Zero Minus Six

Zero Minus Six

1

National Institutes of Health Director Jay Bhattacharya surprised many (including The Brief) by announcing that NIH will not only keep the open access (OA) policy developed during the Biden administration in response to the Nelson Memo, it will actually advance its implementation by six months. We had our doubts that a pricey, Biden-era regulation framed (in the Nelson Memo) as an “equity” initiative  (the title of the memo was “Ensuring Free, Immediate, and Equitable Access to Federally Funded Research”) would survive the Trump administration’s allergy to all things “DEI.” However, searching for policy consistency these days seems something of a fool’s errand. 
 
One significant change is the framing around the policy. Bhattacharya seems to be repositioning it to focus on public trust in science (to be fair, framing of the policy by the previous NIH administration included “trust in research” as a rationale for the policy, alongside “equity”). The immediate public availability of all taxpayer-funded scientific data seems in direct contradiction to the NIH’s actions taken earlier this month when it blocked access to many of its health databases to users in China, Russia, and other countries. The open science requirements of the Nelson Memo mean that everyone, even researchers in (gasp) Canada will have largely unfettered access to results from every project the NIH funds.
 
The costs of this policy remain troubling, particularly in light of the proposed 44% cut in the NIH’s budget. Much of positioning of the Nelson Memo has relied on a head-in-the-sand approach to its costs – less so the public access to published papers (146,000 NIH-funded papers in 2024 at an OSTP-estimated $4,000 per paper works out to $584 million per year); more so for the costs of curating and archiving vast amounts of research data in perpetuity (which likely will dwarf that figure).
 
The announcement is breaking news as we write, and like all things in the Trump administration, subject to “rapid unscheduled disassembly.” It’s hard to scenario plan against a constant flow of ever-changing and sometimes contradictory policies, which are then subject to legal challenge. Even so, organizations that are not already engaging in scenario planning (immediate solid threats as well as long-term “might happens”) are putting themselves at risk. Chaos is bad for business, so this is an important time to look at operational efficiency and to invest in diversification. What opportunities does your organization have to broaden its portfolio of products and services, spreading risks across different business lines?

Vaguely Threatening

2

This past month saw a series of “vaguely threatening” letters (as the Editor-in-Chief of The New England Journal of Medicine characterized them) sent from the Interim US Attorney for the District of Columbia to prominent scholarly journals. The letters were first reported by MedPage Today with a follow-up story on the same day appearing in Science. The letters (here is a copy of the one sent to CHEST) ask questions about potential bias in the journals:

• How do you assess your responsibilities to protect the public from misinformation?
• How do you clearly articulate to the public when you have certain viewpoints that are influenced by your ongoing relations with supporters, funders, advertisers, and others?
• Do you accept articles or essays from competing viewpoints?
• How do you assess the role played by government officials and funding organizations like the National Institutes of Health in the development of submitted articles?
•  How do you handle allegations that authors of works in your journals may have misled their readers?

While the questions themselves seem innocuous (indeed, publishers have given a great deal of thought to these and related questions over the years, developing mature policies and guidelines on advertising, editorial independence, and research integrity), the context in which they are sent has caused alarm. 
 
Amanda Shanor, described in The New York Times article as a First Amendment expert, notes that “It appears aimed at creating a type of fear and chill that will have effects on people’s expression – that’s a constitutional concern.”
 
MedPage Today quoted Adam Gaffney, MD, MPH, a pulmonary and critical care physician at Cambridge Health Alliance in Massachusetts, as saying, “It is yet another example of the Trump administration’s effort to control academic inquiry and stifle scientific discourse – an administration, it warrants mentioning, that has embraced medical misinformation and pseudoscience to reckless effect.”
 
While only four such letters have been reported thus far, we have heard through unofficial channels that the number is significantly higher. It is unclear whether this is part of a coordinated strategy or just a values-signaling exercise by a political appointee facing a difficult confirmation process. Even if it amounts to no more than political theater, it is yet another expense and time sink for editors and publishers, and yet another ominous sign of the continuing assault on the independence of the scientific enterprise. 

China (Further) Rising

3

While the US appears to be reducing its commitment to advancing research, China is further solidifying its position as the world leader in many fields. Just last month, it was shown that China has overtaken the US in cancer research output for the first time in the Nature Index, with a 19% jump in share between 2023 and 2024 (as compared with a 5% increase for the US). Threats against Harvard University, the leading US cancer research institute, and reports of proposed massive cuts to research investment make it likely that this gap will continue to widen.
 
As China is likely to become an even more important market for journal publishers than the behemoth it already is, it is essential to keep a close eye on what is happening in the country, as the government enters Phase II of the Excellence Action Plan (EAP) for China’s Scientific, Technical, and Medical Journals. The goal of the EAP is, according to the STM Association, “to transform individual journal developments into an integrated, ecosystem-based approach, aiming to position China’s STM journals among the world’s top tier by 2035.”
 
A new report offers details on the EAP, a program that provides support for 200 existing domestic English-language journals and 200 Chinese-language journals, as well as the launch of 50 new journals each year in “nominated strategically important and emerging subject areas.” Additionally, the plan calls for the building of publishing infrastructure capacity and platforms within China, as well as training programs for Chinese editing and publishing professionals.
 
At the very least, this means increased competition for every publisher in terms of recruitment of Chinese authors, particularly those active in the prioritized fields listed in the report. Journals not based in China may be at a disadvantage in the recruitment of many papers, due to the continued refinement of the Journal Ranking Table issued by the National Science Library under the Chinese Academy of Sciences, a “critical component of China’s research evaluation system.” The methods used to rank journals remain opaque, but reportedly “adjustments were made to support domestic journals, ‘which generally start at a lower level and need help to compete internationally’.”
 
Over time, this thumb-on-the-scale ranking system may reduce the value to Chinese authors of publishing in international journals. Given the increasing nationalism of the US banning Chinese access to research databases, it would not be surprising to see a response from China, further threatening the idea of science as a global pursuit.

Briefly Noted

4

Science reports that “the Trump administration plans to cut funding for two open-access, peer-reviewed journals published by the Centers for Disease Control and Prevention (CDC)…” Funding for the two journals, Emerging Infectious Diseases and Preventing Chronic Disease, has been eliminated in a draft 2026 budget proposal for the Department of Health and Human Services. A third journal, Environmental Health Perspectives, which has received funding from NIH for over 50 years, has stopped accepting papers due to a lack of confidence that it will have funding for ongoing operations, according to The New York Times. 
 
Library Journal’s annual Periodicals Price Survey predicts a 5.5%–6.5% increase in library materials prices for 2026. We suspect these data were gathered before the current economic upheaval wrought by the US federal administration, and that threatened (if not already curtailed) library budgets will result in drastically smaller price increases.
 
Open Access Scholarly Publishing Association (OASPA) is launching the “Next 50%” project, an attempt to change the conversation around OA and put some thought into what is working, as well as where OA has failed some communities. As detailed in Katina magazine, half of scholarly literature is now OA, but this in many ways represents the low-hanging fruit. New approaches and, as OASPA puts it, a “different conversation” are needed for goals seen as “more complicated, more challenging, and more precarious than in the last decade.”
 
The American Geophysical Union (AGU) has joined with a coalition of plaintiffs challenging President Trump’s executive order calling for large-scale reductions in federal agencies. “AGU’s role in the case will involve illustrating the extensive ways in which scientists and the public will be irreparably harmed by the execution of the President’s order, in particular through proposed mass terminations at NOAA [National Oceanic and Atmospheric Administration], the Department of Agriculture, Department of Energy, the Environmental Protection Agency, and the National Science Foundation.”
 
A new Nature analysis revealing the most-cited papers of the 21^st century confirms the long-prescribed advice for driving citations for a journal, namely, publish more methods papers and papers that contain large statistical datasets.
 
While most publishers have agreed that transparency in reporting use of artificial intelligence (AI) in creating research papers is essential, authors remain unsure of what should be disclosed. The STM Association’s Task & Finish Group on AI Labelling Terminology looks to offer clarity through a draft set of recommendations for Classification of AI Use in Academic Manuscript Preparation. Comments on the draft are open through May 31, 2025.
 
In reading Frontiers’ 2024 Annual Report, one number stood out to us. According to the report, 45% of all rejections were made by the organization’s Research Integrity Team. 
 
Journal of Intelligent & Fuzzy Systems (JIFS) claimed the dubious honor of having the most retractions from a single journal, reaching a total of 1,561 with the latest batch of 678 papers retracted. JIFS entered the Sage portfolio in November 2023 as part of its acquisition of IOS Press.
 
While it looked like the US Department of Education would be shutting down its Education Resources Information Center (ERIC), an online library of 2.1 million education documents, a last-minute reprieve appears to have saved the resource.
 
Sethuraman “Panch” Panchanathan, the National Science Foundation Director appointed by the first Trump administration, has resigned his position, evidently in light of a proposed 55% cut to the agency’s budget.
 
After nearly 25 years at Oxford University Press, Niko Pfund has been named Director of Yale University Press.
 
Wiley has issued a position statement on AI content scraping, declaring that “AI developers and companies must obtain authorisation before using Wiley content for AI development, training, or implementation,” and that, “Transparent attribution and data provenance are essential components of ethical AI development.” Meanwhile, Ziff Davis, owner of more than 45 media brands, is the latest publisher to file suit against OpenAI for copyright infringement. We found two articles useful this month that delve into the way AI, or at least large language models (LLMs), “think,” and why they require such enormous quantities of training data. In the first, Christopher Mims, writing in The Wall Street Journal, explains that AI largely works by memorizing endless lists of rules, or “bags of heuristics” (or perhaps better, “today’s AIs are overly complicated, patched-together Rube Goldberg machines full of ad-hoc solutions for answering our prompts”). The second, by University of Georgia professor (and singer for Camper Van Beethoven and Cracker) David Lowery, discusses the “devaluing of the world’s creative legacy” generated by these processes.
 
Then again, Google’s Gemini AI seems mightily creative, at least as far as inventing meanings for made-up homilies such as “you can’t lick a badger twice.” As explained, generative AI is “ultimately a probability machine … at a base level it’s simply placing one most-likely word after another, laying the track as the train chugs forward,” and, as computer scientist Ziang Xiao notes, “in many cases, the next coherent word does not lead us to the right answer.”

***
“I couldn’t have worked with these asshats anyway.” – An unnamed “high level HHS official shown the door” as a result of cuts at US federal health agencies and quoted in Science, a journal which, to our knowledge, had not previously published the term “asshats”