Bibliography Auditor Simulacrum

What The Tool Does

The Bibliography Auditor addresses a problem that has existed in a mild form for decades and in an acute form since the arrival of large language models: citation errors. Some are honest mistakes — a transposed volume number, a misremembered page range, an author initial dropped — and some are confabulations, citations to works that do not exist. The tool takes a bibliography or reference list and checks each entry against live bibliographic databases, flagging those that are verified, those that contain correctable errors, and those that appear not to exist at all.

The output is an annotated bibliography: each entry is marked *verified*, *error* (with a specific correction), or *unverifiable* (with a reason — no matching record in any database searched, author known but this title not attributed to them, journal exists but no article with this title in the issue cited). The tool does not make the corrections automatically; it reports them and leaves the final judgement to the writer.

Where The Method Comes From

This is a contemporary tool with no classical lineage. Before the LLM era, citation auditing was a manual task performed by reference librarians, journal copy editors, and the more conscientious peer reviewers. The field of bibliographic control — the systematic identification and organisation of published works — has existed since the nineteenth century, and tools like the Library of Congress catalogue, the International Standard Book Number system, CrossRef for DOIs, and large-scale databases like Google Scholar, WorldCat, and PubMed have made citation verification progressively more tractable.

What has changed recently is the scale of the problem. The arrival of fluent language models that generate plausible but fabricated citations — the hallucination problem — has made systematic bibliography auditing a necessity rather than a courtesy. The Bibliography Auditor is one of several responses to that situation, along with journal policies, editorial checks, and the retrieval-augmented generation systems increasingly embedded in research tools.

What It Can And Cannot Do

The tool can check bibliographic entries against CrossRef, WorldCat, PubMed, Google Scholar, and the JSTOR index, flagging mismatches and confabulations. It can handle most reference styles (APA, MLA, Chicago, Harvard, Vancouver) and can process long bibliographies efficiently. It is particularly valuable for checking work produced with AI assistance, for editing student papers before submission, and for auditing one's own reference list under deadline pressure.

It cannot verify the existence of unpublished materials, grey literature, archival sources, personal communications, or works only available in closed collections. It also cannot assess whether a correctly cited source actually supports the claim attributed to it — a citation can be real and still misrepresented. That remains the writer's responsibility.