Uncertainty

Uncertainty

1

With the administrative turnover following the US election, the fate of the Biden Office of Science and Technology Policy (OSTP)’s Memorandum on “Ensuring Free, Immediate, and Equitable Access to Federally Funded Research” (aka, the “Nelson Memo”) is now even more uncertain. No one really knows what the new administration plans to do on many fronts, including science policy, so for now, it’s impossible to say where things will land. 

When a new President from the opposing party takes office, it is not uncommon to see many policies issued by the previous administration rescinded. But the Nelson Memo (outside the relatively tiny niche of scholarly communication) is unlikely to be a priority for the new administration. During the last Trump reign, his OSTP advanced policies very similar to those of the Nelson Memo but abandoned their efforts short of implementation. However, the use of the word “equitable” in the Memo’s title may set off the administration’s extremely sensitive “anti-woke” alarms, raising its profile.

Congress is another wild card. In the last two funding bills, the House Committee on Science, Space, and Technology has shown some clear skepticism toward the Nelson Memo and the OSTP’s complete lack of any cost/benefit analysis. This has resulted in language in these last two bills stripping all payment for Memo implementation. So far, that draft language hasn’t survived to the final passed act, largely due to compromises with the Democratically controlled Senate. With both houses of Congress moving to Republican control, these spending restrictions may stand a better chance of moving forward. It is also worth noting that language blocking the use of the Federal Purpose license for Nelson Memo implementation was included in the latest draft bills by both the House and the Senate and thus may make it through.

The other factor to consider is an increased focus on research security. Both the Biden and the incoming Trump administrations are emphasizing new security measures for research institutions receiving federal funds. Many of these measures are meant to block the transfer of research data to “countries of concern,” including China. This seems in direct opposition to the concept of open science and the open data requirements of the Nelson Memo.

Even if the Trump administration rescinds the specific Memorandum, that still may not prove an end to agency implementation of the policies it contains. NASA essentially already has the Nelson Memo in place. The agencies that believe strongly in the open access and open science agendas may similarly continue to move forward with their implementation plans at their own discretion (just without reference to the rescinded Memo). The agencies largely set their own policies after all. Then again, given what we’re seeing coming regarding plans for major reorganizations of existing agencies, many may simply have other priorities or lower funding levels.

But, at the end of the day, this is idle speculation. John Mulaney’s story of a “Horse in a hospital” remains the best metaphor we’ve seen for what to expect from both the first and now the second Trump administration. “No one knows what the horse is going to do next, least of all the horse, he’s as confused as you are.”

Evidence-based

2

In September of this year, Clarke & Esposito (C&E) supported the MIT Press with a workshop that brought together publishers, librarians, researchers, funders, and policymakers. The workshop, “Access to Science & Scholarship 2024: Building an Evidence Base to Support the Future of Open Research Policy,” convened stakeholders to define the important gaps in our knowledge around what it would take to create and implement effective open science policy. C&E helped build the agenda, recruit speakers, and produce the output report from the meeting, which was released in November. 

The project, spearheaded by Amy Brand (Director and Publisher at MIT Press), is aimed at bringing the voice of the researcher (both at MIT and beyond) into the research communication policy conversation. They are the ones most directly affected by policies such as the Nelson Memo, and yet, largely, the working researcher has been excluded from the policy development process. The goal of the workshop was to help build out an open science research agenda – starting with the assumption that open science is a positive development for research progress, what do we need to know in order to make effective policy to drive it forward?

Video was recorded for each session and is available through links in the report. Particular highlights from our point of view included a panel of society and university press publishers explaining to a room of non-publishers that no, digital publishing is not less expensive than print publishing, in fact it is significantly more expensive. And grows more expensive every day, as author and reader expectations increase and new tools such as those for research integrity checking seem to constantly arrive. 

John Wilbanks, in the session on research data access, made the point that open data policy (and, really, policy in general) is filled with “magical thinking.” Policy is often outcomes-based, and written at a very high level – e.g., make research data and research papers publicly available – with the messy part of making that happen left to others – the individual agencies, research institutions, publishers, and researchers themselves, each with their own agendas and limited viewpoints. Because there is no coordination between these actors, most open science policies end up a jumbled mess of unintended consequences.

As Rachel Burley noted (in the Business Models panel), the reason why large, commercial publishers are so often successful is that they do extensive scenario modeling before enacting business strategy. The research questions raised by the workshop and outlined in the report should enable wider scenario modeling and raise awareness among funders and policymakers of what it will take to effectively realize their goals.

We hope the report will mark a starting point for an ongoing dialogue that will build better future policies. We know that Plan S made progress, but fell short of its goals and suffered from unintended consequences. We may (or may not!) see similar struggles as the Nelson Memo is implemented (or not implemented!) next year. But the more these questions can be answered, the better the chance is for ongoing policy success and the progress those policies seek to achieve.

JIF-less

3

As widely expected, after a period of review, Clarivate has decided to de-list eLife from the Web of Science (WoS), and the publication will no longer receive a Journal Impact Factor (JIF). Essentially, eLife’s new publishing model, in which all papers sent out for peer review are “published” regardless of the reviews received, violates Clarivate’s rules around separating publication from validation. Clarivate has offered to continue indexing eLife in the WoS, or at least those articles that aren’t deemed by reviewers to be “incomplete” or “inadequate,” if eLife can provide a feed that removes those articles. While it is unclear whether eLife will offer that feed (remaining in WoS would bolster discovery for the articles), doing so would not restore the publication’s JIF. Journals that are only partially indexed are not eligible to receive a JIF.

This decision by Clarivate makes absolute sense. The WoS and the JIF, love them or hate them, are a specific and defined index and metric for research journals. When eLife acted as a journal, it fit the necessary criteria for eligibility. Now that eLife is no longer a journal and styles itself a “service that reviews preprints,” it no longer fits in that index or that metric. Keeping eLife in the WoS and awarding a JIF would mean changing the very nature of what the WoS and the JIF are. Rule changes that would allow in the new eLife could also permit inclusion of F1000 (the multidisciplinary open research platform), preprint servers, and really almost anything on the internet where material is posted and commented upon, regardless of the quality of the material or the outcomes of those reviews. We already have that service, it’s called “Google Scholar,” and it serves a very different purpose from WoS (and makes a lot less money for its owner).

What losing its JIF will mean for eLife remains to be seen. A journal being delisted from the WoS and losing its JIF is usually followed by a massive drop in submissions and a resulting drop in publication volume. eLiferemains optimistic that this won’t happen to them, as submissions held relatively steady (with the significant exception of submissions from China, which fell off considerably) during the weeks after eLife was put “on hold” by Clarivate. The analysis offered by eLife is likely premature – there’s a big difference between being investigated and being delisted, and it will take some time for word to get around to potential authors. Since eLife was granted a JIF in 2024, the real change will take place when new JIFs are announced in 2025 and eLife is absent from the listings.

While both eLife and David Worlock suggest that the eLife delisting will serve as a tipping point, ending the dominance of the JIF as an important journal metric for academic hiring, tenure, and funding decision-makers, we at The Brief are less sanguine. eLife is but one journal among many, and even many of its prestigious (and now former) editors disagree with the direction it has taken. 

While it is disappointing to see one of the most interesting experiments in the market maybe coming to a premature end, perhaps the silver lining in the decision is that it now frees eLife entirely from the trappings of the traditional research journal. Despite its recent moves, eLife was still mimicking the form and remaining, as we put it back in 2022, “journalesque.” Now eLife is free to cut those bonds entirely and to become something radically new. Given its ongoing and seemingly limitless funding, delisting from the JIF may mean the end of eLife the journal, but not the end of eLife. 

Tangled up in blue

4

While use of X (the rebranded and reoriented Twitter) by both the research community and those in the scholarly communications industry has been in steady decline, Elon Musk’s role in the recent US elections combined with recent changes in terms of services (which include allowing X to use your posts to feed its AI) seem to have triggered a blue (pun intended) wave of defections. Story after story this past month detail a mass exodus from X to Bluesky, particularly for the medical and science communities. The scholarly communications conversation shows signs of moving there as well.

Is this real? And if it is, does it matter?

First, the positives: Bluesky is a welcome change from many of the more established social media platforms. It has a similar interface to X but feels like the “old-Twitter” to many users – with some improvements. The culture of immediately blocking trolls without offering any form of engagement helps keep the discourse civil, and there’s no need to wade through a horde of blue-check nonsense to get to the interesting replies. Bluesky is based on an open and decentralized social media protocol (the “AT Protocol”). This enables a more stable environment for third-party interfaces, like TweetDeck (now X Pro) used to offer. It also enables you to use your own domain as a handle if you’d like. Refreshingly, Bluesky’s default view or feed is not algorithm-influenced. It is just a chronological list of posts from the accounts you follow. If you get bored with the usual suspects and want to wade into less familiar waters, there are plenty of options to do that. 

And at least so far, Bluesky does not and “has no intention to” use your content to train generative AI.

All that said, Bluesky is still tiny. Traffic is surging post-election, but the latest numbers we’ve seen show the site just passing 22 million users (and gaining around 1 million users per day). That pales in comparison to the 586 million active users still on X, which is itself small compared with the more than 2 billion active on Instagram and Facebook. Many social media experts are still predicting that Threads (275 million signups, of which 35 million in November alone) will be the eventual X successor.

But size isn’t everything. It is far too early to get a clear sense of whether the current Bluesky gains are meaningful, but ultimately it depends on who you’re trying to reach. It may not matter whether Bluesky gets as big as X if your key constituencies are concentrated there. If the scholarly publishing conversation goes there, then it will be a useful place to learn about what’s going on in the community. If the science conversation moves there, then it becomes a useful place for scientists to keep up with their colleagues.

Putting on our marketing hats, we urge caution, as it’s too early to call how this plays out for Bluesky as an essential social media marketing channel. We remind our readers that social media is a small marketing referral channel for journals, although it may have a much larger impact on brand perceptions and community engagement than those raw numbers show. Looking across publishers, over the last year we have observed LinkedIn growing as a traffic source to journal platforms as X declines. Bluesky has been too small to make an impact, but as metrics begin to catch up with post-election behaviors, that may change. Furthermore, Altmetric has recently added Bluesky to its “donut” of measurements, which may further drive its use for raising awareness of journal articles.

More to follow as things develop, but for now, The Brief is officially on Bluesky, and you can follow all of your favorite C&E denizens through this starter pack. At the very least, we can guarantee some pictures of our adorable dogs.

Briefly Noted

5

Google Scholar has launched a new AI-driven research paper summary tool to its PDF Reader, raising concerns among publishers that it will significantly reduce traffic to their journals. This mirrors the worries of the broader content-providing market where traffic from search engines is expected to decline as more and more AI-based discovery tools offer direct answers to search queries rather than pointing users to the sources where those answers can be found. While we at The Brief are happy to see someone offering competition to Google’s search monopoly (in the form of ChatGPT’s new ad-free search engine from OpenAI), ideally that competition will not come at the cost of harming content creators who feed the services. In more positive AI news, Google researchers suggest that invisible watermarks could be added to identify AI-generated text. We also highly recommend reading Ithaka S+R’s new report, “A Third Transformation? Generative AI and Scholarly Publishing,” to help consider future scenarios for the impact of AI on our community. Furthermore, this wide-ranging conversation about AI and trust in science between National Academy of Sciences President Marcia McNutt and JAMA Editor-in-Chief Kirsten Bibbins-Domingo is worth your time. 

Invest in Open Infrastructure offered a report this month, “The Cost and Price of Public Access to Scholarly Publications: A Synthesis.” While the report rounds up many previous analyses and contains some useful data, it does not offer the actual cost or price mentioned in the title; rather, in summary, it concludes that “‘real-world’information on actual costs and payments is lacking and is needed to better support the breadth of stakeholders navigating this landscape.” As with the MIT Press report above, the gaps in our knowledge continue to confound our ability to properly scope and fund policy.

Taylor & Francis has introduced a new model for open access (OA) agreements, and a new abbreviation to remember, CPOP (Collective Pathway to Open Publishing). The scheme aims to help humanities and social sciences journals, where APC (article processing charge) funding is scarce, transition to OA. The model is a blending of transformative agreements and subscribe to open, where the “read” portion of agreements for journals, if it hits a certain threshold, is put toward funding OA for authors at institutions without agreements in place.

Can you improve your journal’s JIF performance by frontloading the January issue with a lot of articles so they have a longer period to collect citations? A new study by Duncan MacRae and Maya Workowski suggests that this conventional wisdom may not be all that accurate. They found that from 2017 to 2021, the average citation performance for articles published in January in Wolters Kluwer journals was only the fifth highest among months of the year. While this is a fairly seat-of-the-pants analysis and there are a lot of confounding factors likely involved (e.g., they looked at issue publication date, not the online publication date of the articles, some of which were already being read for many months before appearing in an issue; some journals included don’t publish a January issue; performance may vary significantly between fields and Wolters Kluwer’s corpus is concentrated in biomedical areas, etc.), the take-home message is sensible: if you want more citations, publish really high-quality, significant articles. A few extra months of collection is probably not going to help uninteresting articles all that much.

We at The Brief are big fans of page and word length limitations for research papers. Such limits instill discipline in authors and reduce the already enormous burden faced by peer reviewers and editors. That said, a recent 35-page, 61-figure article published in Naturehas raised some interesting debate about the value of the occasional huge, large-dataset, “atlas” style paper.

Cambridge University Press & Assessment (CUPA) issued their 2023–2024 Annual Report, showing both increased revenue and increased profits, with profits rising from last year’s £140 million to £203 million. Much of the growth appears to come from CUPA’s educational offerings, as “Academic revenues remained stable while adapting to the demands of open access and online education and serving ever-more scholars.” CUPA also undertook the enormous task of reaching out to all of its authors to gain their opt-in for the use of their work in AI-training deals. While the effort is likely appreciated by authors, the scale of the work here is staggering, with contacts needing to be tracked down for more than 20,000 authors (reminding us of the early days of eBooks when publishers had to figure out how to get in touch with everyone who had ever authored a book for them in order to add eBook rights to their contracts). So far CUPA reports a positive response, “with only a small minority declining to license their content.”

Reports of the death of the bricks-and-mortar books superstore were perhaps exaggerated, as Barnes & Noble seems to be thriving, opening some 60 bookstores this year and reporting that the company is “experiencing strong sales in its existing stores.”

The Researcher App, a discovery and discussion tool for academic research, has been shut down.

Research Consulting’s recent study on the future of learned society publishing was interesting, but the geographical coverage was limited (only UK societies were examined) and some of the conclusions less than compelling. A new global survey on society publishing will at least correct the former issues.

As mentioned in the September issue of The Brief, the STM Association has been focusing on research identity as a route to improving research integrity. A new draft report released for community review means to “explain the background of this subject, and set the scene for forthcoming work that explores possible solutions.”

Problems with guest-edited special issues are not just limited to born-OA journals such as those from Hindawi. Springer Nature is in the process of cleaning up hundreds of problematic papers from several journals.

Speaking of research fraud, The Atlantic has a long piece on the crisis in business school psychiatry research that continues to grow.

Laura Helmuth has resigned her post as Editor-in-Chief at Scientific American following controversy over political comments made on social media.

Matt Hudson is the new Director of Publishing and Publishing Services at GeoScienceWorld. 

Wendy Queen has been appointed the inaugural Chief Transformation Officer at Johns Hopkins University Press. 

Book industry legend Mike Shatzkin has passed away. Our condolences to his family and many friends and colleagues.

Is watching the new listening? The Wall Street Journal says that YouTube is now a dominant player in the world of podcasts, which seems difficult to fathom but here we are.

In a blow to the hopes of those who love literature, monkeys, and randomness, researchers in Australia have determined that the “infinite monkey theorem,” – the idea that, if given an infinite amount of time, a monkey pressing on the keys of a typewriter would eventually compose the complete works of Shakespeare – is technically true, but misleading. Apparently, by the time the current population of chimpanzees on Earth is likely to complete the task, the heat death of the universe will have occurred. We at The Brief are still hoping our simian friends can at least put together a few sonnets before the lights go out for us all.

***
…the business and operating model of book publishing as it has been throughout my 50+-year career is irretrievably broken. And it is increasingly obvious that this is the case across all “content” businesses – newspapers, magazines, movies, TV, and radio – and for very much the same reasons. 

The primary challenge is that no content ever becomes unavailable anymore. So every book in the history of publishing is competing for attention with everything “newly available.” The other delivery forms suffer from a variation of this theme. 

– Mike Shatzkin (2023)