4 Classification
Overview
Subject metadata—particularly classification and subject headings, the respective focuses for the next two chapters—allows metadata creators to communicate what library resources are about using predetermined schemes and vocabularies. This chapter will discuss library classification as it relates to DEI. Rather than attempt to cover the topic exhaustively, it provides readers with a brief overview of the issues and why they matter. It will then describe two practices that can improve the inclusivity of library metadata: keeping the library’s classification current while continuing to use mainstream schemes (e.g., DDC and LCC) and embracing alternative classification. Two real-life case studies will be discussed to illustrate these methods. Note that Chapters 5 and 6 are organized similarly and that some instructions are repeated verbatim except when variation is necessary.
Introduction to library classification
Joudrey, Taylor, and Wisser define classification as “[t]he placing of subjects into categories; in organization of information, classification is the process of determining where an information resource fits into a given hierarchy and often then assigning the notation associated with the appropriate level of the hierarchy to the information resource and to its metadata.”[1] While other classification systems are available, this chapter focuses on the two predominant general-purpose classification schemes used in the United States: Dewey Decimal Classification (DDC) and Library of Congress Classification (LCC).
Introduced by Melvil Dewey in 1876,[2] DDC divides the entirety of knowledge into ten main classes, which are divided into narrower subtopics denoted by three-digit Arabic numerals. Topics are subdivided further with the addition of a decimal point followed by more numbers. LCC, though developed specifically for the Library of Congress’s collection around the turn of the twentieth century,[3] has become widely used by libraries throughout the world. The scheme consists of individual schedules created for various areas of knowledge and uses an alpha-numeric notational system.[4]
DDC and LCC, like all classification schemes, are products of the cultural paradigms of their creators and are subject to explicit and implicit biases. DDC’s Western-centric worldview, for example, is particularly apparent in the 200s—the section dedicated to religion. Christianity is spread throughout 220–289 in the schedule, while all other religions are crowded in the 290s (see Figure 4.1). As information professionals, it is critical that we strive to consider the diversity of our users in library classification. As Sahadath argues, “If we are not proactive in eliminating the inequalities in certain common classification schemes, we could end up alienating diverse and marginalized populations.”[5]
| 200 | Religion |
|---|---|
| 210 | Philosophy & theory of religion |
| 220 | Bible |
| 230 | Christianity & Christian theology |
| 240 | Christian moral & devotional theology |
| 250 | Christian orders & local church |
| 260 | Social & ecclesiastical theology |
| 270 | History of Christianity & Christian church |
| 280 | Christian denominations & sects |
| 290 | Other religions |
Working within DDC and LCC
Fortunately, DDC and LCC are not etched in stone; both have been continually updated to better reflect humanity’s ever-changing knowledge landscape, which includes changes for greater inclusivity. One of the easiest methods of enhancing the inclusivity of your catalog is to ensure that new resources are shelved under the most up-to-date classification numbers. DDC users can follow “025.431: The Dewey Blog” (https://ddc.typepad.com/) for updates and other information regarding the scheme. In addition, the Google Group Dewey Contributors (https://sites.google.com/view/deweycontributors) allows anyone to propose changes.
The Library of Congress’s Classification Web (https://classweb.org/approved/) lists monthly updates to LCC, and metadata creators can subscribe to receive notifications.[6] Libraries can also recommend changes to LCC via the Library of Congress’s Policy, Training, and Cooperative Programs Division (PTCP).[7] When adopting changes to classification, libraries should consider retroactive conversion projects to ensure that materials cataloged before the new policies took effect are up to date with current practices, and to prevent libraries from creating “split” collections in which materials that should be classed together are separate.
We recommend that all libraries using DDC or LCC follow several practices:
- Evaluate choices critically when assigning classification numbers.
- Follow updates to DDC and LCC.
- Connect with your stakeholders (i.e., users, library staff, community members, etc.) and involve them in the process when applicable.
- Know your options for recommending changes to existing schemes.
- Consider using alternative classification (see below) if the problem cannot be adequately addressed using DDC or LCC.
Case study: N-Cutter reassignment at the Iowa State University Library
An example of retroactively converting records to conform to recent developments was performed here at the Iowa State University Library. Some classification numbers in LCC use Cutter numbers, “a method of representing words or names by using a decimal point, followed first by a letter of the alphabet, then by one or more Arabic numerals,”[8] to collocate subtopics. For example, in the E–F schedule (used for history of the Americas) the range for elements of the population (i.e., demographic groups) of the United States is E184–E185.98. Most groups are classified with a Cutter number under “E184,” such as “E184.C97” and “E184.I8” being used for Cuban Americans and Italian Americans, respectively.
This concept was applied at various places in the schedules for African Americans with N-Cutter numbers. The “N” was assigned in the past because it stood for “Negro,”[9] which is no longer a preferred term for Black people. In June 2021 all existing classification numbers with N-Cutters were changed to B-Cutter (Black people). For example, the classification number D810.N4 (Negroes in World War II) was replaced by D810.B53 (“Black people. African Americans”).[10] (See Figure 4.2.)
D1 - D2027 -- History (General) D731 - D838 -- World War II (1939-1945) D810.A-Z -- Other special topics, A-Z |
The Iowa State University library took two actions to address this. First, cataloging unit members were instructed to change any newly acquired titles with N-Cutters in their existing MARC records to their B-Cutter equivalents. Some resources arrive at the library shelf-ready, meaning the records are supplied by vendors, and the physical items typically bypass the cataloging unit. To ensure that local practices were followed consistently, it was decided that all materials with subject matter related to DEI should be forwarded to the cataloging unit.
The second part of the process was to retroactively reclassify existing materials. To avoid creating a split collection—with legacy titles shelved under N-Cutter classification numbers and recent ones shelved under B-Cutter classification numbers—existing N-Cutter titles needed to be identified and assigned new classification numbers. We used an Alma Analytics report that was created by Mark K. Ehlert at St. Thomas University Library and shared with the Alma community.[12] The Analytics feature in Alma was used to retrieve titles with N-Cutter classification numbers in the subfield h ($h) of the 852 field in the MARC holdings record (see example in Figure 4.3). These titles were then retrieved from the stacks, and the books were brought to cataloging to be assigned new call numbers. Most of these obsolete classification numbers were corrected in batch, first by scanning the barcodes on the books to add them to sets and then by using normalization rules to programmatically change call numbers in both the bibliographic and holdings records.
In some cases, obsolete N-Cutter classification numbers were mapped to multiple numbers, thus requiring catalogers to decide which current number best suited the individual title. After the records were updated, the titles were sent to be relabeled and returned to the collection. For books that were out on loan during the initial conversion, the Iowa State University library added a note to the item record instructing circulation staff to forward the item to cataloging after it was returned so it could be updated. This work could serve as a model for similar retroactive conversion projects involving classification number reassignment.
| 852 0# | $a PARKS $c PSGEN $h D810.N4 |
|---|
Alternative classification
The process of updating comprehensive classification schemes can be lengthy. Another option may be to proactively embrace alternative systems. By locally customizing existing classification schemes—or even, as discussed in this chapter’s second case study, developing new ones—libraries can move beyond strict adherence to national standards and spearhead advancements in DEI and library classification. “For some libraries, it is possible to mitigate the shortcomings of dominant classification schemes by creating policy or by making local additions to the schemes themselves,” thus allowing for improvements to be made for specific subject matter rather than a rewriting of an entire scheme.[13]
Alternative classification can serve as an excellent method for concentrating on specific areas of marginalized communities or topics that are not well addressed by existing systems. An early example dates back to the 1930s with Dorothy Porter, a pioneering librarian at Howard University who adjusted her library’s use of DDC “to make Africana materials more easily accessible.” At that time, DDC placed all resources related to people of African ancestry under 325 and 326, the respective classification numbers for colonialism and slavery. Porter created a modified version of DDC to more effectively capture works related to Black people.[14] Another example is the Library and Archives Canada FC class, which is used for Canadian History, and its PS8000 class, which is used for Canadian literature. These extensions to the LCC schedules have “allowed for a more thorough and detailed organization of materials specific to Canadian collections, which are otherwise given less treatment in the basic LC scheme.” [15]
We recommend several steps for libraries considering alternative classification:
- Identify a subject area that may benefit from alternative classification.
- Research existing alternative classification schemes.
- Create a new system if existing alternative classifications are inadequate.
- Implement the new system into your library’s workflow and discovery layers.
The following section discusses several things to consider during this process.
Implementing alternative classification schemes
After your institution has done the necessary research and decided to use alternative classification, the next step is to incorporate the scheme into your systems and workflows. For consistency, it is important to document the circumstances in which to apply alternative headings, as well as how to do so. Below are some questions to assist with this process.
- Which records will be targeted for alternative classification?
Will you target all new records, all existing records, or records for selected collections or types of resources? Will the alternative classification system be used for electronic resources that won’t have a physical call number? - Which alternative scheme will be used?
If using an existing scheme, will it be used in full or in part? If using a locally developed system, or revising an existing one, a schedule will need to be created with a defined hierarchy and clear procedures. How will this be created? Will it be shared only with library staff, or will it be publicly available? - How will you identify the targeted records?
Will you use an automated process to identify records containing specified problematic classification numbers? When automation is not possible, you may wish to browse certain sections of your library’s shelf list. LC Correlations could be used to identify both LCC and DDC numbers commonly used for specific topics. - How will you update the targeted records?
To make the changes, will you use an automated process, such as an integrated library system (ILS) normalization rule that automatically changes the existing classification number to its corresponding alternative?[16] If so, will you use a batch editing tool such as MarcEdit, OpenRefine, or spreadsheets to make global changes? Moreover, will catalogers and metadata creators be instructed to update classification numbers manually in certain scenarios? - How will changes be made to call numbers for physical materials?
How will your library coordinate the updating of call numbers for physical resources? Who will need to be involved in the process? What shifting work will be necessary to ensure that all germane titles are shelved together under the new scheme? - How will records for new resources be addressed?
How will metadata and cataloging staff be instructed to use alternative classification numbers? Which library staff will need to be part of the process? For example, if many of the materials arrive at the library shelf-ready (i.e., they bypass the cataloging department), will receiving staff know the criteria for forwarding resources to cataloging? - Will the alternative system be revised in the future?
Will periodic reviews and revisions of the new classification system take place? Will the library have a system for reporting feedback to technical services? For example, if a user informs a public service librarian of an issue regarding the new system, does the library have a process of recording this for future consideration?
Case study: The Brian Deer Classification system
One example of alternative classification enhancing inclusivity is the Brian Deer Classification system (BDC). Brian Deer, the system’s eponymous creator, was a librarian and Mohawk. While earning his MLIS at McGill University, he determined that the existing mainstream classification systems were inadequate for describing Indigenous communities. In the 1970s he developed an early version of his system to organize the collection of the National Indian Brotherhood and Union of British Columbia Indian Chiefs (UBCIC) in Vancouver, British Columbia.[17]
Tomren discusses several advantages BDC has over mainstream schemes such as DDC and LCC. First, its relatively simple design allows for easier implementation, which assisted Indigenous libraries because they “often had a small staff and limited resources to devote to cataloging.”[18] It allows catalogers to use subclasses to organize narrower topics, such as allowing Indigenous communities to be collocated by “related cultures rather than being simply listed alphabetically as in LCC or grouped linguistically as in DDC.” Finally, it uses endonyms rather than exonyms to describe Indigenous communities.[19]
Another advantage of BDC is its malleability. Deer “believed the classification system used in a particular library should be designed to reflect the concerns of local Indigenous people.” He therefore did not attempt to develop a scheme that could “be applied to all topics of interest to the Indigenous peoples of North America,” and “created a new classification system for each library in which he worked.”[20] Other institutions have continued to update the system to meet their users’ needs.
An example of local revision took place when staff at the UBCIC Resource Centre, where Brian Deer had created an early version of the system,[21] found that multiple updates were needed. For example, the UBCIC acquired large numbers of resources related to environmental and ecological subjects, topics not covered sufficiently by their previous version of the BDC, and it was determined that subject expansion of the system was necessary to accommodate these newly acquired materials. In the previous version, for example, resources on coal developments were shelved under “Nc” (see Figure 4.4).[22]
N -- Energy & resources Nc -- Coal Developments |
After the revision, the subject was changed simply to “Coal” and given the more granular classification number “NRDE.FC” placing it under a more complex hierarchy (see Figure 4.5).[24]
N -- Nature & Ecological Knowledge NR -- Land-Based Resources NRD -- Land Use & Resource Development [Modern/ Contemporary] NRDE -- Energy NRDE.F -- Fossil Fuels NRDE.FC -- Coal |
Also, the top-level N section’s name was changed from “Natural Resources” to “Nature & Ecological Knowledge,” in order to “reflect the relationships Indigenous people have with the land and animals.” Similarly, “Wildlife Management” was replaced by “Wildlife Caretaking & Stewardship.”[26]
Another example of local revision is the Aanischaaukamikw Cree Cultural Institute (ACCI) in Oujé-Bougoumou, Quebec, home to an Eenou community (who are part of the larger Cree peoples), which adopted the system. The UBCIC adjusted the system to dedicate special attention to British Columbia by classifying communities from that province separately from other North American Indigenous communities. Influenced by the UBCIC, the ACCI initially planned a similar approach based on Quebec; however, its collection was also focused on other Cree communities—who are not limited to Quebec but spread throughout several Canadian provinces and a U.S. state. Because dividing works by modern political borders would separate related resources in a way that undermined the scheme’s intent, a decision was made not to separate communities located in modern-day Quebec from communities located elsewhere.[27]
Work on the BDC shows the process that goes into creating an alternative classification system, as well as the benefits it can provide. BDC focuses on Indigenous communities in North America, particularly Canada, but similar methods could be used for other marginalized communities, especially when libraries manage specialized collections with materials related to these groups.
Conclusion
This chapter introduced readers to issues related to DEI in library classification. Two approaches that can make discovery systems more inclusive were discussed: ensuring that the shelf listing of library resources is up to date with recent developments and embracing alternative classification systems. The updating of the LCC N-Cutter numbers at the Iowa State University Library and the development and revisions of the Brian Deer Classification system were highlighted to illustrate these methods in practice.
- 025.431: The Dewey blog
https://ddc.typepad.com/025431/ - Classification Web, Library of Congress
https://classweb.org/ - Dewey Contributors
https://sites.google.com/view/deweycontributors - SACO – Subject Authority Cooperative Program, Program for Cooperative Cataloging
https://www.loc.gov/aba/pcc/saco/ - University of British Columbia, Brian Deer Classification System
https://guides.library.ubc.ca/Indiglibrarianship/briandeer - WebDewey, OCLC
https://dewey.org/webdewey/login/login.html
- Daniel N. Joudrey, Arlene G. Taylor, and Katherine M. Wisser, The Organization of Information, fourth edition, Library and Information Science Text Series (Santa Barbara, CA: Libraries Unlimited, 2018), 626–27. ↵
- Joudrey, Taylor, and Wisser, The Organization of Information, 82. ↵
- “Classification and Shelflisting,” Library of Congress, last modified December 19, 2023, https://www.loc.gov/aba/cataloging/classification/. ↵
- “Libray of Congress Classification Outline,” Library of Congress, accessed May 29, 2024, https://www.loc.gov/catdir/cpso/lcco/ ↵
- Catelynne Sahadath, “Classifying the Margins: Using Alternative Classification Schemes to Empower Diverse and Marginalized Users,” Feliciter 59, no. 3 (2013): 15. ↵
- “Subscribe to E-Mail Newsletters and Alerts,” Library of Congress, https://loc.gov/subscribe/ ↵
- “Classification and Shelflisting,” Library of Congress. ↵
- “Glossary” in Classification and Shelflisting Manual, July 2013, 3, https://www.loc.gov/aba/publications/FreeCSM/glossary.pdf. ↵
- “Summary of Decisions, Editorial Meeting Number 2106,” Program for Cooperative Cataloging, Library of Congress, June 21, 2021, https://www.loc.gov/aba/pcc/saco/cpsoed/psd-210621.html. ↵
- “Monthly List 06 (June 18, 2021),” Classification Web, https://classweb.org/approved/2106.html. ↵
- Figure derived from “LCC Enhanced Browser,” Classification Web. ↵
- Mark K. Ehlert, reply to “‘N-cutters’ have been changed to ‘B-cutters,’" RADCAT@listserv.uga.edu, November 4, 2021, https://listserv.uga.edu/scripts/wa-UGA.exe?SUBED1=RADCAT. Although Iowa State used Ex Libris Alma to retrieve and update records, other Integrated Library Systems (ILSs) can perform similar functions. ↵
- Sahadath, “Classifying the Margins,” 15. ↵
- Janet L. Sims-Wood, Dorothy Porter Wesley at Howard University: Building a Legacy of Black History (Charleston, SC: The History Press, 2014), 39. ↵
- Sahadath, “Classifying the Margins,” 15. ↵
- In this context, normalization refers to “[t]he action of removing unimportant differences from data, especially text, to simplify subsequent processing.” “Normalization,” in A Dictionary of Computer Science, ed. Andrew Butterfield, Gerard Ekembe Ngondi, and Anne Kerr (Oxford: Oxford University Press, 2016), https://www.oxfordreference.com/view/10.1093/acref/9780199688975.001.0001/acref-9780199688975-e-6416. ↵
- Jean Weihs, “A Tribute to Brian Deer,” Technicalities 39, no. 3 (2019): 11–12. ↵
- Holly Tomren, “Classification, Bias, and American Indian Materials” (unpublished manuscript, 2004), 21, http://ailasacc.pbworks.com/f/BiasClassification2004.pdf. ↵
- Tomren, 21. ↵
- Weihs, “A Tribute to Brian Deer,” 12. ↵
- Weihs, 12. ↵
- Alissa Cherry and Keshav Mukunda, “A Case Study in Indigenous Classification: Revisiting and Reviving the Brian Deer Scheme,” Cataloging & Classification Quarterly 53, no. 5–6 (2015): 555, https://doi.org/10.1080/01639374.2015.1008717. ↵
- Hierarchy derived from Cherry and Mukunda, 558, fig. 1. ↵
- Cherry and Mukunda, 558, fig. 1. ↵
- Figure derived from: “UBCIC Resource Centre Classification Plan [Draft],” July 14, 2015, 19, 24–25, https://d3n8a8pro7vhmx.cloudfront.net/ubcic/pages/1166/attachments/original/1507335846/UBCICClassification.pdf?1507335846. ↵
- Cherry and Mukunda, “Indigenous Classification,” 556. ↵
- Annie Bosum and Ashley Dunne, “Implementing the Brian Deer Classification Scheme for Aanischaaukamikw Cree Cultural Institute,” Collection Management 42, no. 3/4 (2017): 280–293, https://doi.org/10.1080/01462679.2017.1340858. ↵