Pseudonymous and Anonymous Authorship

Pseudonymous authorship differs from anonymous authorship in traceability, not disclosure. A pseudonymous author publishes under an invented name the journal can still trace to a real person, typically still accountable under the ICMJE authorship test; an anonymous byline carries no attributable name at all. Both are narrow, conditional exceptions to named authorship.

The operational distinction: pseudonymous vs. anonymous

Three features separate a legitimate pseudonym from true anonymity in scholarly publishing:

• Consistency. A pseudonymous author typically publishes under the same invented name across multiple works, building a durable, citable scholarly identity distinct from — but linked to — their legal one. A single anonymous publication has no such continuity by definition.
• Traceability to the venue. The Committee on Publication Ethics (COPE) is explicit that even where a paper is published with no author names at all, the expectation is that “the identity of the authors is known to the journal but the names have been omitted” — the editor retains a real, accountable identity on file even when readers never see it. A pseudonym with no editor-side record of the real author is not what COPE’s guidance describes; it is closer to a false name submitted without permission, which COPE treats as unacceptable (see below).
• Retained accountability. Whether pseudonymous or anonymous, COPE’s position is that the underlying accountability obligation ICMJE places on authors does not disappear — it is just held privately by the editor instead of publicly on the byline.

Where pseudonymous or anonymous authorship is legitimate

Publication-ethics guidance recognizes a narrow set of circumstances where withholding or replacing an author’s real name is acceptable:

Safety

COPE’s position on handling requests to publish anonymously states that “under special circumstances, anonymising the author or agreeing to the use of a pseudonym is permissible, for example if their safety would be compromised” by publishing their real name, adding that “the safety of humans comes before specific publishing ethics issues.” This is the scenario most often cited in practice: a researcher working on a politically sensitive topic — human rights documentation, research conducted under an authoritarian government, or subject matter that puts the author at risk of arrest, violence, or retaliation in their home jurisdiction — may be permitted to publish under a pseudonym or with the name withheld, provided there is a clear public interest in the work being published at all. COPE recommends journals maintain an explicit policy for this rather than deciding case by case, and that even when a pseudonym is used, the editor-in-chief should still know the author’s real identity.

Historical precedent

Pseudonymous scientific publication has real precedent that predates any modern editorial policy. The statistician William Sealy Gosset published the majority of his output — including the 1908 paper deriving what is now called the Student’s t-distribution — under the pseudonym “Student,” at his employer Guinness’s request that staff not publish scientific work under their own names. Nineteen of Gosset’s twenty-one publications appeared under that name; Guinness knew his real identity throughout. In mathematics, Nicolas Bourbaki is a collective pseudonym adopted since the 1930s by a rotating group of (mostly French) mathematicians publishing a shared, multi-volume treatise on the foundations of modern mathematics — a case of durable, consistent pseudonymous authorship used for a collaborative rather than a safety reason. Neither case involved concealment from a journal or publisher; both are pseudonymous, not anonymous, authorship by the operational distinction above.

Not the same as double-anonymized peer review

Pseudonymous and anonymous authorship are easily confused with double-anonymized peer review, but the two are distinct mechanisms operating at different stages and for different purposes. Double-anonymized (and single-anonymized) review withholds author and/or reviewer identities from each other only during the review process, specifically to reduce bias in the evaluation of the manuscript — the journal itself always knows who the author is, and the author’s real name is restored on the published version once the paper is accepted. Pseudonymous or anonymous authorship, by contrast, is a decision about what appears on the permanent, published record, and it persists after acceptance rather than being lifted at publication. A paper can go through fully identified peer review and still be published pseudonymously (if a late safety concern arises, for instance), and a paper reviewed under double-anonymization is, in the overwhelming majority of cases, published under the authors’ real names.

Where it becomes a research-integrity problem

COPE draws a firm line between the safety-motivated exception above and reasons it explicitly rejects. In a published case discussion, COPE Council considered a manuscript in which one listed author used the surname “999” and several co-authors gave “independent researcher” affiliations to distance the work, submitted outside their employment, from their actual employer. COPE’s council was unanimous that this did not justify a pseudonym: “transparency and accountability are key in scholarly publishing,” and readers are entitled to know who is responsible for a paper’s claims. The same guidance is explicit that being in a dispute with an employer, wanting to distance oneself from an employer, general privacy preferences, or the fact that the content is controversial are not acceptable reasons for a pseudonym — and that it is not appropriate for authors to submit under a false name without the express permission of the editors. Practically, COPE also notes that pseudonyms complicate routine editorial infrastructure, such as linking a persistent identifier like an ORCID iD to the correct author record.

A separate and more serious integrity failure is the use of fabricated identities to manipulate the review or publication process itself — not a disclosed pseudonym on a byline, but an undisclosed fake identity used to deceive editors. Two documented cases illustrate the pattern:

• In 2012, South Korean researcher Hyung-In Moon was found to have submitted manuscripts to journals that allowed authors to suggest their own peer reviewers, then supplied fabricated reviewer names and email addresses he controlled himself, writing favorable reviews of his own papers. The scheme was uncovered after an editor noticed reviews were being returned within 24 hours, and led to 28 retractions.
• In 2014, SAGE Publications retracted 60 articles from the Journal of Vibration and Control after a 14-month investigation uncovered a peer-review and citation ring involving as many as 130 fabricated or impersonated reviewer identities, organized to generate favorable reviews for a specific author and his associates — who in at least one instance reviewed his own paper under an assumed identity.

These are sometimes loosely described as “sockpuppet” reviews. They are categorically different from the legitimate pseudonymous/anonymous authorship discussed above: a safety-motivated pseudonym is disclosed to and verified by the editor, while a sockpuppet reviewer identity exists specifically to deceive the editor and the review system. Fabricated identities are also a recurring feature of organized paper-mill activity, which COPE’s relaunched paper mills working group tracks as a coordinated, evolving threat rather than isolated incidents — distinct from, and more serious than, an individual researcher’s disclosed request for anonymity.

What a journal considering a pseudonym or anonymity request should do

• Have a written policy in advance rather than deciding requests ad hoc.
• Confirm the requesting author’s real identity and retain it on file, even if it will not be published.
• Reserve approval for cases with a genuine safety justification and a clear public interest in publication — not employer disputes, general privacy preference, or a wish to distance the work from an affiliation.
• Ensure the editor-in-chief specifically (not just a handling editor) is aware of the real identity.
• Anticipate downstream friction: persistent-identifier linkage (ORCID), indexing, and post-publication correction or retraction processes all assume a stable, identifiable author record.

Worked examples

Example 1. A statistician publishes a series of technical papers under a consistent invented name at their employer’s request, rather than under their own; the employer and the papers’ venues know the author’s real identity throughout, and the same invented name is used consistently across many publications. This is pseudonymous authorship — durable, consistent, and fully traceable to a real, accountable person even though the public byline differs from that person’s legal name. (Illustrates the pattern documented in Gosset’s “Student” publications, above.)

Example 2. A researcher submits a completed, ethically-conducted study on a politically sensitive topic and requests that their name be withheld because publication under their real name would expose them to a credible risk of arrest or violence in their home country. The journal verifies the researcher’s identity privately, confirms a clear public-interest case for publishing the findings, and publishes the paper with a note that the author’s identity is known to the journal but has been withheld. This matches COPE’s stated safety exception and is legitimate anonymous authorship.

Counter-example. A group of authors, wanting to keep a study they conducted outside their day jobs disconnected from their employer, asks to publish under invented surnames and list themselves only as “independent researchers,” with no safety concern involved. This is the actual scenario COPE Council considered and rejected: distancing a paper from an employer, without a safety justification, is not an acceptable basis for a pseudonym. The appropriate tool for that situation is a conflict-of-interest or affiliation disclosure statement under the authors’ real names, not a pseudonym.