Teaching, although not always highly regarded, does afford a valuable perspective on aspects of a country and in this case – ethnicity. Last year I was fortunate enough to teach on an academic development course which enrolled students from across the country; Wa Chin, Karen- even a young Rohingya man were among our particularly diverse cohort. So when I thought about the upcoming census and the heated question of ethnicity, I considered it in terms of that school and of those students. And I realised there was particular problem when asking someone their ethnicity – it is subjective. It is likely to shift depending on who they spoke to. They were often unreliable narrators of their own identity.

Our application process requested that students write their ethnicity in a box, much like a census. It was an optional request but everyone did it, though I found in the classroom their answers were sometimes different. Some from minorities, the Pa-O for example, would vehemently stick to that ethnicity, though others would simply say ‘Shan’. Those with a unique blend of multiple ethnicities, when faced with the question would often round-themselves up to the easiest answer “Oh I’m Burmese”. All of this without mentioning the numerous students from South Asian backgrounds who are forced, by fault of Myanmar’s flawed ‘135 official ethnicities’ to comply with a completely separate majority. One ethnically Tamil student, for example, appeared proud of her heritage even if in administrative terms it did not exist.

Over the twelve months of the program I built up a broad though often inconsistent understanding of ethnicity in Myanmar. In a formal setting – an application or registration – often the ethnicity of least resistance proved the most useful. If you were part Burmese there are many situations when playing that up could prove beneficial. Or perhaps, for a foreign audience, describing the intricacy of ethnicity was simply too complicated a process. However, when the students were granted both time and anonymity, I discovered there were some indeed interesting changes to what they said.

This was made clear to me when a visiting professor gave the students an activity; they were told to write down what they thought was the most significant historical event in their country. Under their name they were instructed to write their ethnicity. The objective was to gauge what different ethnic groups considered an important event – but the test had an obvious flaw; it assumed that each student represented an ethnic group, while of course their backgrounds diverged wildly. (It does not take much to imagine that a Karen person from the delta would have vastly different outlook than one from KNU territory in the east). But when blessed with the freedom of anonymity, the students were freed to write down whatever they felt like. So those who cautiously wrote ‘Burmese’ before now wrote curious amalgamations such as “Shan- Karen-Burmese-Bengali”. Would they have admitted to a Bengali ancestry in public? Probably not, but the privacy shielded them.

A census that focuses on ethnicity and religion is flawed because it is not always a constant. It is complex and contextual. People may project a particular ethnicity when it suits them, or when they feel safe to do so. That is, even if they are being given the choice which is evidently not the case. The opposite could happen when they feel threatened. One student, who was half Rohingya did a good job of disguising his background for several months, indeed up until the end of the program I was unaware he was Rohingya at all. Instead he called himself Rakhine, much to the dislike of other students from the conflict-ridden state. In this case there was a genuine and understandable need to conceal your origin, but in safer situation – a scholarship interview in this case – he was free to admit his background.

Often my impressions were based on trust; I assumed when people described their background that they were telling the truth. Although they did not lie in the truest sense, ethnic backgrounds were sometimes deliberately omitted and re-branded to reflect compliance with a majority. In Kalay, Sagaing Division, I visited a Burmese student but was surprised to see that the father’s sides of her family were all ethnically Gujariti. Her grand-father, a lively and cheerful man who spoke excellent English, explained that they were originally from India.

So would you describe your family as half Gujariti? I suggested.

Well, now we just say we’re Burmese

It took much time for these distinctions to become clear to me. When I first arrived to Myanmar I found it easy to compartmentalize groups and assume a broad narrative which I could easily understand. But the more I traveled, the more I visited homes I relised that ethnicity was far more complex and subject to change than I had expected. This family had, consciously or not, assimilated into one identity to reflect social norms – the majority Bamar. There is no way a census could record such information.

These changes reminded of the story of a young Karen refugee who resettled to Japan during his youth. His story was filmed by the UNHCR to highlight the particular difficulties in resettlement. After struggling with bullying, he chose to conceal his foreign identity; wearing Japanese clothes, adopting customs and perfecting his Japanese sufficiently so that he could blend in. It was only later, after many years success as a fashion model, did he ‘come-out’ as being a refugee. It was odd to watch his interview; what appeared to be a Japanese man speaking of fond memories of the Myanmar scenery and his grandparent’s home. But it did serve to show that if someone wishes to project a new ethnicity, they can. Few people would know any different.

Having ethnicity and religion on a census form is especially meaningless for this reason, assuming the census would be reliable in other ways to begin with – which will surely not be the case. If reports are to believe the previous census conducted in the 1980s vastly underrepresented the Muslim population in the country. A more accurate count this time around could demonstrate an artificial increase in the number of Muslims – a claim which could be mistakenly interpreted that Islam is spreading wildly in Myanmar. Any data collected on ethnicity will be used to manufacture a distorted view of the country. That is, even if it gets to that stage. We’ve already seen that the government has ignored the UNFPA and removed ‘Rohingya’ from the census, an outcome that was perhaps clear to almost everyone except for the UNFPA themselves.

But almost all my issues with the census can be shown in my one final example; when I visited a student living in Taungoo, a broadly Karen and Burman area. His family, in a small wooden house, was Karen-Bengali on one side and Karen-Burmese on the other. His parents and grandparents were all Muslim, but he and his siblings had converted to Buddhism. It was a curious household, indicative of just how diverse the nation can be. But I thought, how would this family respond to a census? Would they be classed as Muslim household, even though the children were Buddhist? Would each of the family members choose one particular ethnicity? Would everyone even choose the same?

The question was answered when that student joined our program, and simply said ‘Burmese’. Because when faced complicated situation or an application form, you just might just chose what’s easiest. And in a single stroke – everything has been rewritten.

Matthew Gibbons works in youth and community development in London. He has been involved with non-formal education in Myanmar for over four years.