In the article “Objectivity and Subject Access in the Print Library,” Naun describes how librarians try to be objective with subject headings by imparting impartiality when choosing controlled vocabulary. Problems occur when the terms used are not necessarily what the user would use to search. Bias is a problem in search methods as well, even if the subject headings are chosen with care.
Librarians have strongly guided intellectual access through subject headings, particular by subject headings in print catalogs. Sometimes the controlled vocabulary chosen by librarians are influenced by frame of reference, such as “public improvements” in place of “black urban poverty.” Are these the words that users will use when searching by subject though? How the librarian relates the work to a subject heading could be quite different than the way a user looks for it.
Issues with differences between user search terms versus librarian subject headings were shown to have been around as far back as 1904 by Cutter, saying that an entry chosen should be what is first looked under by the class of people using the library. Cutter goes on to say that there is a rule for choosing one term among several possibilities, and only one should be chosen to keep the catalog from bulking up. His views on the relationship between user behaviors to subject content is repeated later by several theorists. Not only is it important to control the size of a print catalog, but also to save time of the cataloger. Subject headings are also placed with digital, automated materials. Newer methods give statistical search results in frequency of the words occurring in the content, rather than searching for only those items with that assigned subject heading, which are far beyond the capabilities of a manual (print) search method.
When subject headings are chosen, bias is bound to show up in the human indexer’s interpretation. Natural language comes with inherent bias, depending on how the person grew up, or how politically correct one is. Full-text searching removes a source of potential bias, though search algorithms are still present and may have their own bias. Other search methods have been attempted, but any intervention of a human mind between materials and a user have possible bias. Numbered classification schemes attempt to avoid bias by fitting subjects into general categories, such as the Dewey Decimal Classification system, but the problem with it is all expressions of the category are grouped together. Literary warrant comes into play when speaking of objectivity. When librarians use literary warrant when choosing words, it’s not in terms of their independent reality, but of the library’s relationship to human users as an ongoing enterprise. Though even this warrant can be skewed by the society they live in.
Impartiality is at the base of librarianship. Bias will always be there in one shape or another, but we are to aim for using subjects that will be recognizable to a majority of the diverse community of users we serve. The title of the article is a bit misleading. Print card catalogs certainly were a precursor to online subject searching, but this article was on subject access in general, with a majority on searching in a digital system or online browser rather than in print card catalogs. As long as there are humans creating the subject headings, there will be bias. Thankfully, as we move away from the card catalog and use mostly digital subject searching, there can be as many subject headings as we like, rather than being stuck with one or two for print-bulk sake.