Special Issues Need a Control Room

Special issues used to sound like editorial optimism: a timely theme, a respected guest editor, a call for papers, and a concentrated set of articles around a field that was moving quickly. That model still has value. It can give smaller communities a place to gather work, help readers see a research area in one view, and let journals cover subjects that the standing editorial board cannot follow in equal depth.

The risk is that the same structure also delegates power at speed. A journal invites outside editors, opens a themed channel, attracts a burst of submissions, and asks existing staff to trust processes that may be less familiar than the regular manuscript stream. If the controls are weak, the problem is not one bad paper. It is a batch problem.

That is the lesson from the latest retraction news. On August 12, 2026, Retraction Watch reported that Elsevier's International Journal of Biological Macromolecules had retracted 120 papers since June after identifying manipulated peer review in guest-edited special issues: https://retractionwatch.com/2026/08/12/biology-journal-pulls-100-articles-and-counting-from-special-issues-for-peer-review-manipulation/. On August 20, Retraction Watch reported that Ceramics International had pulled 56 papers so far this year for concerns including image overlap, data-integrity issues, and suspicious authorship changes: https://retractionwatch.com/2026/08/20/ceramics-journal-pulls-dozens-of-papers-flagged-for-integrity-concerns/. These are not identical cases, and a journal should be careful about drawing more from them than the public record supports. The shared operational point is clear enough: themed publishing lines need stronger live supervision than many journals have historically given them.

The Pattern Is No Longer Ambiguous

The scholarly community has been warning about this exact pressure point. In July 2025, STM announced new COPE and STM resources on best practices for guest-edited collections, noting that such collections had been accompanied by coordinated attempts to manipulate the publishing record and fraudulent behavior by some guest editors: https://stm-assoc.org/new-resources-on-best-practices-for-guest-edited-collections/. The COPE guidance is public, and STM said its companion how-to material emphasized validating guest editors before projects begin and conducting audits during and after peer review.

That sequence matters. Vetting only at proposal stage is not enough. A special issue can look plausible when it is approved and still become risky once submissions arrive, reviewer suggestions cluster, timelines compress, authorship changes appear, or guest editors begin handling work at a volume that no ordinary editor could inspect carefully.

The control question should therefore change from "Do we trust this guest editor?" to "Can we see what this collection is doing while it is doing it?" Trust is personal. Governance is observable.

Guest Editing Is Delegated Authority

Many journals describe guest editors as advisors or curators. In practice, they may influence scope, invitation lists, reviewer choices, revision expectations, acceptance recommendations, and the tempo of an entire collection. That is delegated editorial authority, even when the final decision formally remains with the editor-in-chief.

Delegated authority needs a written boundary. A strong special-issue setup names what the guest editor may do, what they may recommend, which decisions require a standing editor, how conflicts are declared, how self-submissions are handled, what reviewer sources are allowed, and when the publisher may pause the collection. It also names what evidence must exist in the manuscript record.

The last point is often missing. A policy that says guest editors must avoid conflicts is useful, but a journal still needs a record of conflict checks, reviewer selection, editorial overrides, and exceptions. If concerns emerge later, the publisher should not be reconstructing the history from email fragments and staff memory.

Make Volume a Review Trigger

Volume is not misconduct. High submission numbers can reflect a real field, a popular topic, or a journal with global reach. But sudden volume changes should trigger review because manipulation often needs throughput. The International Journal of Biological Macromolecules case is a reminder that a publication spike changes the integrity workload, whether or not the spike is itself suspicious.

Journal leaders should define thresholds before the call for papers opens. If a collection receives twice the expected submissions in a month, who reviews the pattern? If one guest editor handles an unusually high share of papers, what happens? If acceptance times fall below the journal norm, who checks reviewer depth? If author affiliations, reviewer domains, or topic clusters look unusually concentrated, is there a pause mechanism?

These triggers should be boring and automatic. They are not accusations. They are management signals, the editorial equivalent of reconciling deposits before closing an issue. A journal that waits for PubPeer comments, sleuth emails, or retraction notices has already moved the audit downstream of publication.

The Reviewer List Needs Its Own Fence

Peer-review manipulation frequently exploits reviewer identity, reviewer suggestion workflows, private email addresses, compromised accounts, weak editor checks, and networks of friendly or fabricated reviewers. Special issues add another route: guest editors may bring legitimate networks, but those networks can also narrow the reviewer pool until independence becomes difficult to judge.

A practical control is to treat reviewer selection for guest-edited collections as a separate governance surface. Require institutional or persistent identity checks for new reviewers where feasible. Flag repeated reviewer use within a collection. Compare reviewer domains against author domains and prior coauthorship when possible. Require standing-editor approval for reviewer suggestions supplied by authors or guest editors in higher-risk cases. Preserve the reason a reviewer was chosen, not only the invitation record.

None of this means a journal should distrust every guest editor. It means reviewer independence is too important to leave inside a black box. The standing editorial office should be able to explain why the review pathway was credible for any accepted special-issue article.

Audit During the Issue, Not After It

Post-publication audits are necessary, but they are expensive, public, and slow. They also place correction work on readers, indexers, authors, and institutions after the article has already entered the record. A better model inserts light but regular audits while the collection is active.

• At proposal approval, verify guest-editor identity, affiliations, publication history, conflicts, and topic fit.
• At call launch, publish the collection scope, editorial responsibility, conflict rules, and expected review standards.
• During review, sample manuscripts for reviewer independence, timeline anomalies, authorship changes, image or text-integrity signals, and decision consistency.
• Before publication, require a standing editor to review exception reports and high-risk accepts.
• After closure, archive a collection-level audit note that records volume, acceptance rate, escalations, and unresolved concerns.

This does not require every society journal to build a large investigations unit. It does require the collection to have a dashboard, a responsible owner, and a stop button. A special issue without a stop button is not an editorial project. It is a production lane with a theme.

Shared Screening Is Helpful, But It Is Not Governance

Industry infrastructure can help. The STM Integrity Hub describes itself as a cloud-based environment for publishers to check submissions for research-integrity issues, using patterns associated with paper mills and other concerns while respecting data privacy and competition-law constraints: https://stm-assoc.org/what-we-do/strategic-areas/research-integrity/integrity-hub/. Shared signals are valuable because bad actors do not organize themselves around one publisher's internal workflow.

But screening tools cannot decide a journal's appetite for delegated editorial authority. They cannot set the conflict policy, define collection caps, choose the escalation owner, or explain to authors why a manuscript pathway was paused. Those are governance choices. Technology should surface patterns early enough for editors to act, not become the excuse for editors to look away.

For platforms such as Publicator, the useful role is not to promise that a system can certify integrity. It is to make governance visible: AI-assisted submission checks, reviewer matching with oversight, role-scoped access for guest editors, audit trails for reviewer invitations and decisions, analytics that flag unusual collection patterns, and DOI/Crossref-ready metadata that can support correction work if an issue later requires it. The product value is strongest when it helps a journal prove what happened, who decided, and which exceptions were reviewed.

A Practical Takeaway For Journal Leaders

Before approving the next special issue, run a collection-risk review in writing. Keep it short enough that editors will actually use it. Name the guest editors, their verified affiliations, conflicts, expected article volume, reviewer-selection rules, self-submission policy, decision authority, audit cadence, pause triggers, and closure report owner.

Then apply the same review to every active collection. Do not wait for a scandal to discover that one guest editor is handling too much, one reviewer network is overused, or one themed issue is accepting papers faster than the journal can inspect them. The goal is not to make special issues rare. The goal is to make them legible.

Special issues are not inherently suspect. Weakly governed special issues are. The journals that keep the model healthy will be the ones that treat guest editing as delegated power, monitor volume as an integrity signal, fence reviewer selection, and leave an auditable record while the collection is still alive. Once the retraction notices begin, the control room has already opened too late.