Editorial policies

The rules, in full

These policies govern every journal, book and article on this platform. They follow the COPE Core Practices, and where this page and a journal's own summary disagree, this page wins.

Peer review

Model: Double-anonymous by default: reviewers do not know who wrote the manuscript and authors do not know who reviewed it. Open review is available when both parties opt in, in which case reports are published, signed, alongside the article.
How many: Two reviewers minimum for research articles. Where reports conflict on a material point, a third is sought rather than the editor casting a deciding vote.
Who: Matched on subject and on method. Author-suggested reviewers may be considered but never constitute the required minimum, and suggested opposed reviewers are honoured unless the field is too small to allow it.
Editor submissions: A manuscript from an editor of the journal is handled entirely by another editor, and the submitting editor has no visibility of the file at any stage.
Timeliness: Every journal publishes its median time to first decision and updates it quarterly, including the quarters it worsens.

Authorship

Criteria: Substantial contribution to conception, design, acquisition, analysis or interpretation; involvement in drafting or critical revision; approval of the submitted version; and agreement to be accountable for the work.
Not authorship: Funding acquisition, general supervision, and provision of materials or a dataset are contributions and belong in the acknowledgements, not in the author list.
Order: Decided by the authors. We do not arbitrate order, but we do require that every author has agreed to the order submitted.
Changes after submission: Require written agreement from every author, including anyone being removed, and a reason. Author lists that change between submission and acceptance are a documented misconduct signal and are treated as one.
Corresponding author: Responsible for communication, for confirming that co-authors have approved the version, and for responding to queries after publication.
Group authorship: Consortium and group names are supported, with the individual members listed and indexed so each is individually citable.

Competing interests

A competing interest is anything a reasonable reader might think could have influenced the work. Declaring one is almost never disqualifying. Failing to declare one usually is.

Authors: Declare at submission: funding, employment, consultancies, patents, holdings, board memberships, and personal or political relationships bearing on the subject. Every article carries a statement, including one saying there are none.
Reviewers: Declare at invitation. Same institution, recent collaboration, a competing manuscript in preparation, or a personal relationship with an author.
Editors: Declare at handling. An editor never handles a submission from their own institution, a recent collaborator, or a former student within five years.
The publisher: Has no role in any editorial decision, cannot see a manuscript file without the handling editor's knowledge, and has no mechanism by which to overturn a decision.

Research ethics

Human participants: Approval from a named ethics committee, with the reference number, plus a statement on informed consent. Research without approval is not published here retrospectively.
Consent to publish: Required for any identifiable individual, including in images and case reports. Anonymisation by cropping is not anonymisation.
Animal research: Institutional approval, the guidelines followed, and ARRIVE-compliant reporting.
Vulnerable participants: Additional scrutiny where participants are children, prisoners, patients of the researcher, or otherwise unable to refuse freely.
Field research: Permits, and — where the work is conducted in a community other than the researchers' own — an account of how the community was engaged and what it gets back.
Dual use: Where findings could enable serious harm, we consult the ethics team before publication and, rarely, decline to publish detail that adds nothing scientific.

Plagiarism & research integrity

Screening: Every submission is screened with iThenticate at desk stage. The report is read by an editor rather than acted on by a threshold — a heavily-cited methods paragraph is not misconduct, and a lightly-paraphrased introduction may be.
Self-plagiarism: Reusing your own text without citation is still a problem, though a smaller one. Reusing your own data across papers without disclosing the overlap is a larger one.
Image integrity: Figures are checked for duplication, splicing and inappropriate adjustment. Adjustments to brightness or contrast must be applied to the whole image and disclosed.
Data fabrication: Allegations go to the research integrity team, follow the relevant COPE flowchart, and are put to the author in writing before any action.
Paper mills: We screen for the known signals — templated text, implausible authorship changes, citation rings, tortured phrasing — and we will decline on that basis alone, saying so.
Citation manipulation: Editors may not require citation of their own work or of the journal. Reviewers may suggest citations only where directly relevant, and their suggestions are visible to the editor.

Generative AI

Authors: AI tools may be used for language editing and must be disclosed in the methods or acknowledgements. They may not be listed as authors, because an author must be accountable and a tool cannot be. You remain responsible for every claim and every citation in your text — including the ones a tool invented.
Reviewers: Do not paste a manuscript into a public AI tool. It is a confidentiality breach regardless of the tool's data policy. Using AI to polish your own report is fine and should be disclosed to the editor.
Editors: The same confidentiality rule applies. No editorial decision is made by, or with the assistance of, an automated system.
AI-generated images: Not permitted as data. Permitted for illustration where disclosed and where the caption says so.

Data availability

A data availability statement is required on every research article, including one that says data are not available and why. See Data, code & protocols for the practical guidance.

“Data available on request” is accepted only with a named contact who has agreed in writing to be it. The phrase is, in the published literature on this exact question, overwhelmingly a way of saying no.

Corrections & retractions

Correction: For errors that do not affect the conclusions. Published as a linked notice; the article is amended and the amendment is dated and described.
Retraction: For errors or misconduct that invalidate the conclusions. The notice names the reason and who requested it. The article stays online, watermarked — removing it would break every citation and hide what happened.
Expression of concern: Where a serious question exists but an investigation is unfinished. Replaced by a correction, a retraction or a withdrawal of the concern once it concludes.
Who can raise one: Anyone, including anonymously. We assess the substance of a claim rather than the standing of the person making it.
Timeline: Acknowledged within five working days. Where an institution is investigating we generally wait for it, unless readers are being actively misled in the meantime.

Appeals & complaints

Appealing a decision: One appeal per decision, to the Editor-in-Chief, within sixty days. Address the reviewers' specific points; an appeal that restates the paper's importance will not succeed.
What happens: The Editor-in-Chief may uphold the decision, seek a further review, or reopen. They may consult a board member who has not seen the file. Answered within twenty-one days.
Complaints about process: Handling delays, conduct, or a policy applied inconsistently — to the editorial director, who is outside the journal and answers within fourteen days.
If we get it wrong: We say so and we say what we changed. A publisher that has never published a process failure has either never had one or does not publish them.

Raise a concern

Copyright & licensing

Open access articles: You retain copyright. The article is published under CC BY 4.0 or CC BY-NC 4.0, as you choose at acceptance.
Subscription articles: You grant us an exclusive licence to publish; copyright stays with you. You keep the right to deposit the accepted manuscript in a repository immediately, with no embargo.
Books: Terms are set in the contract and explained clause by clause before signature. Open access options are available for monographs with institutional or funder support.
Third-party material: Permissions are the author's responsibility, and we check that they exist before production rather than after publication.
Text & data mining: Permitted for all content, by licensed institutions and by anyone at all for the open access portion.