A Correction Notice Is Not a Footnote

The awkward part of correcting the scholarly record is not always the editorial decision. Often the harder part comes five minutes later, when the journal has to make that decision visible everywhere the article now lives. The article page needs a status signal. The PDF may need a label. The correction notice needs a DOI. Crossref metadata needs an update relationship. Indexers, libraries, reference managers, repositories, and readers all need a path from the old claim to the current record.

Crossref made that path easier to start this week. On August 20, 2026, it announced that Metadata Manager now supports richer registration workflows for books and chapters, reports, dissertations, and post-publication updates for journal content: https://www.crossref.org/blog/registering-rich-metadata-for-books-chapters-reports-dissertations-and-updates-just-got-easier/. Crossref said the tool lets members work without direct XML editing and can capture contributors, affiliations, funding information, references, and update metadata. The feature sounds procedural. For journals, it is a prompt to revisit how a correction becomes part of the record rather than a note sitting beside it.

That distinction matters because readers increasingly meet articles away from the publisher site. A PDF saved before a correction, a citation pulled from a reference manager, a search result, a discovery index, and a systematic review screening tool may all carry different pieces of the same history. If the update is only a page on the journal site, the record depends on a reader arriving in exactly the right place.

The Update Has To Travel Without The Editor

Editors are good at explaining what changed. Infrastructure is good at carrying structured relationships. A healthy correction workflow needs both. The notice can say that a figure legend was wrong, that an author affiliation has been amended, that a statistical value was corrected, that an expression of concern remains unresolved, or that the article has been retracted. Metadata must then express the relationship in a way machines can follow.

Crossref guidance on registering updates describes the usual pattern: when a significant update occurs, the publisher issues a separate correction or retraction notice with its own DOI and metadata, and that notice points back to the original article DOI: https://www.crossref.org/documentation/register-maintain-records/maintaining-your-metadata/registering-updates/. Crossref also notes that the Crossmark update field can be used to register update relationships even when a publisher has not implemented the full Crossmark button experience.

The operational implication is simple but often missed. A correction notice is not just editorial prose. It is a publication object. It needs a landing page, a DOI, a publication date, a relationship to the affected article, and enough metadata for external systems to recognize what kind of update it is.

Metadata Manager Lowers The Threshold

Many smaller journals have avoided richer update metadata because XML deposits are intimidating, vendor capacity is uneven, or staff touch Crossref records only when something breaks. Metadata Manager does not remove the need for governance, but it lowers the technical threshold for a journal team that needs to register an update correctly without building a custom deposit pipeline.

Crossref's August 11 community explanation is useful because it is direct about the workflow. The new option is not a separate content type in the broad sense; it is a way to indicate that a registered publication has been updated, most commonly for retractions, corrections, withdrawals, and errata: https://community.crossref.org/t/ticket-of-the-month-june-2026-post-publication-updates-in-metadata-manager/16253. The DOI being registered is the DOI of the update notice, while the DOI of the original article appears in the notice metadata as the item being updated.

That may sound like a deposit detail, but it protects the chronology of the record. The original article remains identifiable. The notice remains citable. The relationship between them becomes queryable. A reader, indexer, or integrity analyst should not have to infer from a title string that one object changes the status of another.

The Notice And The Article Are Two Records

Journals get into trouble when they treat a correction as a replacement paragraph. A production editor fixes the HTML. A PDF is silently swapped. A footnote is added to the article page. The editorial team may feel the matter is resolved because the public page looks current. Downstream systems may still see the old article record with no durable update relationship.

The better model keeps the notice and the article distinct. The article page should tell readers that the record has changed. The notice should explain what changed and why, using neutral language and appropriate editorial detail. The DOI metadata should connect the notice to the article. The PDF and HTML should carry visible status cues where appropriate. Internal case files should preserve the evidence and approvals that led to the public action.

NISO CREC, the Recommended Practice for communicating retractions, removals, and expressions of concern, frames this as both metadata and display work. NISO describes the goal as consistent, timely transmission of status information to human and machine readers across the publishing and discovery ecosystem: https://www.niso.org/standards-committees/crec. COPE has also endorsed the NISO CREC Recommended Practice: https://publicationethics.org/guidance/endorsed-guidance/niso-crec-recommended-practice.

Policy Pages Are Part Of The Deposit

A journal cannot improvise its update rules every time a problem appears. Crossref participation guidance for Crossmark starts with an update policy page: a single page explaining how content is updated after publication, assigned a DOI and registered so it can be persistently linked: https://www.crossref.org/documentation/crossmark/participating-in-crossmark/. The policy must cover corrections, retractions, withdrawals, and other updates.

This is not bureaucracy for its own sake. A DOI for the policy gives the journal a stable reference point when it registers content, publishes notices, trains editors, answers author questions, and handles platform migrations. It also forces the journal to define what happens to corrected and retracted works. Does the original remain available? Is it watermarked? Which title prefix is used? Where does the notice appear? How are readers told about the change in HTML and PDF?

The policy page should not be a graveyard of legal phrasing. It should be operational enough that a managing editor can follow it on a busy publication day. It should name update categories, minimum notice content, approval roles, display expectations, metadata registration steps, and the handoff to indexing or discovery partners.

The Missing Step Is Internal Ownership

Tools make updates easier to deposit. They do not decide who owns the chain. In many journals, a correction touches at least four groups: editorial, production, platform support, and metadata or DOI administration. Research integrity or legal review may also be involved. If each group completes its local task without a shared checklist, the public record can still split apart.

• The editor approves a correction, but nobody creates a separate notice DOI.
• Production updates the article page, but the PDF downloaded last month has no status signal.
• Metadata staff deposit the notice, but the original article title and landing page do not indicate that the record changed.
• The journal publishes an expression of concern, but the policy page does not define when it should be replaced, withdrawn, or escalated.
• A platform migration preserves article DOIs but drops the update-policy relationship from older records.

These are not rare edge cases. They are predictable handoff failures. The fix is not to ask every editor to become a metadata specialist. The fix is to make the correction workflow boring enough that staff can run it under pressure.

A Practical Drill For Journal Leaders

Choose one published correction or retraction from the last two years and trace it from decision to discovery. Start with the editorial file. Then inspect the notice page, original article page, PDF, Crossref record, Crossmark status if used, indexing feeds, and any repository or aggregator pages your team can check. Write down each place where the update appears, each place where it does not, and who could fix it.

Then run the same exercise prospectively. Pick a hypothetical correction accepted today and assign names to each step: drafting the notice, assigning the notice DOI, updating the article page, marking the PDF, registering the update relationship, checking the DOI metadata, notifying indexers if needed, and documenting the case closure. If one step has no owner, it is not a workflow. It is a hope.

Crossref has made it easier for journals to register post-publication updates without wrestling directly with XML. That is useful, especially for teams with limited technical support. But the leadership task is larger than learning a new form. Treat every correction notice as a publication object with a lifecycle, not as a footnote attached to an embarrassed article. The scholarly record corrects itself only when the correction can travel.