“Protect the anonymity of individuals raising concerns.” The line appears in AIP Publishing’s guidelines for editors and reviewers, in the same passage that requires editors to respond to reported concerns in accordance with COPE guidelines. Thomas Harvey, a postdoctoral researcher in high energy theory at MIT, raised a concern with the editors of Physics of Fluids, an AIP journal. According to Retraction Watch’s October 13 report, the editors passed that concern to the paper’s authors along with his name, his institution, and the fact that he was a postdoc.
Obviously we screwed up.
The authors’ reply came back through the journal. It suggested that Harvey be “encouraged to withdraw” his criticisms, and it told him what had happened to the last junior researcher who crossed them.
Protect the anonymity of individuals raising concerns.
A cosmology paper in a fluids journal
The paper is “An alternative cosmological model for an expanding universe,” published in 2025 by William George, an emeritus professor at Chalmers University of Technology in Sweden who spent his career on turbulence, and Thomas Johansson, who is also retired from Chalmers. The model claims to do away with the need for dark matter and dark energy. Harvey sat in on George’s talk about it at the American Physical Society’s summit in Denver in March and judged it “very clearly wrong.” He posted his objections under a pseudonym on PubPeer, then sent what he calls a “polite email” to the journal’s editors.
His main objection is narrow and checkable. Harvey says a central equation contains an algebraic error, one he calls “elementary,” and that once the error is fixed the model describes a universe with nothing in it. George conceded the algebra on PubPeer, Retraction Watch reports: “Obviously we screwed up.” He promised an erratum, maintained that the rest of the analysis holds, and added, “We are surely not withdrawing the original paper.”
The reply that reached Harvey through the journal went well beyond the algebra. George and Johansson wrote that if his criticisms were published, they would make them public and “take our chances in the public arena.” They described an earlier critic, a “similarly aggressive young post-doc,” whose comments they said they had presented “in every forum for several years.” In their initial response they called their critics “apostles of the status quo.”
Harvey said the reply “felt a bit like a threat.” George, writing to Retraction Watch, called it “certainly not a threat” and said he wanted Harvey to understand that “things you say in print last forever.” Nothing in the reporting settles which reading is right. What it does establish is the sequence of events. The journal attached a junior researcher’s name and rank to his critique and sent it to two senior authors, and those authors answered by recounting how they had treated the previous junior researcher who challenged them. A comment posted under George’s name beneath the Retraction Watch story says he learned Harvey’s rank from Harvey’s own website. The editor’s apology, below, does not dispute that the journal sent it.
Harvey wrote to the journal’s co-editors in chief, Savvas Hatzikiriakos and Gerald Fuller, that a researcher should be able to flag a mathematical error without that happening. To Retraction Watch, he called the episode “basically editorial malpractice.” Hatzikiriakos apologized and called the disclosure an “oversight.” He also pointed out that Harvey’s email had not said it was confidential, and he directed Harvey to submit a formal comment, which he described as “the only formal process available” for correcting flawed published work.
The confidentiality point does not survive a reading of AIP’s own text. The guideline does not limit protection to people who ask for it. It tells editors to “protect the anonymity of individuals raising concerns,” with no condition attached, so the burden falls on the editor. A postdoc should not have to know to ask.
Eight referees and no retraction
Harvey filed the comment. According to Retraction Watch, 8 referees reviewed it, and none of them defended the original paper. Several recommended retraction. Some said the authors’ response “did not resolve the issues.” That response keeps the paper’s conclusions intact, with what the authors call a “single exception”: the corrected equation.
AIP’s guidelines say that an editor presented with convincing evidence should promote a correction or retraction “in a timely fashion.” Hatzikiriakos took a different route. He said that publishing Harvey’s comment together with the authors’ response would be “of greater benefit to the scientific community,” and he attributed that course to the referees’ suggestions. Retraction Watch reports that no such suggestion appears in the referee comments it saw. As of October 13, Harvey was still waiting for the journal to publish his comment alongside the authors’ response, and the paper whose authors admitted the error remains in the record unretracted.
The handling looks different once you know how the paper got into the journal. George told Retraction Watch that Physics of Fluids has been his “go-to journal for controversial ideas” since his first paper there in 1973. He said arXiv told him and Johansson they were “not qualified to submit a paper on cosmology or astrophysics,” so he pitched the paper directly to the Physics of Fluids editor in chief and supplied a list of potential referees. (Retraction Watch disclosed that its cofounder Ivan Oransky chairs arXiv’s board but is not involved in its day-to-day editorial decisions.)
None of this shows that the referee list decided the paper’s acceptance, or that a five-decade relationship shaped how the editors treated the postdoc who challenged it. Taken together, the documented facts are these: a cosmology paper by two fluid dynamicists entered a fluids journal through a direct pitch to its editor in chief, with referees its authors had suggested. When a critic showed that its central equation was wrong, his identity went to the authors in a single exchange, while the referees’ calls for retraction became a comment and a reply that have yet to appear.
Harvey said he “thought I was doing them a favor.” From now on, he said, he will state his expectations of anonymity up front. That is the lesson the journal taught the one researcher it exposed: protection is something you have to ask for, even though the publisher’s guidelines say it is owed without asking.
After Retraction Watch contacted AIP Publishing, the publisher’s research integrity and ethics manager emailed Harvey to promise a “thorough review.” Hannah Daniel, AIP’s media relations specialist, said: “We are unable to comment further while the review is underway.” AIP has not said what the review covers or when it will end. Hatzikiriakos and Fuller still hold Harvey’s refereed comment, unpublished, along with 8 referee reports, several of which recommend a retraction that has not been made. The review that might explain why sits with AIP’s research integrity and ethics manager, and no finding has been issued.