese and foreign races	1.5
Chinese	0.8
Indians and Pakistanis	1.9
Other foreign races	1.2

Source: Maung (1986).

Hence, the first comprehensive post-independence census may be regarded as a precursor to the 1989 citizenship law as it served as the basis for establishing a list of ethnic groups. Interestingly, Burma had recorded 144 races (reduced to 135 races in 1990)—including six Muslim groups—as native ethnic groups at the time of the 1973 census. The six Muslim communities that were listed in the Instruction Book ‘How to fill up the form’ for the 1973 Population Census issued by the Ministry of Home and Religious Affairs on 9 December, 1972, were, namely, Burman Muslims (including Pathi, Myedu and Zerbadee), Rakhine Muslims (including Rohingya and Kaman), Panthay Muslims (Chinese) and Pashu Muslims (Malay) (Thein, 2012). The Kaman were later acknowledged as the national race (even though they do not figure on the list of the 135 races).

The new citizenship law of 1982 was the logical follow-up to the 1973 census. It stipulates very stringent membership criteria. How did it achieve this outcome? In 1976, preparations had begun for drafting a new citizenship law whose intent was to safeguard dominant positions for the Burmese ethnic majority group. The process from drafting the law to applying it was unusually long and time-consuming. Thus, it took 6 years for the bill to become formally adopted and another 7 years to actually enter into force. Dr Aye Kyaw, a historian from Rakhine and close associate of Ne Win, has been one of the key protagonists of the 1982 citizenship law, which he drafted together with the legal expert Dr Maung Maung and the education ministry official U San Thar Aung. In an interview, Kyaw spoke very frankly about how the three came up with choosing the year 1824 as baseline for deciding who to include and exclude from future citizenship:

Q. How do you define an ethnic nationality?

A. In 1978, while under the Burma Socialist Programme Party rule, me, Dr. Maung Maung (the late President), and U San Thar Aung discussed a law on ethnic nationality. Dr. Maung Maung was an academic on law, I am an historian and U San Thar Aung was director general of the higher education department at that time. We discussed the matter in a room in the State Council office.

I said that for recognizing an ethnic nationality in Burma, there was a census record during the Bodaw Phaya reign, made in the 18th century. It listed all nationalities living in Burma, and it mentioned Arakans, Karens and Mons (Talaings) in the survey. The document can be taken as a base, I suggested. Dr. Maung Maung said that survey was too early. Then I suggested the year of 1824, a turning point in Burmese history when the British annexed lower Burma. Dr. Maung Maung agreed on that date, and we drafted a law that people living in Burma during 1824 were recognized as ethnic nationalities. We found no such word as Rohingya in that survey. (The Irrawaddy, 2009)

The precedence of nationality (i.e., ethnicity and race) over citizenship was the key essence and novelty of the 1982 law. Ne Win stated in one speech that ‘Racially, only pure-blooded nationals will be called citizens’ (Ne Win, 1982). Dr Maung Maung reiterated this point that being a national meant much more than just being an associated or naturalised citizen (or no citizen at all):

Before promulgating the new law, Chairman of the Law Commission, Dr. Maung Maung, argued that Members of racial groups who had settled in Burma anterior to 1823 are citizens[…]they are nationals. Every national is a citizen. Persons born of parents, both of whom are nationals, are nationals and citizens. However, not all citizens are nationals. (Haque, 2014; Working People’s Daily 21, April and 16 October 1982 cited in Kin, 1983). ()

In what sense is the 1982 Citizenship Law xenophobic? It is merely a law which ensures that persons whose ancestors arrived in Myanmar after 1824, when Myanmar was first colonised, and before 1948, when Myanmar became an independent state, could become full citizens if they wished. Many post-colonial societies adopted similar policies. Of course, he [Ne win] was concerned about the loyalty of immigrants to the country of their adoption and in this he was also like many other leaders of former colonial states which inherited significant economically powerful immigrant communities from the colonial period. (Frontier Myanmar, 2016)

Taylor omits to mention that only nationals were considered full citizens. Indeed, the law is discriminatory towards Rohingyas for three main reasons: first, Rohingyas were not deemed to have lived in British Burma prior to the baseline year 1823 when Arakan was ceded to the British and are therefore not recognised as nationals; second, many Rohingya families migrated to and settled in Arakan between 1823 and 1948 during the British occupation of Arakan which would immediately exclude them from citizenship; and third, Rohingyas who came after 1948 are excluded from naturalisation because they were born to parents who are neither both of them citizens. There is another reason which makes the law potentially xenophobic: citizenship except citizenship by birth can be revoked at any time. Suffice it to say that the difference between associate citizens and naturalised citizens—both categories apply mostly to ethnic Indians and Chinese—is a minimal one as Ne Win himself mentioned in a speech: ‘The difference lies in whether they [associate citizens] applied for citizenship or not [after independence in 1948]’ (Ne Win, 1982).

To put it simply, those who omitted to apply for citizenship—for whatever reason in 1948—are less trustworthy and must go through the naturalisation procedure. Nevertheless regardless of the distinctions made, neither associate nor naturalised citizens may apply for official positions or political functions or serve in the army. The 1982 citizenship law only provides one single legal avenue—and window of hope—for Rohingyas to access citizenship: the law does not stipulate proof of citizenship for existing citizens. Hence, in principle, Rohingya who can make proof that they are still holding their National Registration Cards given to them after independence in 1948 should be deemed citizens. Indeed, Art 6 of Chapter II says: ‘A person who is already a citizen on the date this Law comes into force is a citizen’. Moreover, the 1982 law is not retroactive, a provision that should have functioned to prevent mass statelessness by enabling documented citizens to transfer over to the new regime without having to prove their lineage under changed terms and rules.

Even Tonkin, who has been harshly criticised for denying a historically grown Rohingya identity, acknowledges that most Muslims have lived long enough in Arakan to qualify for citizenship:

It is uncontested, and should in any case be self-evident to us all, that some Muslim residents of Arakan could, if only they had the documentation, trace their ancestry back several centuries, while most others, even if ‘Chittagonian’, have been in Myanmar for four or more generations and so qualify in principle for citizenship, even under the much criticised 1982 Act. (Tonkin, 2014)

The biggest obstacle to peace in Rakhine, according to the final August 2017 report of the UN-brokered Rakhine Advisory Commission, is the issue of citizenship. Almost all other issues are linked to citizenship—for instance, access to education and the right to vote and work. ‘If this issue is not addressed, it will continue to cause significant human suffering and insecurity, while also holding back the economic and social development of the entire state’, the report says (Advisory Commission on Rakhine State, 2017). The commission called on Myanmar’s government to establish a clear strategy and timeline for the citizenship verification process and recommended the revision of the 1982 citizenship law.

The changes to ID cards in 1989 represent the final picture in the legalisation of ethnicity. One year after the new military junta had taken power in 1988, the citizenship law was put into effect and with it a major change to national Burmese ID cards requiring inclusion of the ethnicity status of ID card bearers was carried out to reinforce the distinction between nationals and non-nationals (Aung, 2007).

It is important to note that the full practical implementation of the citizenship law in 1989 corresponds roughly to the time when the old NRCs (National Registration Card), the only proof of citizenship for Rohingya, were confiscated and replaced with new ID cards (white cards or TRCs [Temporary Registration Card]). At the same time, the regime introduced also new ID cards for full Burmese citizens: they were issued NRCs that comprised a major change: from 1948 to 1989, ID cards (of full citizens) had contained no entry for ethnicity (for a chronology, refer Table 3). According to Cheesman, ‘even after 1964, no government agency had the authority to issue an official document indicating that a person belonged to a particular national race’ (Cheesman, 2017, p. 471). This changed however in 1989. Since then, all residents in Burma have to carry identity cards, showing the citizenship status, normal place of residence, date of birth, name of the father and so on. The format of the ID card was changed to include not only all of the above but—most importantly—also ethnicity and religion (Aung, 2007). As mentioned already, Rohingyas had to return their NRCs in return for new ID cards (white cards or TRCs). Cheesman has pointed to the legal ambiguity of the revised ID card system: it has created a—what he calls—‘proxy’ or ‘interim’ category of cardholders who are neither citizens nor foreign aliens:

These cards are as their name suggests clearly intended under paragraph 13 of the Union Residents Registration Rules only as interim documentation. They were never intended as alternative documents for non-citizens: in that case, the person should be issued with a certificate under the Registration of Foreigners Rules, 1948. Nevertheless white cards were from the early 1990s used as proxy permanent identification documents for some 700,000 people to whom they were issued. (Cheesman, 2017b, p. 473)

According to Cheesman, the Rohingyas were de facto rendered stateless by the ‘deliberate breach and selective application’ (473) of the 1982 law. He concludes that precisely because nationality has become the new ‘gold standard’ (473), the only way for the Rohingyas to be accepted as citizens is to claim that they too are part of the nation. Regardless of the question whether Rohingyas should rather be viewed as aliens or residents with a special proxy status, which largely depends on the goodwill of the Burmese majority that drafted the 1982 law, the military regime allowed all Rohingyas holding a TRC to cast their votes in the 2010 and 2012 by-elections. After the 2012 conflicts in Rakhine State, President Thein Sein’s government withdrew Rohingya voting rights and those Rohingyas who held a white card had their cards revoked on 31 March 2015. Some 760,000 Rohingyas were affected by this measure. In this way, they were unable to vote in the November 2015 elections. In 2016, Suu Kyi relaunched the effort of finding a solution but tried to skirt the issue of ethnicity by eliminating race and religion from a planned new ID document. But many Rohingyas oppose the proposed cards, saying the omission of ethnicity is part of a continuing effort to eradicate any record of their existence in Rakhine. In 2014, when the former Thein government conducted a pilot ID card verification project, it ordered the Rohingya to identify themselves as Bengali, a label they oppose. Rohingyas say that even if they accepted the new ID cards mentioned by Suu Kyi, many would have little chance of meeting the law’s onerous requirements for attaining citizenship, including conclusive evidence of lineage in Myanmar. A new temporary green card replacement ID introduced in 2017 does not resolve or address the citizenship problem.

OPEN IN VIEWER

Table 3.

National ID Card Evolution

	Chronology ID cards 1948–2017
1948	Rohingyas given National Registration Certificates (NRCs)
1974	Rohingyas given alien (Foreign Registration cards) but many refuse to render their NRCs
1989	Rohingya NRC cards replaced with TRCs (white cards)
1989	New citizenship law enacted
1990	Full Burmese citizens receive NRCs mentioning ethnicity and religion
2010	Rohingyas vote in elections
2012	Rohingya vote in elections
2014	Pilot project by the U Thein Sein grants 200 Kaman and Rohingya applicants citizenship if they agree to ‘Bengali’ label
2015	President Sein orders invalidation of white cards
2016	Out of 759,672 white cards disbursed, 469,183 have been returned and exchanged for new green cards
2017	New green cards (Identity Card for National Verification) are only temporary (valid for 2 years)

Source: Own data file compiled from multiple sources.

Phase 3: From Unitarian Statehood to Religionisation of Politics

Before moving to the post-1988 military junta era, the logic of the military campaigns undertaken by General Ne Win against the Rohingya is briefly discussed. According to Selth, it is rather improbable that personal grievances may have motivated Ne Win’s targeting of Rohingya:

It has long been rumoured in Burma that one reason why Ne win imposed harsh measures against Arakan after1962 was because some Rohingyas had upset his business plans after the Second World War. While Ne Win was well known for pursuing such personal grievances, the measures taken by his regime against the Rohingyas reflect more complex motives. (Selth, 2003, p. 36)

The more complex motives may reflect the wish to maintain a strictly unitarian state amidst secession and liberation struggles, and to divert attention from internal problems. As Arakan had become the poorest province in the country, the Arakanese were forced to leave in search for new opportunities in Eastern Burma such as the Shan and Karen States and the Moulmein area. The 1973 census showed that Arakanese Muslims had spread to these eastern borders and other commercially vital areas such as Mandalay, Pegu, Prome, Maolmein, Bassein and so on. Ne Win did not want that. The Muslims were to stay in Arakan, so that the Arakanese Buddhists and Arakanese Muslims could be used against each other. This was the best way to keep the national liberation movement of the Arakanese under control. The history of Arakan since the 1962 military coup is marked by the emergence, fragmentation and disappearance of numerous insurgent groups, both Arakanese and Muslim. Since 1948, many different Arakanese groups had been fighting the Burmese government—among them several armed Rohingya groups (Kingsbury et al., 2015).

In 1974, a new constitution made Arakan a state within the Union of Burma. Even though Ne Win and his legal advisor Dr Maung Maung agreed to the creation of an Arakanese state, they ensured that the nominally declared state possessed no real autonomy. After having given in to Rakhine demands, Ne Win took his chance to take revenge not only on the Rohingya but also on the Bangladesh government and launched in 1978 an anti-Rohingya military operation under the pretext of curbing illegal immigration. More than 200,000 Rohingyas sought refuge across the border in Southern Bangladesh amidst widespread reports of army brutality. Under international pressure, Burma agreed to take back the Rohingyas. The 1978 repatriation agreement stipulated that Burma ‘agrees to the repatriation at the earliest of the lawful residents of Burma (stressed by author) who are now sheltered in the camps in Bangladesh on the presentation of Burmese National Registration Cards’.

By the end of Ne Win’s rule, circumstances had changed. In 1988, civilian opposition confronted the military regime resulting in violence, many deaths and the suspension of economic aid by the leading Western powers. The powerful State Law and Order Restoration Council (SLORC) that formed the new government was chaired by General Saw Maung, a loyal follower and supporter of Ne Win who remained behind the scenes for some time after his removal. The biggest change to come was the changing attitude of the military towards the Buddhist clergy: instead of ruling—as Ne Win had done—by bypassing the Buddhist clergy, the military junta began to engage with Buddhism. This was by no means a harmonious or peaceful process as the suppression of the 2007 Saffron Revolution demonstrates. Suffice it to say that the military framed Buddhism in terms that served its own interests. Unlike Ne Win who had prevented Buddhism from becoming a politicised religion, the military junta employed Buddhism as a tool for nation-building. As Foxeus put it, ‘the SLORC-SPDC implemented a modern reinterpretation of the precolonial hierarchical cosmological Buddhism’ (Foxeus, 2017, p. 225). In the meantime, the situation in Arakan had deteriorated. In late 1989, the SLORC began resettling Burmese in the predominantly Muslim areas of Arakan. As in other states featuring ethnic minority insurgency, military intervention was stepped up. Forcible relocation for urban redevelopment and counter-insurgency led to the uprooting of large numbers of people. A border development programme was launched in Arakan with military backing in September 1991. Violence against Rohingya led to mass exodus. A repatriation agreement struck with Bangladesh in 1992—the joint statement by the foreign ministers of Bangladesh and Myanmar issued at the conclusion of the official visit of the Myanmar Foreign Minister to Bangladesh 23–28 April 1992—stated that refugees should be resettled ‘to enable them to carry on their livelihood as members of Myanmar society’ [stressed by author] (Joint Statement, 1992). In later agreements, the term ‘member’ was replaced with ‘resident’ which is less suggestive of being a permanent member of a political community. Despite the stated resettlement promise, refugees had to go through first a registration procedure. This meant that only refugees registered by the Bangladesh government could be considered for Myanmar verification. What matters most is that according to the 1992 agreement—and similar later agreements—even those who met the criteria for citizenship must provide bona fide evidence of their residence in Myanmar, with documents such as ID cards or proof of their addresses. After the 1991 mass exodus, for example, it took the UNHCR 4 years to convince Burmese authorities to document the Rohingyas by issuing TRCs (white cards) to them.

The 1991 campaign against the Rohingya precipitated the reshuffling of SLORC. Then Shwe became Myanmar’s head of state and government in 1992, succeeding Saw Maung as Chairman of the military junta. Then Shwe immediately began to repair Burma’s image amidst the international criticism directed against Saw Maung’s earlier campaign in Rakhine.

Phase 4: Increased Elite Competition for State Control (2011–Present)

With the uncertain and incomplete democratic opening of the political system after 2011, violence resumed and it became a function of intense electoral competition between the powerful military, the civilian NLD opposition and the local Rakhine political parties. In order to undermine both the popularity of the local Rakhine parties and counterbalance the growing influence of the oppositional NLD, the military exploited the Rohingya voting issue in Rakhine State as Thawnghmung has pointed out (2014). The four following episodes capture how the Rohingyas have suffered more than any other group from the consequences of this locally and nationally fought power struggle.

The August/September 2017 Pogroms Against Rohingya

Persecutions were triggered by the attack of Rohingya armed units on government military posts in August 2017 that left many dead. The Arakan Rohingya Salvation Army (ARSA), a new Rohingya insurgent group which made its first appearance in 2016, has claimed responsibility for these attacks. Not much is known about its goals, supporters and networks (Japan Times, 2017). Leading genocide scholars such as Thomas MacManus, a researcher at Queen Mary University of London who authored a report ‘Countdown to Annihilation: Genocide in Myanmar’ with de la Cour Venning and scholar Penny Green in 2015, have qualified the 2017 violence in Rakhine as a genocidal state crime, whereas the UN cautioned that violence may have amounted to crimes against humanity. Even though no one has been so far held accountable, and even though the exact number of victims remains unknown, there is enough reason to believe that crimes have been committed and have been committed by particular persons. Indeed, recent reports suggest that crimes against civilians were committed and that the military elite bears responsibility. In December 2017, for instance, the US administration blacklisted a senior Myanmar army general who it said oversaw human rights abuses committed by security forces against Rohingya in the August 2017 army offensive in Rakhine State. In a related statement, the US said that the secretary of State had determined on 22 November 2017 that the situation in northern Rakhine State in Burma constituted ethnic cleansing (Press Statement US Department of Treasury, 2017). The indicted army General Maung Maung Soe was reportedly removed from his command post in Rakhine State.

In a December 2017 report based on field surveys, the international NGO ‘Doctors without Borders’ (MSF [Medecins Sans Frontieres])—which had been working in Rakhine since 1994 until its suspension in 2014—has estimated that 11,391 Rohingyas had been killed in Rakhine State in the first 31 days following the outbreak of violence on August 25 (Médecins Sans Frontières [MSF], 2017). According to the same report, 69.4 per cent of the deaths had occurred by shooting with firearms which suggests the involvement of security forces. The report revealed further that the mortality rate for men was double that for women and that 1.7 per cent of the killed had been children below the age of 5. In January 2018, finally, the international community was taken by surprise when Myanmar’s military conceded for the first time that its security forces and Buddhist villagers had wrongly killed 10 Rohingyas whose bodies had been found in a mass grave (ABC News, 2018).

Meanwhile, the terms of the Rohingya repatriation have been casting doubt on Myanmar’s willingness to guarantee the safety and basic rights of Rohingya refugees. Bangladesh said it would complete the process of returning within 2 years. The communiqué that was issued in January 2018 is of a general nature and does not specify the terms and details of reintegration:

(…) The Physical Arrangement stipulates that the repatriation would be completed preferably within 02 (two) years from the commencement of repatriation. Verification and return will be based on considering the family as a unit. The meeting also finalized the “Form” for verification. Modalities for repatriation of orphans and children born out of unwarranted incidence have been incorporated in the said Arrangement. Under the Physical Arrangement Bangladesh would establish 05 (five) transit camps from which returnees would be received initially in 02 (two) reception centers on Myanmar side. Myanmar would shelter the returnees in a temporary accommodation at the Hla Pho Khung and expeditiously rebuild the houses for the returnees to move in there. Myanmar would consider resettling the people staying at the zero line on a priority basis. Myanmar has reiterated its commitment to stop outflow of Myanmar residents to Bangladesh. Both countries agreed to form 02 (two) Technical Working Group, one on verification and the other on return. The Physical Arrangement has included modalities of the relevant aspects of the return. (MOFA Bangladesh, 2018)

Before moving to the conclusion, the re-ethnicisation milestones that have been elaborated and presented in the article are consolidated into Brass’s (extended) violence framework (Table 4). It provides an accurate and complete representation of the colonial legacies and legal frames that underlie violence and of the sequence of events, interests, networks, discourses as well as the administrative patterns that explain and may even predict the eruption of violence. It confirms that violence in Rakhine cannot be attributed to Rohingya ethnicity claims per se but must be conceived as the—preventable—outcome of a collusion of national and local interests that have misused ethnicity and religion and contextual factors such as internal migration caused by an economic crisis and liberation struggles at the periphery during the rule of Ne Win or the post-2011 uncertain and incomplete transition to a less authoritarian regime. Finally, the framework elucidates and names those who bear direct responsibility for the violence, namely: the 969 movement, MaBaTha, the pro-military USDP and the local ANP (former RNDP).

OPEN IN VIEWER

Table 4.

Brass’s (extended) Model Applied to the Anti-Rohingya Violence in Myanmar

Colonial legacies (extended)	Legalisation of ethnicity (extended)	Post-2012 triggering events	Elite interests	Post-2012 ‘IRS’ networks	Communal discourse	Incompetent state administration	Full-scale violence
British classification by ethnicity/race	1973 census; 1982 citizenship law; 1989 changes to ID cards	Series of incidents between June 2012 (rape and murder of Rakhine women) and August 2017 (ARSA attacks on border posts)	Before 2011: Unitarian state, pacify periphery; After 2011: electoral power competition for state control	Preparation: 969 movement and later Patriotic Association of Myanmar (MaBaTha) Conversion: Pro-government USDP; party/military; and local Rakhine party (RNDP, later ANP) Explanation: Derogatory reporting by state media and by social media platforms	Before 1988: Ethnicity narrative only After 1988: Ethnic othering and appeal to Buddhist sentiments	Civil and police administration and the local authorities failed completely in their most elementary duties of maintaining law and order, prosecuting crimes and punishing the perpetrators	1978: 290,000 displaced 1991/1992: 260,000 displaced 07/2012: 225,000 displaced 08/2017: 600,000 displaced

Source: Based on data compiled by author.

Conclusion

The article explored the roots of the Rohingya issue in Myanmar and argued that internal migration flows during British colonial rule and racial classification systems dividing populations into indigenous and aliens categorisation laid the foundations for the re-ethnicisation of politics. The historical centrepieces are the 1973 census, the 1982 citizenship law and the 1989 changes to national ID cards. The post-2011 electoral competition and struggle for power and state control reached unprecedented ethnicisation levels as Rohingyas were totally delegitimised by having their resident status converted from quasi-citizens (NRCs) via proxy permanent residents (white cardholders) to undocumented, de facto non-residents. The extended framework of Brass has been useful in identifying and framing the Rakhine violence as an institutionalised and organised process. It also confirms that it will not be easy to break or reverse the pattern of violence in Rakhine. By now, the Rohingya community in Rakhine State has been so weakened and marginalised and social relations between the Rakhines and the Rohingyas so damaged that a viable solution seems way far off. Moreover, it seems clear that the military is not willing to compromise on the issue of citizenship as the strongman and head of the Burmese Army, Senior General Min Aung Hlaing, has made very clear when he ruled out any future Rohingya citizenship. Citizenship as a secular and civic concept should be based on the premise that each and everyone—regardless of ethnicity—who has resided for a reasonable amount of time within the borders of Myanmar and who respects the Constitution and legal state order should be entitled to full citizenship. In an ideal state of full and equal citizenship, there should be no need and no room for a special or separate ethnic minority status for the Rohingya or other groups. However, as long as significant numbers of Arakanese Muslims are denied access to health services and education and prevented from speaking their language, exercising their religion, feeling physically secure and moving freely, discussions about citizenship are disingenuous at best. As long as fundamental human rights are violated, it seems inappropriate or at least premature to think about Rohingya citizenship because—from a victim perspective—it is difficult to conceive how anyone would like to become the citizen of a state that does not want to guarantee the most basic rights for all.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.