Bias in subject representation: Convergences and divergences in the international literature

Abstract

Terms chosen to represent document subjects, the classification notations assigned to them, the abstracts and indices contain biases in two contexts. In a negative context, document surrogates are constructed with bias when they leave out diverse features, disenfranchise groups and topics outside an accepted norm (Olson, 2002). In this way, these document surrogates would convey inclinations, or prejudices. In a positive context, slanting or tendency (Hjørland, 2008a, McIlwaine, 2003) conveyed by the document surrogates to ensure specificity to particular discursive communities or knowledge domains can be observed. Considering that topics related to biases are scarce in the Information Organization literature, this paper proposes a discussion on the characteristics and occurrences of bias in subject representation. This study is exploratory and bibliographic, and adopts an inductive method.