Abstract

After a brief general consideration of the census, state-formation and moral regulation, together with the use and abuse of this form of 'numbering the people', two major illustrative examples are provided which expose the problematic nature of seemingly objective censal data: illiteracy and ethnicity in the Canadian Censuses of 1921 and 1981 particularly. The bulk of the paper consists of an examination of the trends in ethnicity and gender in relation to varying measures of educational achievement in the Canadian Census 1921 through 1981. The summarised findings are that two of the three founding peoples' ('British' and 'French') are located in the middle range, whilst the third ('Native peoples') is located at the bottom; all other ethnic groups are more polarised. This is a consistent pattern across the sixty years surveyed, although it is often (in popular and academic writing) treated as a 'new phenomenon'. With regard to gender, the paper qualifies the popular myth that the education system is now less gender ascriptive than it was previously: wherever females have been better than males, the males have tended to close the gap or even reverse the situation; wherever males have been better than females, the gap has closed slowly, if at all. This confirms the recent study by Pineo and Goyder, based on the 1981 census alone, that "the Canadian educational system acts more ascriptively upon women than on men." For all the problems of Census data, the article argues that it can be used to discern aggregate trends over time which qualify many contemporary myths.