Census categories and identity in Punjab

From schedules to slips

H. A. Rose’s report explains what happened after the 1901 census schedules were collected. Under the slip system, particulars for each person were copied onto a separate piece of paper. Colour represented the recorded religion, while printed symbols represented sex and marital status. Sorting then produced tabulations for successive characteristics. Rose insisted that unfamiliar occupational descriptions should remain separate at this stage, leaving classification to the compiling staff. A printed table thus stood at the end of several operations, rather than reproducing an untouched answer.[1]

Checking and compilation

His report also records organisational problems. Work was divided between several centres when a single large Lahore office proved impracticable. Rose criticised the weakness of the compilation office and proposed more decentralised copying and sorting at the next census. These recommendations must be separated from procedures already carried out. They reveal how staffing, checking and classification affected the published result.[1]

Comparing changing labels

Guilhem Cassan’s 2015 study confronts the later problem of comparing categories across censuses. He describes assembling an 1881–1921 panel, tracking altered names and classifications and merging entries into comparable groups. Changes to district boundaries required further mergers or exclusions. Such harmonisation is an analytical construction, not evidence that the original labels meant precisely the same thing in every enumeration. The paper makes its grouping choices part of the method being assessed.[2]

His legal classification uses Shadi Lal’s commentary; it does not establish any individual’s identity. Cassan treats census responses as an imperfect proxy for responses faced by the bureaucracy. His account explicitly distinguishes census administration from the administration applying the land law. Aggregate movements between recorded categories therefore require interpretation. They cannot identify a particular person’s ancestry, show that every changed label was dishonest, or settle how a community should describe itself.[2]