Not So Special

Not So Special

1

It is extraordinarily difficult, but not, as it turns out, impossible, to lose US $400 million in a single day in scholarly publishing. Wiley’s stock price plummeted 17% following its March 9th earnings call (see Figure 1 below) reducing its market cap from $2.4 billion to $2 billion before the day was out. This precipitous slide in Wiley’s stock price resulted from a 5% year-over-year decline in earnings, which were lower than analysts’ estimates. The decline in revenue was largely attributed to two circumstances. The first is “demand pressure” (lower than expected sales) in its higher education business, attributed, in part, to declining enrollments. The second, and more notable, circumstance is the “pause” in publication of special issues by Wiley’s Hindawi publishing unit following the discovery of widespread publication fraud. 

The “publication fraud” Wiley is referring to consists of extensive fabrication of articles and related peer review by (or coordinated by) paper mills. Retraction Watch reported on this problem back in September 2022, when Wiley announced the retraction of more than 500 articles published in Hindawi titles:

Hindawi’s research integrity team found several signs of manipulated peer reviews for the affected papers, including reviews that contained duplicated text, a few individuals who did a lot of reviews, reviewers who turned in their reviews extremely quickly, and misuse of databases that publishers use to vet potential reviewers. 

Richard Bennett, Vice President of Researcher and Publishing Services for Hindawi, told us that the publisher suspects “coordinated peer review rings” consisting of reviewers and editors working together to advance manuscripts through to publication. Some of the manuscripts appeared to come from paper mills, he said. 

Wiley reported in the earnings call that it had suspended publication of special issues by Hindawi in October 2022, shortly after this retraction announcement. According to reporting by Retraction Watch, the special issue program restarted in January [Editor’s note: We have updated The Brief to reflect that Hindawi restarted special issues in January. Originally we reported that the special issue program was still paused in March]. Wiley projects that the pause in special issues will result in a $30 million reduction in revenues in the 2023 fiscal year. Even more problematic for Wiley is that special issues appear to be remarkably profitable, resulting in a $25 million impact on Adjusted EBITDA (earnings before interest, taxes, depreciation, and amortization). That is to say, Wiley earns an 80% margin on special issues, resulting in an outsized margin impact on overall revenue.  

Some background on special issues is warranted. Special issues are journal issues on a specific topic or theme. They are often edited by a guest editor—usually an expert in the topical area—who typically recruits papers and selects peer reviewers. Special issues have been used for decades by a great many journals, including many well-known and respected titles. 

In addition to providing a way to highlight papers on a narrow topic, special issues have become an engine of growth—one used extensively by Hindawi competitors MDPI and Frontiers (Frontiers refers to them as “Research Topics”). Back in 2016, Hindawi published approximately the same number of papers as both MDPI and Frontiers (see Figure 2 below). Since then, however, Hindawi has lost considerable market share, especially when compared with MDPI. While Hindawi has grown its publication output from just under 23,000 articles in 2016 to just over 50,000 articles in 2022, MDPI grew from around the same number of papers in 2016 to over 300,000 papers in 2022. Frontiers meanwhile grew from 20,000 articles in 2016 to nearly 130,000 in 2022. In both cases, this extraordinary growth can be attributed, in large part, to special issues. Hindawi appears to have only recently adopted the special issue growth strategy in an attempt to make up lost ground. 

We discussed MDPI’s special issue growth in the September 2021 issue of The Brief, noting at the time that MDPI had a totally bonkers 39,587 special issues with a closing date in 2021. Times Higher Education reports that this number has grown, with MDPI aiming to produce an even more bonkers 56,000 special issues in 2023. As THE notes, MDPI’s two flagship journals—the International Journal of Molecular Sciences and Sustainability—have announced, respectively, 3,514 and 3,512 special issues with a closing date in 2023. That works out to a mind-boggling 9 special issues a day—for each journal! They will likely only publish a fraction of this number as some will be abandoned or consolidated, but even if they publish only a quarter of the planned issues it is a huge number.

As noted above, a salient aspect of the special issue model as practiced by MDPI, Frontiers, and Hindawi is the use of “guest editors.” One benefit of the guest editor model is that guest editors are often closer to the topical areas than the regular journal editor and better positioned to identify contributors and peer reviewers. They also provide additional labor, which has allowed MDPI and Frontiers to scale their programs. A critical drawback to special issues, however, is that they provide a particularly attractive target for paper mills. 

Paper mills are a growing problem for scholarly publishers. The Committee on Publication Ethics (COPE) defines paper mills as “profit oriented, unofficial and potentially illegal organisations that produce and sell fraudulent manuscripts that seem to resemble genuine research.” In a recent article in the Financial Times (FT), researcher Bernhard Sabel, estimates the global market for paper mills at “a minimum of €1bn a year and probably much more,” with much of the business focused on Chinese researchers. The business model of paper mills is to charge researchers for authorship on a paper published in a peer reviewed journal. The FT, for example, notes a recent advertisement from a paper mill seeking to sell authorship on a paper to be published in a mid-tier Chinese journal for $800. Authorship in international journals can cost much more. 

Once the paper mill secures paying customers they fabricate a research paper. The trick then becomes publishing it. Special issues have provided an attractive target for paper mills, as the guest editor model is easier to manipulate. Paper mills will often impersonate the identity of a researcher to become a guest editor. They will then fabricate papers and manipulate the peer review process by impersonating reviewers. The outcome is a special issue filled with fabricated research inside of a legitimate, peer reviewed journal with a Journal Impact Factor (JIF). One published, the “authors” of the paper can then list it on their CVs for career advancement—which works so long as (1) no one on the promotion committee actually reads the paper, and (2) the paper is not retracted in a widely publicized publication scandal. 

Wiley purchased Hindawi in 2021 for nearly $300 million, a remarkable figure that makes Hindawi the second-largest acquisition in Wiley’s 214-year history. We covered this acquisition in the January 2021 issue of The Brief, noting that Wiley published far fewer articles annually across its journal portfolio than its competitors (Elsevier and Springer Nature). This is not a problem for subscription journals, as the unit of value is the journal, not the article. In open access (OA) publishing, by contrast, the unit of value is the article and a publisher’s revenue is based on its article output. Wiley has embraced Gold OA publishing and continues to sign transformative (read and publish) agreements, which seek to shift the locus of payment from subscriptions to article output. Wiley’s CEO, Brian Napack, noted in the March 9 earnings call that Wiley has signed over 60 such agreements and they are “foundational to our Research strategy.”

The purchase of Hindawi was meant to help better position Wiley for OA publishing by bolstering Wiley’s overall article output and accelerating its article growth. The ensuing mess at Hindawi points to the heightened economic risk borne by publishers shifting to Gold OA models (including transformative agreements). Under a subscription model, not only are the incentives different (there is far less incentive to attempt to rapidly grow one’s article output), but also the economic consequences of missteps are reduced. Were Hindawi publishing under a subscription model, 2023 subscription revenues would be largely unaffected by the suspension of its special issue program. With a Gold OA model, any pause or reduction in published output results in an immediate impact on revenues (and, in this case, an outsized impact on margins)—and one that continues until publication output resumes at previous levels. 

This heightened risk profile raises the importance of quality control. Hindawi is not the first publisher to run into fraud associated with special issues. We have previously commented on similar circumstances of publication fraud (albeit on a smaller scale) with regard to Springer Nature and Elsevier journals. 

In pausing Hindawi’s special issue program, Wiley did the right thing, despite the significant short-term financial implications. While the drop in Wiley share price is undoubtedly painful for both Wiley management and shareholders, the share price will likely recover (this is not investment advice!). Perhaps more concerning for the longer-term outlook for Hindawi was the announcement by Clarivate on March 20th that 19 Hindawi titles have been delisted from the Web of Science Core Collection, and hence will no longer be issued a JIF (see Item 2 below). According to Dimensions, those 19 journals accounted for 50% of Hindawi’s published article output in 2022. 

The decision by Clarivate to delist these 19 titles, despite Wiley’s efforts to rectify problems—combined with the reaction from the market—is likely to scare other publishers away from taking similar remediation measures in cases of future widespread publication fraud (of course, it will also likely scare publishers into developing strong editorial safeguards in the first place). If the likely reward for being open about publication fraud and taking corrective steps is being punished by the market and delisted, publishers have every incentive to deny problems and attempt to quietly sweep things under the rug. 

Wiley has put in place more rigorous fraud-detection measures. According to a communication by the company to The Brief, they have “increased rigorous new checks throughout our publishing workflows, added additional expert staffing to assess all manuscripts prior to publication, and introduced additional AI-based screening tools.” As the special issue program at Hindawi resumes with these measures, the question is whether Hindawi will be able to meet Wiley’s expectations for growth? Not only will Hindawi likely be more cautious with special issues, it will also have 19 journals (including some of its most prolific titles) effectively sidelined.

Delisted

2

Many academics still care deeply about Journal Impact Factors (JIFs), despite initiatives such as DORA (the Declaration on Research Assessment). As a result, publishers care about impact factors too and often complain when their latest journal is not selected to receive the metric. 

In 2022, Clarivate announced that every journal in the Web of Science Core Collection will receive an impact factor in June 2023. This means that 9,000 additional journals will be listed for the first time. This should reduce some inequities. Attracting submissions is hard if a journal doesn’t have an impact factor; getting an impact factor is challenging for new journals that haven’t published many papers. This vicious circle should now be eliminated for most scholarly publishers.

In tandem, Clarivate has been investigating more than 500 Web of Science journals to determine whether they have committed nefarious publication practices and so should be delisted. According to a recent blog post, more than 50 journals have so far failed to meet the quality criteria when they were re-evaluated and have been removed from the Web of Science Core index (we assume this means that as many as 450 more journals may be delisted in the future). Clarivate has not published an official list of the journals that have been removed. However, Retraction Watch published the names of the delisted journals a few days after the announcement (this list was subsequently confirmed by Clarivate in reporting by Science). 

As noted above (see #1), 19 of the delisted journals are published by Hindawi. Two MDPI journals were delisted because they fell foul of Clarivate’s “content relevance criterion.” Essentially this means that the two MDPI journals published papers that were not in the editorial scope of the journals. MDPI was quick to point out that their journals were not removed from Web of Science because they published low-quality papers. One of the MDPI journals, International Journal of Environmental Research and Public Health, was the second largest journal in the world in 2022; only Springer Nature’s Scientific Reports published more papers. 

The implications for those journals are likely to be significant, as a recent Viewpoint in JAMA makes clear: “One of the earliest specialized mega-journals, Oncotarget, grew rapidly to the point of publishing 10,336 items in a single year (2017) at its peak; then Clarivate delisted it from Journal Citation Reports under contested circumstances. Without an impact factor, its output decreased to a mere 159 articles published in 2022.”

The lack of transparency around which journals were removed from Web of Science and why is concerning. Publishers are at the mercy of decisions made by the Web of Science editorial team, yet there is no official list of which journals were delisted nor much in the way of an explanation as to why they were removed. At the moment, there is an unofficial list that contains the names of journals that were removed because they published fraudulent papers alongside journals that have been delisted simply because of a name change (and have been relisted under the new journal name).  

Clarivate is listed on the New York Stock Exchange and a recent earnings callwith investors provided some insight into why changes are being made to Web of Science. Jonathan Gear, the CEO, said that Web of Science and Derwent (a patent database) “are dragging down our growth rate.” He also noted that the Web of Science user interface was “a little dusty” and that allowing that to happen was “very much an own goal.”

In the Q&A session, Clarivate’s CFO, Jonathan Collins, said that the expansion of the JIF plus the addition of a Preprint Citation Index to the Web of Science, which was announced in February, “will continue to firm up [subscription] renewal rates, helping us to be able to capture some of that value in the form of pricing.”

The biggest risk to Clarivate’s cash cow is more likely to be CoARA (Coalition on Advancing Research Assessment) than DORA. At the start of March, CoARA announced that 423 members had joined since December last year. The Agreement on Reforming Research Assessment, published in July 2022, includes a commitment to move away from using JIFs as a proxy for academic performance. If the CoARA academics get their way, Web of Science may gradually drag down Clarivate’s growth rate even further. However, long-running attempts to wean academia off JIFs have largely proven fruitless, so it’s unclear whether CoARA will be able to pull this off.

A New Flagship at Frontiers

3

It is a truth universally acknowledged that a publisher in possession of a good fortune must be in want of a flagship. It is therefore no surprise that a new, selective, multidisciplinary journal, Frontiers in Science, was launched by Frontiers this month. What is unusual, however, is that this was a soft launch—there was no press release, no fanfare. Even the boilerplate on the press office page claims the journal is “launching in 2022.” This new selective journal does not appear to be positioned as a revenue driver. Rather, it appears to be part of a brand-building initiative.

The Frontiers in Science publishing model is somewhat unusual. The journal publishes “lead articles” alongside opinion articles that discuss the lead article and what it might mean for science and society; these form Article Hubs. The lead articles are all invited and some authors will be part of the “Frontiers Forum Deep Dive,” which is similar in format to a TED conference and has been held every year since 2015. Frontiers in Science will publish 20–25 lead articles a year, so this is not a commercial endeavor. Rather, the new journal is about improving the perception of the Frontiers brand as a quality publisher. 

The research community is slowly moving away from associating OA with low quality, but the perception is still there, fueled by the rapid, exponential growth of “born OA” publishers and the proliferation of special issues (see #1). In recent years, Frontiers’ marketing material has focused on impact and citations—“3rd most-cited publisher” appears at the top of the Frontiers’ homepage alongside “6th largest publisher.” It’s also noteworthy that Frontiers’ tagline is “Where scientists empower society. Creating solutions for healthy lives on a healthy planet.” As we described in a previous issue of The Brief, publishers are increasingly focusing their branding around solving global society’s Grand Challenges. Frontiers in Science further supports this approach.  

The growth of Frontiers has been extraordinary. According to Digital Science’s bibliometrics tool Dimensions, Frontiers published 129,132 articles in 2022, up from 87,399 in 2021 and 49,624 in 2020 (see also #1 for the role of Research Topics in this growth). The portfolio has 196 journals, some of which are very broad (for example, Frontiers in Medicine), and some of which are rather niche (Frontiers in Arachnid Science is unlikely to catch many papers, although it presumably has a good chance of being included in Web of Science). The largest journal in the portfolio is Frontiers in Psychology, which published an astonishing 8,524 articles last year. Thirty-three journals in the portfolio published more than 1,000 papers in 2022; between them those journals published 87% of the portfolio’s articles.

As publishers think through how best to transition to OA, they would do well to remember that a flagship journal can be many things. Frontiers in Science does not appear to be designed to be part of a transfer cascade. Rather, the journal is about enhancing the reputation of the publisher. Selective journals are expensive to produce, but those costs may be tolerable if the return on investment appears in other intangible ways.

AI and Content Licensing

4

Authors and artists are increasingly concerned as they see AI text and image generators ingesting their content and regurgitating strikingly similar works to their original creations. The Authors Guild has gone so far as to recommend authors add a clause to their contracts preventing use of the work in AI training. For scholarly journals and their publishers, the question is less about seeing original work repurposed (we all “stand on the shoulders of giants,” after all), but instead whether, in an OA world, this sort of reuse could provide a valuable revenue stream that could help lighten the financial load on authors and their institutions.  

The revenue that different journals earn from licensing, reprints, and secondary rights varies tremendously, but the more money that comes in from non-APC (article processing charge) sources, the lower those APCs can be. 

Licensing scholarly articles for reuse in AI training represents a new revenue stream for many publishers. The rights and legal requirements around this sort of reuse remain unclear, but as the Copyright Clearance Center’s Roy Kaufman explains in a recent Scholarly Kitchen post, a handful of ongoing court cases will likely help set the rules going forward. The cases essentially involve two questions: (1) is reusing images and text to train an AI to be considered “fair use” (and thus does not require paying copyright holders for permissions); and (2) if the materials are licensed under open source or Creative Commons attribution terms, is the AI required to cite all input sources for its output?

If the courts decide that AI use is transformative, then all of this is moot, but it is worth considering the alternative, and adopting strategies that allow for the retention of licensing revenues. If the courts decide that rights licensing is required, then avoiding licensing terms that voluntarily give away those rights (i.e. the CC BY license) wherever possible may be the most prudent course to take. Many journals have already adopted this practice, licensing OA articles under more restrictive terms (i.e. CC BY-NC-ND) except where funders require CC BY. 

The attribution question presents a different can of worms, and another unintended consequence of OA funder policies. If copyright remains with the author (a largely valueless proposition for the author when a CC BY license is required, as we wrote last month), then when AI companies wish to go beyond the terms of that CC BY license (e.g., using an author’s work in a training process but not providing attribution), they’ll need to seek permission from the copyright holders (authors). That would require herculean efforts and AI companies will never reach all authors, thereby leaving out potentially valuable information from their systems. Depending on which way the courts rule, it may make more sense to revert to copyright transfers from authors to publishers—with publishers then issuing licenses and thereby (re)consolidating permissions. What is old may be new again.

The Dog that Caught the Car

5

Now that a move to OA in journal publishing appears to be accelerating, libraries and funders who have pushed for it for years suddenly seem to be realizing just what it means, and sounding a bit like the dog that caught the car.

After leading the push to author-pays, APC-based OA, the EU appears to be having second thoughts. A leaked draft policy demands that OA be “the default mode in publishing, with no fees for authors.” The policy offers no details on exactly how this is supposed to work, but it fits well with the long-running notion that everybody wants OA but nobody wants to be the one stuck paying for it. 

Meanwhile in the US, the Ivy Plus Confederation (libraries from Ivy League universities plus a few other research-heavy institutions such as MIT, University of Chicago, Duke, and Johns Hopkins) have sent a somewhat late-in-the-game letter to the OSTP (US Office of Science and Technology Policy), offering their concerns about having to cover the costs of the new US policy. 

It has long been known that the APC route to OA concentrates costs on productive research (“publish”) institutions while reducing costs for institutions that read journals but don’t author a lot of articles (“read” institutions). Back in 2009, Phil Davis described it as “a publication model that costs much more than the system we have in place today.” Blogger Pablo K saw this happening back in 2012 when the Research Councils UK policy was first being put in place. All of this likely seemed an abstract threat in the US, and something worth putting on the back burner in the drive to OA, at least until the Nelson Memo turned it into an urgent concern.

A look at the numbers makes the scale of this problem more evident. According to Dimensions, MIT published some 8,337 articles in 2022. At an average APC of $3,000 (perhaps a bit higher than the average APC, but MIT authors are likely to be publishing in higher-ranked and more expensive journals), that works out to $25 million in payments per year. Harvard published 32,291 articles in 2022, which would be close to $97 million in annual author fee payments. Remember, Harvard has been a leader in the push to OA, and back in 2012 stated that they could not afford to pay $3.5 million in annual journal subscriptions.

While it’s hard to muster a great deal of sympathy for Harvard, with its $53 billion endowment, $5.8 billion in annual operating revenue, and operating surplus of over $400 million in 2022, the arguments made about the inequities inherent in the APC model shutting out non-Harvard researchers are worth considering. Unfortunately, the letter offers little in terms of solutions, and fails to make any policy “asks” of the OSTP. Given the enormous amounts of money that funders already provide to universities in terms of grant overheads, which are supposed to pay for support and facilities needed by researchers to do their work, a response from federal agencies along the lines of “we’ve already given you this money” would not be unreasonable.

As there is no additional money available to funders to support these policies, potential solutions for research-intensive institutions could be either increasing the size of grants to include additional funds to cover publication costs (meaning fewer, but larger grants) or restricting some percentage of grant overheads for this purpose (which would reduce funds for other aspects of research). 

A further problem for libraries is that, while including funds for OA fees in research grant overheads seems like an apt solution (putting aside the question of where such funds will come from), it is not at all obvious that they should be the ones to manage such funds, as opposed to the university’s Office of Research or purchasing department. Some libraries may therefore find themselves in the position of “saving” money that previously went to subscriptions, only to see their budgets reduced accordingly. 

For now, file this one under “be careful what you wish for, you just might get it.”

Attempted Coup at eLife

6

Earlier this month, eLife Editor-in-Chief Mike Eisen signaled via Twitter that he might be leaving his post, writing that he has been subject to “some pretty nasty and direct personal attacks… from colleagues.” A few days later, Nature published a news story adding more details to the drama:

On 9 March, 29 eLife editors—including the journal’s former editor-in-chief, Randy Schekman—wrote to Damian Pattinson, executive director of the journal’s non-profit publisher, eLife Sciences Publications in Cambridge, UK, asking that Eisen be replaced “immediately”. They added that they had no confidence in Eisen’s leadership, because he had dismissed their concerns and had not considered compromise positions. One of the journal’s five deputy editors had already stepped down from that leadership position, and “significant numbers” of reviewers and senior editors were “standing ready to resign”, they wrote.

The kerfuffle is related to the new direction at eLife, which we wrote about in October. Eisen plans on shifting eLife away from a traditional journal to what is effectively a “service that reviews preprints.” While clearly a radical departure from standard journal publishing, this is a direction that aligns with the views of at least one of eLife’s three funders (both of the authors of this paper work at Howard Hughes Medical Institute [HHMI]). 

It is understandable that eLife’s founding editor and other editors that built the journal and its reputation from scratch would be concerned about seeing their work effectively discarded by the new editor. However, Mike Eisen is not the person that one hires to be a traditional journal editor. When it comes to journal publishing, Eisen is not a “status quo” guy. He was also well known to eLife’sfunders—HHMI also funds Eisen’s lab. We joked in our October piece that he probably didn’t wear his “Journals Aren’t Real” t-shirt to his eLife job interview—but maybe he did! There is no chance that anyone on the hiring committee said something in the vein of “We’d like to maintain the status quo at eLife, maybe with an improved Journal Impact Factor.” Eisen is not the person you hire for that job. He was almost certainly hired to do exactly what he is doing, a point ratified by the eLife board in a statement made in response to the Nature story.

People

7

Springer Nature has reorganized its leadership, with Steven Inchcoombe moving from Chief Publishing Officer to a newly created position, President, Research. Harsh Jegadeesan, formerly the company’s Chief Solutions Officer, moves into the CPO role. 

Chris Marcum, who oversaw much of the Nelson Memo, has left OSTP for a position in the White House Office of Management and Budget.

Delia Mihaila has left her leadership position at MDPI to become Academia.edu’s Chief Publishing Officer.

Deanna Raineri has joined Wiley as the company’s new Senior Vice President for University Strategy and Market Innovation.

Sarah Tegen has taken on a new position at the American Chemical Society as Acting Publisher, Chemical & Engineering News.

***

Gordon Moore, cofounder of Intel and inventor of Moore’s Law, died this month. Aside from the obvious technological advances that have revolutionized scholarly communication, the Gordon and Betty Moore Foundation provided essential support to get PLOS off the ground and running, offering a revolution of a different sort.

Patricia (“Pat”) Schroeder, former CEO of the Association of American Publishers (AAP), has passed away. Schroeder led the AAP when the organization filed (on behalf of member publishers) a lawsuit against Google and its plans to digitize books without the permission of copyright holders—a case with parallels to the situation now faced by publishers and AI companies (see #4 above). Prior to her tenure at AAP, Schroeder served for 24 years in the US House of Representatives, representing the State of Colorado’s 1st District.

Briefly Noted

8

As we go to press on this month’s The Brief, a federal judge has issued an overwhelming defeat to the Internet Archive (IA) in the lawsuit filed by four publishers over their National Emergency Library program and controlled digital lending (CDL) in general. In making a summary judgment against IA without the case going to trial, the judge declared that IA’s claims failed on all four factors used in determining Fair Use. IA’s book scanning program was declared to be non-transformative, as all it did was make direct copies of the works (unlike Google’s efforts which were used to create something new, an online content index). IA plans to appeal. The Scholarly Kitchen has a helpful set of commentaries, including one from Joe Esposito.

Sage refreshes its branding. 

Sage has purchased the Epigeum online course platform from Oxford University Press (OUP). The platform originally launched out of Imperial College London and was bought by OUP in 2015.

Informa’s (owners of Taylor & Francis) 2022 full-year financial results show a “…continued strong performance in traditional pay-to-read publishing combined with further progress in Open Research…,” which delivered 3% revenue growth in 2022 (up from 2.4% in 2021).

Is this the supercontinent we’ve been waiting for? R Discovery from Cactus Communications has become the world’s largest repository of OA journal articles.

cOAlition S has released its 2022 Annual Report. The statistics offered for the 79% of funded articles in 2022 that were compliant suggest that, as expected, this is largely a Gold OA policy, with 80% following this path. The 2024 deadline for ending hybrid support looms large as nearly one-third of those compliant articles were published in hybrid journals. Also of interest, despite increased attention paid to the Rights Retention Strategy in 2022, the number of articles taking this route (18% of the total) declined by more than 26% from 2021.

Twitter has announced the prices for access to its (formerly free) API (application programming interface)—a range between $42,000 and $210,000 per month, depending on what percentage of tweets one wants to see. This astronomical price increase leads us to wonder about the viability of continuing to include Twitter data in Altmetric. 

The Royal Astronomical Society announced that it is flipping all of its journals to OA by the beginning of 2024. In response, Lisa Janicke Hinchliffe wonders when these sorts of announcements become so commonplace that they no longer grab our attention.

The National Academies of Sciences, Engineering, and Medicine have released a report on Advancing Anti-Racism, Diversity, Equity, and Inclusion in STEM Organizations: Beyond Broadening Participation.

The best explainer we’ve seen yet on how ChatGPT works, from Stephen Wolfram. 

Analyzing what is surely a terrifying sign for society, The New Yorker takes a deep dive into the massive decline in the number of undergraduates majoring in the Humanities.

The results of a 2021 survey of scholarly monograph publishers have been released, showing an increased transition from print to digital and lingering uncertainty about OA. Meanwhile, preliminary results from the TOME (Toward an Open Monograph Ecosystem) pilot are out. Given the long lead time in publishing a monograph, most of the books covered came out in the last few years of the five-year project, so the numbers only look at a smaller number of books released earlier in the pilot. The good news is that the digital OA monographs outsold “comparable” titles in print; the bad news is that those results are loaded with a significant number of caveats and may not accurately reflect future buying patterns in a fully-OA market.

Russia’s invasion of Ukraine continues to impact science, as a new study suggests that the Russian scientists shut out of working with many international partners are now shifting their collaborations to China and India.

Did a journal publish a scientific study of a plastic toy shark?

***
What a school thinks about its library is a measure of what it feels about education. —Harold Howe (US Commissioner of Education under President Lyndon Johnson)