Anderson’s chapter tours three institutions through which the colonial state imagined its dominion; this inventory stops at the first. The census is the most literal catalogue of the three: it assigns every person a label, a class, and a place. Cataloguing the catalogue shows how the machine works.
The vanishing entry. Across successive Malayan censuses, labels come and go: “Hindoos” disappear after 1871, and “Parsees” last until 1901, packed under “Tamils and Other Natives of India” (Anderson, 2006, ‘The Census’). The result is an “extraordinarily rapid, superficially arbitrary, series of changes,” with the politically powerful categories always leading the list (Anderson, 2006, ‘The Census’). The entries do describe, but each revision records the administrator’s need rather than a stable people; the census was never neutral.
The borrowed label. The 1911 Federated Malay States census files “Javanese,” “Sakai,” and “Bugis” under “Malay Population by Race,” although most groups on the full census list originated in the neighboring Dutch East Indies, whose census-makers imagined “Malays” as one minor ethnicity among others (Anderson, 2006, ‘The Census’). The British list was drawn with “eyes modestly lowered to their own colonial borders” (Anderson, 2006, ‘The Census’). The two taxonomies are incompatible because each was built inside its own jurisdiction: the labels record the classifier’s vantage as much as the classified.
The reserved seat. Under each racial group sits a subcategory of “Others,” and these are “absolutely not to be confused with other ‘Others’” (Anderson, 2006, ‘The Census’). The residual box is what lets the census claim completeness: whatever refuses a label is absorbed by a label for refusal, which Anderson later calls “a splendid bureaucratic trompe l’oeil” concealing “all real-life anomalies” (Anderson, 2006, ‘The Museum’). With anomalies filed away, the census sustains its fiction that everyone is in it, each in one, and only one, “extremely clear place. No fractions” (Anderson, 2006, ‘The Census’).
The angry record. In a late seventeenth-century Cirebon case, the local record would have known an accused murderer only by his court title, Ki Aria Marta Ningrat; the VOC files “angrily identify him as a Chinees,” the “single most important piece of information about him that they convey” (Anderson, 2006, ‘The Census’). Anderson’s contrast is exact: the court classified people by rank and status, the Company by something like “race” (Anderson, 2006, ‘The Census’). The classification did not stay on paper. On the basis of this “inventive census,” the VOC insisted that those it called Chinezen “dress, reside, marry, be buried, and bequeath property” accordingly (Anderson, 2006, ‘The Census’). The label does not simply fail to fit the man; it proceeds to govern him.
Anderson’s own prose performs the operation it exposes. He inventories the census, quoting its labels and preserving its “sic”s, until the grid shows its strain; then he names the whole style of thinking a “totalizing classificatory grid,” whose effect is to render everything “bounded, determinate” and, in principle, “countable” (Anderson, 2006, ‘The Museum’). The census did produce knowledge, but knowledge shaped for rule: legible, bounded, and administrable. The chapter shows a list of people becoming an instrument for imagining a nation, a fiction with very clear places and no room for fractions.
References
Anderson, B. (2006) ‘Census, Map, Museum’, in Imagined Communities: Reflections on the Origin and Spread of Nationalism. Revised edn. London: Verso. Available at: https://ebookcentral.proquest.com/lib/ual/detail.action?docID=5176951 (Accessed: 15 November 2026).


