Deliberate bias in Knowledge Organization? Advances in Knowledge Organization, vol. 11, 2008, pp. 256-261.

"Bias" is normally understood as a negatively loaded word, as something to be avoided or minimized, for example, in statistics or in knowledge organization. Recently Melanie Feinberg suggested, however, that "if we cannot eliminate bias, then we should instead attempt to be more responsible about it and explicitly decide on and defend the perspectives represented in information systems". This view is linked to related views: That knowledge organization is too much concerned with information retrieval and too much described in the mode of scientific discovery, as opposed to the mode of artifact design: "From the literary warrant of Hulme to the terminological warrant of the Classification Research Group (CRG), to Hjørland’s domain analysis, the classificationist seems like one who documents and compiles, and not one who actively shapes design." This paper examines these claims, which may be understood as questions about subjectivity and objectivity in classification and about positivism versus pragmatism in research. Is KO an objective and neutral activity? Can it be? Should it be? A dominant view has been that knowledge and KO should be understood as a passive reflection of an external order. This has been termed the mirror metaphor of knowledge and is related to empiricism and positivism. The opposite view - which is in accordance with both Feinberg and Hjørland - states that knowledge organization should be functional and thus reflecting given goals, purposes and values. It is related to pragmatism in philosophy