Dead Letters in the Catalog: What Obsolete LCSH Headings Tell Us About the Library's Changing Soul

Every controlled vocabulary accumulates ghosts. In the case of Library of Congress Subject Headings, those ghosts are formally documented—recorded in authority files with the designation cancelled or deprecated, sometimes accompanied by a brief note explaining what replaced them and when. To read through a sampling of these obsolete terms is to move through a compressed history of American intellectual life, its assumptions, its blind spots, and the slow, often contested process by which institutions revise their own frameworks.

The study of abandoned LCSH headings is not merely an exercise in cataloging archaeology. It has direct consequences for contemporary researchers, digital preservation specialists, and the library systems that serve them. Understanding why certain headings fell out of use—and what happened to the records they once described—is essential for anyone working at the intersection of historical scholarship and modern information retrieval.

The Anatomy of an Obsolete Heading

Not all deprecated headings are deprecated for the same reasons. A useful taxonomy distinguishes at least three categories: headings retired because the terminology they employed became socially unacceptable; headings superseded because the subject matter was reorganized or renamed within the LCSH structure; and headings that were technically valid but functionally abandoned because researchers stopped using them as information-seeking behavior evolved.

The first category draws the most public attention. Headings that encoded racial, ethnic, or diagnostic terminology now considered offensive or clinically outdated have been the subject of sustained advocacy campaigns, formal proposals to the Library of Congress, and, eventually, official revision. The replacement of Illegal aliens with Noncitizens and Undocumented immigrants—a change finalized in 2021 after years of lobbying by library associations and immigrant advocacy groups—is among the most widely discussed recent examples. Less prominent but equally significant are the dozens of earlier revisions that replaced pathologizing language around disability, sexual orientation, and mental health with terminology more consistent with current medical and social science usage.

The second category is quieter but affects a larger volume of catalog records. When the Library of Congress reorganizes a subject hierarchy—splitting a broad heading into more specific ones, or collapsing a cluster of narrowly defined terms into a single broader entry—the retired headings do not simply disappear. They persist in older catalog records, creating a retrieval problem that can bedevil researchers working with legacy databases or conducting longitudinal studies of a topic across multiple decades of library acquisition.

Search Traffic and the Long Tail of Disuse

The third category—headings that are technically current but functionally orphaned—is perhaps the most revealing from a technological standpoint. Modern integrated library systems generate transaction logs that record, among other data, which subject heading searches are actually conducted by patrons. Analysis of these logs across multiple large academic library systems consistently reveals a long-tail distribution: a relatively small number of headings account for the overwhelming majority of subject searches, while thousands of authorized terms receive no queries whatsoever within a given year.

Some of this pattern reflects the obvious reality that certain subjects are simply more heavily researched than others. But a portion of the orphaned-heading phenomenon reflects something more interesting: a structural mismatch between the vocabulary LCSH uses to describe certain topics and the vocabulary that contemporary researchers employ when they formulate queries.

Consider the heading Aeronautics, Commercial. The term is precise and historically grounded, but a researcher today is far more likely to enter airline industry, commercial aviation, or even a specific airline's name into a discovery layer. If the system does not successfully map those natural-language queries to the authorized heading, the records cataloged under Aeronautics, Commercial become effectively invisible—not because the heading was deprecated, but because the gap between controlled vocabulary and colloquial usage has widened past the point where the system can bridge it automatically.

What Abandonment Reveals About Institutional Memory

The persistence of deprecated headings in older catalog records raises a question that library technologists and digital humanists are actively debating: should retrospective cataloging projects systematically update legacy records to reflect current authorized headings, or does the original cataloging itself constitute a historical artifact worth preserving?

Arguments for systematic updating emphasize access. A researcher searching under a current heading should, in principle, retrieve all relevant materials regardless of when they were cataloged. Leaving older records under deprecated terms creates artificial gaps in search results, effectively hiding materials from users who have no reason to know that the vocabulary once used to describe them has since changed.

Arguments for preservation emphasize historical integrity. The heading under which a record was originally cataloged reflects the intellectual context of its time. A monograph cataloged in 1955 under a heading that would be considered offensive today was processed by a cataloger working within the assumptions of that era; updating the record erases evidence of those assumptions. For historians of librarianship and information science, the original cataloging data is itself a primary source.

Most large research libraries have arrived at a pragmatic middle position: authority control processes automatically update headings in linked records when an authorized term is changed, but the change history is preserved in the authority file itself, allowing researchers who need it to trace the terminological evolution.

The Technological Pressure on Controlled Vocabulary

The emergence of full-text search, faceted navigation, and, more recently, machine learning–assisted discovery has altered the environment in which LCSH operates in ways that directly affect which headings remain useful and which drift toward obsolescence. When a researcher can retrieve a relevant document by searching any word that appears in its full text, the subject heading loses some—though not all—of its traditional retrieval function.

This shift has intensified internal debates at the Library of Congress and within the broader cataloging community about the appropriate scope and granularity of subject heading work. Some voices argue for a leaner, higher-level vocabulary that interoperates cleanly with linked data environments and modern discovery systems. Others contend that the depth and specificity of LCSH—including many of its less-trafficked headings—remains irreplaceable for precision retrieval in specialized research contexts.

What the graveyard of obsolete and orphaned headings ultimately discloses is not a system in decline but one in continuous, sometimes uncomfortable negotiation with its own history. Each deprecated term is a record of a moment when the Library decided that accuracy, or justice, or utility required changing the language it used to describe the world. Those decisions accumulate into something larger than any individual heading revision: a portrait of an institution trying, imperfectly and incrementally, to serve the researchers of the present without entirely abandoning the cataloging work of the past.