Conflicts between Authenticity and User Retrieval in Classification Work

The field of knowledge organization theory shows major divides in priorities and conclusions: whether the core of classification is structure or meaning (Beghtol, 1986; Mills, 2004; Spärck Jones, 1970), whether a classification designer can or should rely solely on the literature to determine terminology (Fidel, 1994; Hulme, 1911; Mai, 2005), and whether it is the global principles (Ranganathan, 1962) or the individual classification designer’s knowledge and insight (Albrechtsen and Pejtersen, 2003; Feinberg, 2007; Hjørland, 2002) that creates a good system. The field of classification research offers no clear consensus on how classification construction occurs in practice or its ideal version (Park, 2008). Nor do we understand how classificationists negotiate conflicts between these principles and approaches in daily work. For example, if literary warrant is insufficient to make semantic decisions (Fidel, 1994; Mai, 2005), how do classification designers justify and reconcile – to themselves, their institutions, and their users – the inevitable use of personal and contextual information with the belief that the system should represent the documents without bias? Similarly, at what point (if any) in a domain analysis process does the classification designer encounter the impact of his or her system on the field being organized, and how does he or she seek to minimize or justify this distortion