simultaneously defining and delimiting indigeneity itself, determining who ‘has been here’ and who is ‘foreign’. If indeed the 1931 and 1953-54 censuses are the basis for the now official 135 groups, as the government claims, there are major inconsistencies. The 1931 colonial census (the final completed colonial census) listed approximately 135 groups based on language, but those groups differed significantly from the 135 ‘National ethnic groups’ listed by the Council of State in 1982 (Cheesman, 2017; Ferguson, 2015). The 1953 census lists Burmese, Karen, Shan, Chin, Kachin and Karenni groups, along with “other indigenous races”. In a second section, Indian and Pakistani, Chinese, European and American are considered foreigners. The 1983 published data uses the same categories for national races, with the addition of Mon and Rakhine, but includes Indian, Chinese, Bangladeshi, Nepalese, “mixed foreign and Burmese”, and “other foreign” as non-taingyintha groups (1986). The 2014 Census, the first carried out in 30 years and the first under the new democratic regime, asked about ethnicity in its questionnaire, yet no ethnicity-related data were released in the Overview and Results (Ministry of Labour, 2018b), the Census Atlas (Ministry of Labour, 2018a), or the Census Main Report (Ministry of Labour, 2015). The reason behind this inconsistency was undisclosed, and little is known about what was done with the data. Within the 135 official taingyintha races, some groups are listed according to exogamous names, while others have endogamous names (Ferguson, 2015). Some are listed by language, others by general location. Some are listed twice. In short, the taingyintha list is the subject of nearcontinuous controversy, yet it maintains great power in limiting access to citizenship. Perhaps most important for the purposes of this paper is examining which groups are not listed among the 135 official taingyintha, as those groups are effectively barred from citizenship by birth and struggle to attain it in any other way, rendering many of them stateless. In many cases, their noninclusion continues despite significant evidence that they have lived in the territory of modern Myanmar since long before the Anglo-Burmese War, and thus should definitionally be included. These groups primarily include Panthay Chinese Muslims, other overseas Chinese communities, Burmese Indians, Gurkha, Pakistanis and Rohingyas (Ferguson, 2015). Legacies of pre-colonial and colonial migration merge together here—Chinese and Islamic migration patterns are swept up with British imperial migration and all of the communities are excluded from membership, despite often long pre-dating the 1823 A.D. cut-off date for potential inclusion in Article 3 the Citizenship Law. Here, ideas of indigeneity and religion also overlap. The largest and most controversial group among those excluded are the Rohingya, who have been primarily categorized as ‘Bangladeshi foreigners’, including officially in the 1983 census, and in a significant amount of government and media discourse since. The Rohingya are a Muslim-majority community living in a Buddhist-majority state, in a region of the country bordering a     23