a blog about genomes, DNA, evolution, open science, baseball and other important things

At the now famous 2001 meeting that led to the Budapest Open Access Initiative – the first time the many different groups pushing to make scholarly literature freely available assembled – a serious rift emerged that almost shattered the open access movement in its infancy.

On one side were people like me (representing the nascent Public Library of Science) and Jan Velterop (BioMed Central) advocating for “gold” open access, in which publishers are paid up-front to make articles freely available. On the other side was Stevan Harnad, a staunch advocate for “green” open access, in which authors publish their work in subscription journals, but make them freely available through institutional or field specific repositories.

On the surface of it, it’s not clear why these two paths to OA should be in opposition. Indeed, as a great believer in anything that would both make works freely available, I had always liked the idea of authors who had published in subscription journals making their works available, in the process annoying subscription publishers (always a good thing) and hastening the demise of their outdated business model. I agreed with Stevan’s entreaty that creating a new business model was hard, but posting articles online was easy.

But at the Budapest meeting I learned several interesting things. First, Harnad and other supporters of green OA did not appear to view it as a disruptive force – rather they envisioned a kind of stable alliance between subscription publishers and institutional repositories whereby authors sent papers to whatever journal they wanted to and turned around and made them freely available. And second, big publishers like Elsevier were supportive of green OA.

At first this seemed inexplicable to me – why would publishers not only allow but encourage authors to post paywalled content on their institutional repositories? But it didn’t take long to see the logic. Subscription publishers correctly saw the push for better access to published papers as a challenge to their dominance of the industry, and sought ways to diffuse this pressure. With few functioning institutional repositories in existence, and only a small handful of authors interested in posting to them, green OA was not any kind of threat. But it seemed equally clear that, should green OA ever actually become a threat to subscription publishers, their support would be sure to evaporate.

Unfortunately, Harnad didn’t see it this way. He felt that publishers like Elsevier were “on the side of the angels”, and he reserved his criticism for PLOS and BMC as purveyors of “fools gold” who were delaying open access by seeking to build a new business model and get authors to change their publishing practices instead of encouraging them to take the easy path of publishing wherever they want and making works freely available in institutional repositories.

At several points the discussions got very testy but we managed to come to make a kind of peace, agreeing to advocate and pursue both paths. PLOS, BMC and now many others have created successful businesses based on APCs that are growing and making an increasing fraction of the newly published literature immediately freely available. Meanwhile, the green OA path has thrived as well, with policies from governments and universities across the world focusing on making works published in subscription journals freely available.

But the fundamental logical flaw with green OA never went away. It should always have been clear that the second Elsevier saw green OA as an actual threat, they would no longer side with the angels. And that day has come.

With little fanfare, Elsevier recently updated their green OA policies. Where they once encouraged authors to make their works immediately freely available in institutional repositories, they now require an embargo before these works are made available in an institutional repository.

This should surprise nobody. It’s a testament to Stevan and everyone else who have made institutional repositories a growing source of open access articles. But given their success, it would be completely irrational of Elsevier to continue allowing their works to appear in these IRs at the time of publication. With every growing threats to library budgets, it was only a matter of time before universities used the available of Elsevier works in IRs as a reason to cut subscriptions, or at least negotiate better deals for access. And that is something Elsevier could not allow.

Of course this just proves that, despite pretending for a decade that they supported the rights of authors to share their works, they never actually meant it. There is simply no way to run a subscription publishing business where everything you publish is freely available.

I hope IRs will continue to grow and thrive. Stevan and other green OA advocates have always been right that the fastest – and in many ways best – way for authors to provide open access is simply to put their papers online. But we can longer pretend that such a model can coexist with subscription publishing. The only long-term way to support green OA and institutional repositories is not to benignly parasitize subscription journals – it is to kill them.

In 2005 I submitted an essay about science publishing to a political magazine. I got a polite reply back saying that the article was interesting and the issue important but that my approach wasn’t right for them. My piece was too straightforward. Too persuasive. They preferred articles that had a simple “hook” and, most importantly, were “counterintuitive”.

Zoom forward a decade and I finally get what they were looking for. In the last few months two articles about open access have appeared in political magazines, both having “counterintuitive” points.

The first,  “The Duck Penis Paradox: Is too much Internet pop science drowning out the serious stuff?” by Alice Robb appeared in September in The New Republic. I spoke to Robb extensively as she worked on the article (although I got labeled “voluble” for my efforts), and as I started to read it, I was reasonably pleased. Although she was a bit flippant, Robb did a credible job of describing the motivation behind PLOS ONE and our rise in the publishing world.

But then she got to her “counterintuitive” point:

So, in many ways, Eisen has won. More people have more access to more studies than ever before. Science has never been so democratic. It’s just not clear whether democracy is what science needs.

Robb goes on, describing how actually reading about the variety of science people are doing gave her a headache, and laments the potential loss of filters:

The traditional journals may be inefficient, but they serve a purpose. By establishing a hierarchy, they help direct scientists’ and journalists’ limited attention to the research that deserves it.

So, basically, Robb was complaining that PLOS is bad because it is anti-elitist – that we may not like elitist journals, but we NEED them, lest we leave poor science journalists dangling in the wind, forced to actually read papers and figure out what’s interesting on their own.

Nevermind, that said meritocracy is demonstrably flawed. Nevermind that the current system of peer review sucks at identifying good quality and important science. Nevermind that anyone who pays attention to science – and Science – should know “high quality” journals routinely publish crap. After researching the issue, Robb concluded that even a dysfunctional elitist hierarchy is better than no elitist hierarchy.

In retrospect, this should not have surprised me. For as long as I can remember – and long before that too – The New Republic has been a great defender of our current “meritocracy” in all areas of life. So why should it be a surprise that they view efforts to democratize science as a bad thing.

Robb’s piece of reminiscent of an editorial that appeared in The Harvard Crimson shortly after PLOS ONE was launched:

Getting into Harvard is hard, very hard. Yearly the gatekeepers in Byerly Hall vet thousands of applicants on their merits, rejecting many times the number of students that they accept. But getting a scientific paper published in Science or Nature, today’s pre-eminent scientific journals, is oftentimes harder.

Science, like much of academia, has its own admissions committee. Though over a million manuscripts are published in journals yearly, many more are submitted and rejected. The gatekeepers of science—peer reviewers who are reputable scientists and well versed in a particular field—advise journal editors whether to reject a manuscript outright, send it back for revisions, or publish it.

…

Without a peer review process to separate the revolutionary papers from the merely good from the rubbish, scientists will have no way of knowing which discoveries and experiments merit their time and interest. Instead, they will spend inordinate amounts of time wading through the quicksand of junk science to get to truly interesting work. Peer reviewers are chosen as peer reviewers for a reason—unlike the hoi polloi that roam the Internet, they have the knowledge and experience to judge scientific research on its merits.

I responded at the time:

As a Harvard graduate and co-founder of the Public Library of Science (PLoS), I was appalled by your editorial, “Keep Science in Print” in which you condemn our new journal PLoS One. The article is too ill-informed and riddled with factual inaccuracies to be taken seriously as an attack on our efforts to rejuvenate peer review by opening up the process to all members of the scientific community. I would normally feel compelled to correct all these errors, but fortunately I don’t have to. Perhaps sensing the opportunity for delicious irony, the “hoi polloi that roam the Internet” have identified and corrected your mistakes in the open commentary you provided for this article.

They did not, however, respond to your repellent effort to rally the forces of elitism to derail a project whose primary aim is to rapidly bring scientific knowledge to everyone. Elite scientific journals are, you argue, like the Harvard admissions committee—carefully separating revolutionary papers from the merely good, just as Byerly Hall culls the unworthy from the ranks of each year’s freshman class. I couldn’t agree more. The two are very similar—and both are deeply flawed. It is impossible for even the smartest scientists to recognize the true merit of a paper before it is published, just as it is impossible to identify the smartest and most talented scholars on the basis of their high school grades and SAT scores.

Think, if you will, of PLoS One as a large public university—our doors are open to papers that might not earn admission to Science or Nature. But, over time, many of these papers will turn out to be outstanding. Once they see PLoS One, we are confident that consumers of scientific papers will discover what employers have long ago: If you’re looking for the imprimatur of greatness, try Nature or Harvard—but if you want the real thing, try PLoS One or Berkeley.

Although I am disappointed that the conversation about PLOS ONE hasn’t really changed in a decade, both The Crimson and TNR were right in calling PLOS ONE an attack on elitism in science. We just differ in whether we think that’s a good thing.

With this critique of PLOS in mind, it was surprising to read an article published earlier this week, “Free Access to Science Research Doesn’t Benefit Everyone” by Rose Eveleth that comes at open access (and open science in general) with a different “counterintuitive” point. She too starts off with a generally favorable outlook on openness, but quickly comes to a different conclusion: that PLOS is TOO elitist:

Making something open isn’t a simple check box or button—it takes work, money, and time. Often those pushing for open access aren’t the ones who will have to implement it. And for those building their careers, and particularly for underrepresented groups who already face barriers in academia, being open isn’t necessarily the right choice.

…

Melissa Bates, a physiology researcher at the University of Iowa, says that when it comes to making papers open access, it’s not fair to ask graduate students and early career scientists to bear the brunt of the responsibility. “There’s this idea that open access is this ethical and moral thing, that it’s a morally and ethically grounded movement, and I can appreciate in a sense that it is,” she said. “But there’s also a business model to how science is done.”

That business model isn’t all that different in science publishing than it is in any other kind of print publishing. Putting out a journal costs money. And someone, whether it’s the university, the scientists, the government, the public, or some benevolent billionaire, has to pay for it. Much scientific research is funded by taxpayers. But the editorial process—the printing, the hosting, and the rest of it—is not. “In principle, Open Access is what I call doing the right thing,” said Alan Leshner, the executive publisher of Science, a journal the keeps its papers closed for the first year after they’re published, and then opens them up to the public. “It would be great if we could afford open access to everything we publish immediately. The problem is it costs $50 million a year to publish Science.” Somebody has to foot that bill, he says.

When a paper is accepted to a journal that isn’t automatically open access, in some cases scientists can pay a certain amount of money to release it to the world. Those publishing fees can be thousands of dollars for each paper. Open-access advocates argue that it’s worth the money to put the work out there, but Bates points out that often grants will have a limit to how much someone can spend on publishing fees. Gezelter says that that economic tension is a big one in labs. “Would you rather publish these 10 papers open access or would you rather hire a grad student for a year?” he asks. “It leaves individual scientists in an ethical quandary,” Bates said. “The answer for me is always going to be: I’m going to pay a person.”

There’s a lot to unpack here, but Eveleth’s basic argument is that open access is a high-minded ideal being pushed on young scientists by an elite who don’t understand, or don’t appreciate, the challenges of doing science in the modern world. I agree completely that the system as a whole pushes people away from open access, both in terms of career development (pressure comes from many directions pushing people to publish in the highest impact journals, irrespective of how they are run) and financially (universities heavily subsidize the costs of getting access to subscription journals, but do little to offset the costs of open access journals). There has been a tendency in the OA community (myself included) to put our hopes in young scientists (since the publishing behavior of most established scientists has proven themselves to be beyond amendment.  But that’s not fair or reasonable (something previous interactions with Bates helped me to appreciate). So there is value in this piece for shining light on a aspect of open access that hasn’t received a lot of press play.

I really like Eveleth’s writing. But I feel that this piece did not do justice to the past or present of open access in several important ways.

One of the central premises of the story is that costs associated with open access publishing (or open source software) make it a luxury that many can not afford. There is some truth to this – the move to open access publishing has shifted the way in which money is transferred from scientists to journals. Although it costs the system way less when people publish in open access journals – the average revenue for subscription journals is around $6,000 a paper, more than even the most expensive open access journals, and several times more than the cost of publishing in PLOS ONE – subscription costs are almost completely subsidized by universities, while open access charges rarely are. Thus the money it takes to publish in an open access journal comes out of research funds, while subscription costs do not.

However Eveleth raises this issue as if it’s something new – an unexpected, and unappreciated, side effect of open access publishing. But this is not a new problem. Supporters of open access have long been aware that until the ~$10b currently spent every year on subscriptions is diverted and used to support publishing in other ways, mechanisms must be developed to help authors whose grant funds are insufficient to cover up front charges to publish in open access journals. And Eveleth fails to mention the many initiatives designed to address this issue. PLOS (and many other OA publishers) offer fee waivers to authors who are unable to pay the publication fee, and to my knowledge PLOS has never turned away a paper on financial grounds. Furthermore, many funding agencies will cover the costs of OA for their grantees. And an  increasing number of universities have open access funds that will cover or help defray these costs for scientists at their institutions. Bates’ own University of Iowa has such a fund, although it is limited to researchers without grants.

And the situation with publishing costs is far more complicated than the story lets on. Many subscription journals also charge authors who publish there – in some cases more than it costs to publish in open access journals. Bates, for example, published an article in the Journal of Applied Physiology last year.  This journal charges authors a $50 submission fee, and $75 per page in the final PDF. At 9 pages, this article would have cost them $725. That’s a bit less than publishing in PLOS ONE, but not more than it would have cost to publish in Peer J. And this is low compared to the cost of publishing in other subscription journals. PNAS charges $1700 per article, for example. While it’s “free” to publish in Science, Nature or Cell, they charge ~$1000 if you have a color figure (which most articles do). Thus, at an institution that has funds to support open access publishing, it might actually be cheaper to publish in open access journals than in many subscription journals.

I am not denying that that people are under severe financial pressure these days, and there are certainly many authors who do not have access to institutional funds to cover these costs. It’s a systemic failure when funding agencies (e.g. the NIH) and institutions that claim they support open access publishing but leave authors in a position where they have to choose between publishing in open access journals and having some extra research funds. But it was incorrect of Eveleth to suggest that these financial challenges are unique to open access.

This has been a disturbing trend in journalism about open access lately. It’s become fairly common for people to take a problem with publishing, note that this problem applies to open access journals, and make this a problem for open access. The most egregious example was the “open access sting” carried out by John Bohannon in which he submitted a bogus paper exclusively to open access journals, found that many accepted it, and concluded that open access journals had a problem with peer review. If we are worried about ensuring all scientists have unfettered ability to publish their work – as we should be – we should worry about obstacles to publishing in all journals, not just open access ones.

Leaving the author charges issue, Eveleth  chose to wade briefly into the broader economics of scholarly publishing, quoting Alan Leshner, the outgoing publisher of Science, citing the fact that it costs $50,000,000 to publish Science, and complaining, “Somebody has to foot that bill.” But this point is left hanging – with no discussion or response. By doing this Eveleth says to her readers – many of whom, because the story was published outside of the science press, are learning about open access for the first time – that these is a valid and open criticism of open access, for which there is no response. When, in reality, Leshner has been saying the same thing for over a decade, and I and other open access advocates have a detailed response. I don’t necessarily expect Eveleth to rehash the whole open access debate, but to leave it seeming that this is some kind of new, unanswered critique of open access does not do justice to the history of this subject.

There are several problems with Leshner’s statement. Yes, it costs $50,000,000 to publish Science. And there is no way these costs could be covered by the thousand or so authors of research articles it publishes each year ($50,000 a paper would tax even the most well-heeled labs). But the fact that Science can not come up with a business model that would allow it to make the papers it publishes freely available is not a problem with open access, it’s a problem with Science.

One of the main reasons that Science is so expensive (its cost of ~$50,000 per paper is roughly 10x the industry average, which is already absurdly high) is that it employs highly paid editors to screen papers, and rejects the vast majority of them. I don’t know the exact numbers, but probably only one in fifty submissions is ultimately published. Thus, even with a fairly gilded staff, their cost per submitted paper is a much more reasonable $1,000. The problem with Science (and Nature, Cell and other high profile journals) is that this “review but reject most papers” is that it’s a relic of the print age, when space in a printed journal was limited by the cost of paper and shipping. But those costs are gone. And instead Science maintains a false scarcity to drive up the value of its brand. The alternative is a system in which we decouple the act of publishing and review – to have a system in which all papers are rigorously assessed, but where the assessment – whether good or bad – is simply published alongside the paper, rather than used as the basis for an absurd partitioning of papers into the 20,000 silos we call journals. (I’ve written about this more extensively here and here).  People might not agree this is a better solution – but given that Eveleth raised this issue, it is a disservice to the topic and her readers that she didn’t contextualize Leshner’s quote properly.

After raising the cost issue, Eveleth moves on to argue that open access is a also luxury of a non-financial sort – that only to people who are well established, and that publishing in open access journals is intrinsically bad for one’s career. I know that everybody believes that a paper in Science, Nature, Cell, NEJM or JAMA  is a ticket to career success, and that to some extent this becomes a self-fulfilling prophecy. But I think that – despite this near universal perception – that the effect isn’t nearly as strong as people think. There is certainly a correlation between career success and publishing in these journals. But as we all know, correlation does not necessarily imply causation, and I think it’s very possible that that people get jobs/grants/tenure as well as big 5 publications because the same criteria are applied in hiring, promotion and funding as are applied in selecting papers for publication.

I understand why Bates and most other scientists say that they will always choose to publish in these journals if offered the chance, because it is something under their control that they believe might lead to greater career success. But it’s disappointing that an excellent journalist like Eveleth just takes this assumption at face value instead of questioning it or at least pushing back on people who assert it as if it is a fact.

Finally, Eveleth makes the point that open access is elitist because it is particularly dangerous to pursue for scientists early in their careers. It is, of course, obviously true that scientists at different stages of their careers face different challenges. I am, personally, more able to take risks than, say a postdoc looking for a job, or an untenured, unfunded new PI. But nearly every paper I have ever published, and nearly every paper anyone ever publishes, has primary authors who are not well established. It’s the way science works. A graduate student, postdoc or other young scientists is the first author on the vast majority of papers published. And so nearly every paper involves someone in a vulnerable position in their career who would stand to benefit from whatever boost one gets from publishing a high impact paper. Thus the oft-repeated idea that there is some special subset of open access papers where the authors can safely publish in open access journals, while the authors of other papers can not, is, to a large extent, not true.

In saying that I am not trying to argue that Bates or any other scientist should be asked to gratuitously endanger their careers for the greater good. Or that everyone faces anything remotely like equal challenges in building a successful career in science. Rather I think it is important to note that the concerns Bates expresses apply far more broadly than the article implies. Indeed, as successful as open access publishing has been, it is one of the movement’s great failings that we have not succeeded in upending the system to the extent that people like Bates, who appears to genuinely support the ideals of open access, feel like publishing in open access journals is the best way to build their careers. Until we change this, the movement for greater openness in science will not succeed.

So, despite its failings in accurately representing open access, Eveleth’s piece serves a useful purpose. I believe the open access movement is driven primarily by anti-elitist sentiments – a desire to free information, to remove its control from the forces of commerce, and to break down the elitist hegemony of high-profile journals. But the elitist risks in open access are real. I don’t think they’re the fault of the open access movement – we have tried from the beginning to have the powers that control the funds used on subscriptions use them instead to fully subsidize open access fees; we have tried to undermine and ultimately destroy the impact factor driven culture of high-profile journals and their perceived role in hiring, funding and promotion. But the forces of inertia have, so far, been too strong. But our fault or not, it is crucial that we listen to the concerns of young scientists like Bates and try to make sure that open access really is accessible to everyone.

[NOTE: In the original version of this piece I suggested the Iowa open access fund would have covered Bates’ open access fees. It wouldn’t have as it was restricted to researchers without grants. I apologize for suggesting otherwise and for being an asshole about it.]

Macmillan, the publisher of Nature and 48 other Nature Publishing Group (NPG) journals, announced today that all research papers published in these journals would be “made free to read in a proprietary screen-view format that can be annotated but not copied, printed or downloaded”.

If you believe, as I do, that paywalls that restrict the free flow of scientific knowledge are a bad thing, then anything that removes some of these restrictions is a good thing.

This move is fairly typical of Nature as of late. Despite its place as one of the oldest and most august big Kahuna in the subscription publishing world, Nature – and especially its Digital Science division – have been far more attuned to the ways that the Internet has changed publishing than their competitors. And, because of the rise of open access publishing and funder efforts to provide access to their papers, people increasingly expect to be able to access scientific publications, and Nature is responding to that expectation.

There are really two parts of this announcement.

1. A smallish (~100) media outlets and bloggers will be able to provide a link to Nature papers they are writing about that will allow readers that will allow them to access them for free.
2. Subscribers to Nature and other NPG journals will be able to generate and share such links by email, on Twitter, etc…

It’s actually kind of brilliant on Nature‘s part. They are giving up absolutely nothing. Readers of news stories about Nature articles were never going to pay to access the actual articles (like other publisher Nature has tried a pay-per-view system that has completely failed). And individuals and institutions that subscribe to Nature aren’t going to give up the convenience of being able to read articles on demand for the challenge of finding a link on Twitter (unless someone were to set up a database of these links…. hmmm….).

And let’s remember that subscribers to Nature were already sharing copies of downloaded PDFs quite abundantly. This was not, as Nature argues happening in an inconvenient way in the dark corners of the Internet. This was happening in email and on Twitter. The problem was that Nature had no control over this sharing. So, really, they’re not changing people’s ability to access Nature very much – what they’re doing is changing where they access it – likely with the hope that they will figure out ways to monetize this attention.

Thus Nature gets lots of goodwill, more people reading their papers, and they lose nothing in the process. At least not immediately. Because the irony of a system like this is that it can’t ever actually do what it purports to do. If it ever actually made it possible to find and get free access to any Nature paper, then people actually would stop subscribing and they’d have to end this kind of access.

At the end of the day, this is a pretty cynical move. I’m sure the people at Nature want as many people as possible to read their articles. But this move is really about defusing pressure from various sources to provide free access. Yet Nature knows that they can’t really provide free access without giving up their lucrative subscription business model, which they are unwilling to do. So they do something that makes it seem like they are promoting free access, while doing nothing to address the real obstacle to free access – subscription publishing.

It is also worth noting how Nature is defining access down. First we had “open access” in which people can download, read, reuse and redistribute content. Then we had “public access” in which people can download and read content. Now we have “free access” in which people can read for free in a proprietary browser, and can’t download or print. This is going in the wrong direction, and it would be a disaster for science if – as Nature clearly hopes – this is the definition of access that sticks.