A count is also a decision about categories

Census categories shape history by deciding which answers can appear together in a table. A striking Punjab example appears in W. Chichele Plowden’s 1883 report on the 1881 census. Quoting Punjab census superintendent Denzil Ibbetson, it acknowledges difficulty drawing boundaries between religious affiliations and describes practices that officials found hard to fit into their recognised labels. The uncertainty belongs to the report itself, not just to a later historian’s criticism.[1]

Ibbetson’s stated working rule placed someone under the census heading Hindoo if the person could not define a creed or supplied a name outside the recognised religions and their sects. Consequently, the resulting total was partly the product of a classification rule. It should not be paraphrased as proof that every person in that column independently chose an identical religious identity or held the same beliefs.[1]

Modern classifications also need their instructions

The Office for National Statistics provides a useful comparison from England and Wales, rather than a direct continuation of colonial Punjab. Its Census 2021 bulletin explains a two-stage ethnic-group question, combining broad categories with more specific responses and write-in answers. A Roma response option was new in 2021. A category’s appearance in a table therefore cannot, on its own, establish when a community first existed.[2]

The bulletin also reports Sikh write-in identities under more than one broad ethnic category and distinguishes these answers from the separate religion question. A historian counting only the religious-affiliation column would answer a different question from one studying Sikh ethnic self-description. Neither column is an automatic substitute for the other, and the relationship should be investigated rather than assumed.[2]

Compare definitions before comparing totals

Plowden’s quoted discussion also describes inconsistent rules in earlier Punjab returns. Before interpreting a change as conversion or demographic growth, check how officials grouped the answers. The report is evidence of administrative practice too. Its dismissive descriptions of communities remain the official’s judgments, not neutral definitions adopted here.[1]

For modern comparisons, the ONS warns that population change, self-identification and the online search-as-you-type function can all affect results. The practical answer is to read the questionnaire and classification alongside the totals. A number remains useful when the reader states exactly what was counted, where and when, and resists turning one administrative heading into an unchanging account of a community.[2]

Sources & notes

Source records include access dates and, where known, limits on what was examined. Checks used AI assistance; independent human or specialist review has not been completed.

  1. W. Chichele Plowden; quoting Denzil Ibbetson’s Punjab report, paragraph 195. Report on the Census of British India taken on the 17th February 1881 ↗primary colonial census report · Source date: 1883 · Locator: Volume I, printed pp. 21–22, paragraph 38; PDF pp. 48–51 in the digitised compilation. Accessed 2026-11-01.Actual transcription of the specified report pages inspected. The evidence is the administrator’s account of classification, not an endorsement of his religious or caste judgments. Ibbetson extracts and later reprints of the same passage are one original work for the shared budget.Return to A count is also a decision about categories, paragraph 1 ↑Return to A count is also a decision about categories, paragraph 2 ↑Return to Compare definitions before comparing totals, paragraph 1 ↑
  2. Office for National Statistics. Ethnic group, England and Wales: Census 2021 ↗primary government statistical bulletin · Source date: 2022-11-29 · Locator: Sections 2–3: question structure, Roma option, self-identification, write-in functionality and Sikh answers across questions. Accessed 2026-11-01.Actual bulletin inspected. England and Wales is an explicit methodological comparison; no direct equivalence with the colonial Punjab classification is asserted.Return to Modern classifications also need their instructions, paragraph 1 ↑Return to Modern classifications also need their instructions, paragraph 2 ↑Return to Compare definitions before comparing totals, paragraph 2 ↑
Help improve this entry. Suggest a correction with a source.
Record PH-JOU-0016 · English edition · human review not performed