This case study will analyze the status of marked and unmarked binaries related to social identities in LCSH.
by Emily Drabinski, Coordinator of Library Instruction, LIU Brooklyn
The problem of bias in library classification and cataloging structures has been well documented and analyzed. Efforts to intervene in these systems have largely taken the form of advocating for added or revised subject terms to reflect the language of diverse users and diverse library content. This case study will analyze the status of marked and unmarked binaries related to social identities in LCSH. What might effort to name difference mean for the capacity to retrieve information about dominant identities and ways of being? What is reified as natural and normative to such an extent that it is not visible as subject to analysis by diverse systems that relentlessly minoritize while leaving the so-called “universal” unmarked even as it is profoundly white, male, Christian, and western? Such an analysis might shed productive light on other forms of digital information organization that focus on full and fair representation of non-dominant identities and works, leaving the non-dominant un-named and therefore uninterrogated.
Problem and Context Statement
This case study addresses the status of marked and unmarked categories in knowledge organization structures. Efforts to remedy the problem of bias in library classification and cataloging structures have focused on adding or correcting language in thesauri in order to facilitate information retrieval. Examples span the 1971 publication of Sanford Berman’s Prejudices and Antipathies1 to the 2018 arrival of Cataloging Lab, a collaborative online space for generating and correcting Library of Congress Subject Headings. These projects take for granted the value of being named and therefore visible in information systems.
Such efforts at representation produce multiple effects. For example, recent efforts by activist cataloger Netanel Ganin led to the inclusion of the term Gender non-conforming in the Library of Congress Subject Headings (LCSH), naming works about this identity in the normative system for the first time. These materials, of course, have long existed in library collections. The inclusion of the controlled term authorizes those materials, rendering them smoothly identified with the system. By installing the term in the list of subject headings, librarians establish “Gender non-conforming” as real in the ordering scheme of the library.
There is, however, no authorized heading in the system for Gender conforming. In the case of majority identities, ideologies, and knowledges, this absence does not indicate exclusion from the system, but dominance of it. The emphasis on marking differences in knowledge organization structures means that normative categories move through the cataloging and classification scheme invisibly. The push to mark categories of difference in the system is to make them retrievable, and therefore objects of study and analysis. By leaving the norm unmarked, catalogers and classifiers render dominant categories similarly inaccessible to interrogation.
Knowledge organization structures embed ideology in the intellectual framework of library collections and in physical space. Cataloging librarians understand this power well, and regularly work to remedy exclusions of non-dominant people and ideas in the catalog. This approach to “correcting” the library catalog focuses on adding terminology to the system to enhance inclusion. An idea or identity must be named in order to be represented in the system, and so the project becomes one of naming difference.
Representation–in knowledge organization systems as in other areas of political life–is not an uncomplicated good. Entities do not exist separately from each other, but are constituted through relationships. Hope Olson (2007)2 argues that library classification structures are rooted in Aristotelian logic, producing a “linear, hierarchical structure made up of mutually exclusive categories” (p. 509-510). Items can exist in one and only one dichotomous category, often constructed as “A/Not-A” (p. 513), and power is always at play. A is dominant in relation to Not-A, and the binary construction limits the proliferation of third ways of being, doing, or knowing.
Cataloging and classification efforts that focus on adding and naming the “Not-A” reproduce the Aristotelian hierarchies—and accompanying power dimensions—that render classification structures part of the problem in the first place. Gender non-conforming now exists in the catalog as a “Not-A” entity due to recuperative efforts on the part of catalogers. The controlled term implies Gender conforming, entering a hierarchical and exclusive dichotomous relationship into the catalog along with the term.
The catalog is not an analytic machine, however, so the dominance of gender conforming people is hidden at the same time that it is reproduced within the system. As items are cataloged using the marked term Male nurses, the unmarked term Nurses moves through the system as general term that assumes women. Women astronauts are marked while Astronauts are invisibly men. African American women are marked while white women move through the catalog simply as Women. As categories of difference are named in the catalog, these silences proliferate as well. In leaving one side of the binary unmarked, these efforts render non-dominant subjects open to interrogation while dominant moves invisibly through the knowledge organization structure. Naming in these systems facilitates information retrieval. Failing to name the side of the binary that reflects norms of white supremacy and patriarchy makes investigation into these phenomena more difficult.
Investigating library knowledge organization systems for evidence of bias has an extended and productive history in information studies. Such efforts have usefully highlighted the problems of controlled vocabularies developed by catalogers to reflect dominant ways of knowing and being. This work has resulted in the removal of subject headings like Yellow peril, the revision of headings like Negro, and the addition of headings that reflect diverse racial, sexual, and gender identities. More work will be done in these areas as catalogers continue to address terms like Illegal aliens and Indians of North America and work to add terms that capture the broad diversity of social identities. Such efforts ought to include attention to marking the unmarked normative categories. Such work could render normative categories more easily retrieved in library systems, and more legible as dominant elements of various binaries that pervade socially constructed systems like the catalog.
Next steps or research agenda
Library of Congress Subject Headings are not the only tool researchers use to navigate library collections. The problem outlined in this case study does not prevent research into dominant ideological systems, and does not determine what researchers find in catalogs, especially in a digital era when linear card systems are not the only way to gain access to materials in a given collection. It is, of course, possible to research white supremacy without the consistent marking of whiteness throughout the catalog. However, the structural analysis offered here can serve as a useful tool for teaching students about other political and economic structures that govern their lives. Who counts as a criminal and who is a troubled kid, and how are they treated differently in the criminal justice system? What counts as affirmative action and what is simply hiring for “good fit”? The clarity of marked and unmarked subject headings concretizes these issues, making the catalog a useful analytic text for other lessons about the ways that power works.