The DOI Needs a Chain of Custody

A DOI Is Also a Custody Claim

When a journal changes publisher, the visible event is usually commercial: a society signs a new agreement, a portfolio changes hands, or a platform contract ends. The scholarly record experiences something more specific. A new organisation becomes responsible for keeping the title identifiable, keeping old articles reachable, preserving the meaning of the DOI record, and responding when a published item needs correction or retraction.

That is the practical significance of Crossref''s draft Member Practices, now open for community feedback through October 9, 2026. Crossref says the final practices are planned for December and may ultimately provide grounds for suspending or revoking membership when a member shows a systematic pattern of irresponsible conduct or does not make reasonable efforts to improve after concerns are raised: https://www.crossref.org/blog/share-your-feedback-on-the-new-crossref-member-practices/.

The draft does not turn a DOI into a quality seal. Crossref is explicit that it does not assess the quality of published research or intervene in every individual dispute. But it does make the stewardship claim harder to ignore: the organisation registering the record is identifying itself as the party responsible for the title and its metadata. A transfer therefore has to move responsibility, not merely content files.

The Transfer Is More Than a New Website

The draft Member Practices asks members to register records only for content they own or are appointed to publish, avoid duplicate DOI records, keep DOI landing pages active and publicly accessible, include ISSNs where available, and give permission for a title transfer without unreasonable delay. It also says that a member should not resell DOI registration services through its own membership and prefix while making another organisation look like the publisher of record.

Each point sounds narrow until a journal moves between organisations. Then they become one dependency chain. If the receiving publisher registers new DOIs instead of taking responsibility for existing ones, citations split. If the old landing pages disappear, library links and author CVs start failing. If the ISSN, ownership, or appointed-publisher relationship is unclear, a metadata record can make the wrong organisation appear accountable. If a platform vendor is doing the deposits, the journal still has to know who owns the decision and who can repair the record.

NISO’s Transfer Code of Practice v5, published March 31, 2026, reaches the same problem from the continuity side. It gives transferring and receiving publishers guidance on digital assets, URLs, access rights, usage data, perpetual access, and ISSN/DOI information. NISO says the goal is to keep content accessible and minimise disruption to manuscript submission and production when a title moves: https://www.niso.org/publications/rp-24-2026-transfer.

Four Handoffs That Should Never Live in Email

1. Legal responsibility before the announcement

Before a transfer is announced, identify the journal owner, the appointed publisher, the Crossref member that currently holds the title, and the party that will hold responsibility after the effective date. Build an inventory of the journal’s ISSNs, DOI prefix, title-level records, article-level records, current resolution URLs, archive locations, and any content that is still in production. A contract can allocate rights between two parties, but it does not automatically update the public scholarly record.

2. DOI responsibility on the effective date

The receiving publisher should know which existing DOI records it will maintain and how it will update their URLs and metadata. The transferring publisher should know when it is expected to grant permission for the Crossref title transfer and which legacy pages or redirects must remain live. This is also the point to test whether submission, production, hosting, and deposit systems agree about the same effective date. A journal can be commercially transferred on Monday and still be technically split across two systems on Friday.

3. The record after a withdrawal or retraction

Continuity is not only a transfer concern. The Crossref draft says a withdrawn or retracted item should continue to resolve to an active landing page that makes the change of status clear; the original DOI should not simply become a 404. A DOI record for the post-publication update should be created and linked to the original item, and the original metadata should show that the item was removed. This is where a transfer file becomes an integrity file: the receiving publisher needs the authority and technical access to preserve the history of decisions made before it took over.

4. The request from a well-meaning institution

The draft also addresses a common edge case. A university that is a Crossref member may be asked by faculty to register a DOI for an article published in a journal owned by another member. Crossref says the university should not do that. It would be claiming responsibility for the journal title and its metadata without being the owner or appointed publisher. The repository can display the publisher’s DOI; if the institution hosts a genuinely separate preprint, it can register a preprint record when it is responsible for that repository.

What the Journal Should Be Able to Prove

A policy page saying that the journal values persistent access is not enough. During a transfer, a publisher should be able to open one evidence file and answer five questions without reconstructing the story from old inboxes:

• Who owns the title, who is appointed to publish it, and which organisation is responsible for the Crossref records?
• Which ISSNs, DOI prefixes, article records, URLs, and archive assets are in scope for the transfer?
• When will each party update resolution targets, title ownership, metadata, access arrangements, and public transfer notices?
• How will the journal preserve corrections, expressions of concern, withdrawals, and retractions created before or after the transfer?
• Who can approve and execute a repair when a DOI resolves incorrectly or the public metadata identifies the wrong steward?

The fifth question is often the one that exposes the real weakness. A vendor may perform the deposit, a platform may host the landing page, and a society may own the title, but the journal still needs a named decision-maker with authority across those systems. Stewardship cannot be outsourced into ambiguity.

Do Not Over-Read the Draft

Crossref’s proposal is important precisely because it is limited. The organisation says it will not decide the merits of a legal ownership dispute, judge the quality of research, or investigate every isolated failure. The draft is aimed at patterns: deceptive or fraudulent activity, systematic problems, and members who do not respond constructively when concerns are raised. It is a membership expectation, not a substitute for editorial judgment, a transfer contract, or a preservation plan.

That distinction should shape how publishers respond to the consultation. The useful question is not whether the draft guarantees that every transfer will be smooth. It is whether the expectations are clear enough for a small society, university press, platform provider, or large publisher to identify its own control gaps before a transfer or integrity case makes them public.

A Practical Audit Before 9 October

Choose one journal that has changed ownership, platform, or publishing partner in the last three years. Start with a DOI from the archive, not the newest article. Follow it through the resolver, article landing page, PDF or HTML, Crossref metadata, ISSN record, and any public transfer notice. Then test a correction or retraction path, even if the title has not needed one. Record who owns each step, what breaks when a URL changes, and which system can update the public record.

If the chain is intact, the exercise gives the journal evidence it can use in a transfer negotiation or board review. If it is not intact, the journal has found a continuity project with a clear scope. Crossref is accepting feedback on the draft Member Practices until October 9, and that deadline is a useful forcing function even for organisations that do not plan to submit comments.

Practical Takeaway for Journal Leaders

Treat every journal transfer as a chain-of-custody event for the DOI record. Before changing platforms, vendors, or publishers, name the responsible steward, inventory the existing identifiers, test resolution and status-update paths, and preserve the evidence that lets a reader understand what changed. The DOI is persistent only when the organisation behind it keeps behaving like a steward.