Correspondence analysis and the construction of social fields

Abstract

The use of statistical correspondence analysis can be very useful for the analysis of social fields and social space, but it can also lead to misunderstandings. To clarify these misunderstandings, one must take into account the sophisticated thinking that underlies it and that deals with some general aspects of the use of statistics in sociology. Firstly, this article highlights the fact that the statistical practices of Pierre Bourdieu were rooted in a specific sociological tradition that is very old but sometimes overlooked. Secondly, it emphasizes the importance of the stage that consists in establishing the data and the role that field theory plays at this stage. Thirdly, it shows how the analysis and the interpretation of the results of correspondence analysis are, in some respects, dialectic: field theory can benefit from some of the geometrical properties of correspondence analysis, but it should also take into account the limits of statistical tools to the extent that social fields cannot be entirely considered as if they were made of mathematical code.