bioinformatics, genomes, biology etc. "I don't mean to sound angry and cynical, but I am, so that's how it comes across"

Ripping apart that terrible Atlantic piece on Open Access

It’s genuinely not often that I read an article and disagree with every single point, but The Atlantic managed it with their terrible piece on Open Access.  However, when  one of your chief sources is the executive publisher of Science, a closed access glamour-journal, it’s perhaps not so surprising that the article is so bad.  A bit like asking Turkeys if they think Christmas is a good idea….

The author, Rose Eveleth, mentions a few times the moral side of open access, but states that it’s not as simple as just looking at morals and ethics.  For me, the moral, ethical argument is the strongest of all the arguments – that those who created the research, and those that paid for the research, should have free and open access to the outputs.  Surely this is self evident.  Add a third group, those who most benefit from the research i.e. students.  All 3 groups should have free access to research outputs.  Forget money, forget business, forget everything else – open access is the right thing to do.  

The article staggers from one blunder to the next:

“Making something open isn’t a simple check box or button—it takes work, money, and time.”

This ignores the argument that making something closed also takes work, money and time.   Compare if you will how much it costs to run arXiv or bioRxiv, compared to the cost of running Nature or Science.  It’s not the “making open” that costs money.  How about the cost of PLOS?  PLOS ONE publishes more papers than any other journal, is open access, and is cheap to publish in.  The argument above is nonsensical.  Open and closed journals cost money, and I would argue that closed journals actually cost more to run.  Perhaps the above sentence should be rewritten:

“Making money from science publishing isn’t easy – it takes work, money and time”

We then move on to a bizarre section, where Dan Gezelter, who will probably regret this for the rest of his career, seems to suggest that people don’t release their code because they would have to document it if they did:

“It’s only scientists who are relatively secure who can spend the time [to document], and it does take extra time to make sure that their stuff is released correctly”

First of all, coders document.  They have to.  If they don’t, even they don’t understand what on earth their code does.  Also, if you don’t write user documentation, then only you will ever be able to use your code, and you are not doing your job properly.  Documenting code, including creating user documentation, is part of being a coder.  If you’re not doing it, you’re an idiot, and a bad scientist.  This should never be a barrier to releasing code.

Not to mention that released, open-access but undocumented code is still a million times better than no code at all.  I’m seriously embarrassed that some of these arguments made it into print.

After that, we move on to a section with quotes from Melissa Bates, who apparently argues:

it’s not fair to ask graduate students and early career scientists to bear the brunt of the responsibility

But is anyone actually doing that?  This is a massive straw man.  Of course it would be unfair to ask young scientists to bear responsibility, and it’s for that exact reason that no-one is asking them to!  Melissa then goes on to state:

“But there’s also a business model to how science is done.”

Sure.  There’s a business model behind drug cartels too – that doesn’t mean they’re right, and it doesn’t mean we shouldn’t try and abolish them.

The article staggers on to a section where Alan Leshner, executive publisher of Science, talks about the cost of putting out a journal.  Given that open access threatens the business model of journals such as Science, it’s no surprise Leshner seems against it:

“The problem is it costs $50 million a year to publish Science. Somebody has to foot that bill”

That is incorrect.  No-one needs to foot the bill.  We don’t need Science.  Nothing bad would happen if Science didn’t exist.   Life, and research, would go on.

Moving on to some more quotes from Gezelter, we get to a stage where the article compares apples with oranges and comes up with 5.  Apparently, the additional cost of making a closed access paper open access is too high and funders won’t pay it.  This ignores the fact that open access journals, such as the PLOS journals, PeerJ and FrontiersIn are all far cheaper to publish in than Nature, Science and Cell.  It costs money, many thousands of dollars, to publish in closed access journals.  Then one has to pay additional thousands to make the article OA.  Publishing in a straight open access journal in the first place saves money.  How can anyone get it this wrong?

Next we move onto the impact factor argument, and here I have some sympathy – the article states:

At a time when the job market in science is extremely competitive, the institutions combing over resumes aren’t looking for someone’s commitment to the open-access cause, they’re looking at their potential for big research.

And I think that is correct.  In the current climate, if you are a young scientist and you have a chance to publish in Nature, you should grab it with both hands.  It will make your career.  But lets not overstate this – I’m an open scientist and I have a good career.  My first paper was in BMC Bioinformatics, and I haven’t stopped publishing in Open Access journals since.  Titus Brown is an open scientist and has tenure.  There are others.

Moving on to impact factor itself, the most atrocious piece of cherry picking occurs – comparing Nature‘s IF of 42 to PLOS ONE‘s impact factor of 3.5.  Of course, PLOS’s flagship journal is PLOS Biology with an IF hovering around 11 (Genome Biology, another OA journal, has a similar IF).  Comparing Nature‘s flagship IF with PLOS’s lowest is a very low, deceitful thing to do – shame on you Rose Eveleth.  You would never make it as a scientist, we don’t cherry pick data to suit our arguments.

Obviously Nature‘s IF is still higher than the best PLOS can produce, but let’s not forget, Nature has been around since 1869, whereas PLOS Biology was created in 2003.  PLOS has a lot of catching up to do.  Conspicuous by its absence in the article is the fact that impact factor correlates highly with the number of retractions, and that there have been calls for a retraction index to be published alongside the impact factor.  I wonder if Rose Eveleth knew about this and chose to ignore it?

We begin to get some sense when Virginia Barbour enters the article, arguing that impact factor is a bad way of measuring how good science is, and that open access is about culture change, which takes time.  She says culture change is hard, and I agree.  We can all think of sins of the past that required culture change which was hard – but we did it anyway, because it was the right thing to do.  It’s what people do, it’s what each new generation does – we fix the mistakes of the past.  Open access fixes mistakes in scientific publishing.  Melissa Bates somehow manages to frame this natural, essential process as a bad thing – I don’t know how, and I cannot agree.  Some may try and make the new generation “cannon fodder” (as Melissa says), but quite simply, we should not and will not let them.

Finally, Leshner writes his journal’s own epitaph:

“I don’t see Science becoming open access in the near future”

I liken this to high street stores in the 90s, looking at the internet and saying “Meh, that will never catch on”.  10 years later they were dead, crushed by Amazon and eBay.  Adapt or die, that’s where publishing is right now.

What strikes me about this whole thing is how little value closed access publishers actually add.  Leshner is right when he says that the “circulation model” will die – the last time I went to the library and read a paper journal must be over 10 years ago.  This way of publishing science is dying, soon to be dead.  So what other role do publisher’s play?  In the era of the internet, they actually add nothing.  But here, perhaps, is their salvation.  Could they somehow add value?  I’ll draw an analogy to the cinema – why do I pay to go to the cinema when I can watch all sorts of great movies free at home?  Well, because the cinema has a bigger screen, 3D surround sound, popcorn, ice cream, hot dogs etc.  I enjoy the cinema.  It is a better experience than watching movies at home.  The cinema adds value.

So what value could publisher’s add?  Could they release all papers as open access, yet then provide additional tools (under a subscription model) that allow people to explore the data?  Maybe.  I don’t know what the “added value” is, but publishers will have to find it somewhere, because the current closed model of scientific publishing is dying; conversely open access is alive, is growing, and is the future.

As I said above – adapt or die.  Your choice.


  1. I would have quite a bit to say if I were responding to the original article, but the post does a great job of that. However, I have a bit of response to the closing question of the post.

    The premier closed journals do add value.

    The first kind of value they add is in the kind of peer review they generate, which becomes a sort of imprimatur. There is a stamp of acceptance, importance, and above all, interest. The papers there apparently generated motivation from the original researcher to submit – despite various opportunity costs, especially the high rate of desk rejection, and the long and often adversarial peer review. They also made it past these gates. Some informed people think they are widely interesting. F1000 was supposed to make this seal available more broadly; now, Altmetrics? It is interesting to contemplate their relative degrees of success.

    The glossy journals also provide a substantial amount of press. This cannot be underestimated. If the various better OA journals each found one scientist – per paper – to write a short precis for a general audience, and published it alongside, that would add immense value for authors in that journal. Further, Nature and Science tend to coordinate with science writers, newspapers, and so on.

  2. Do they do anything that an open access publisher couldn’t do? Our press releases are coordinated by the university – no need for Nature or Science

  3. University press releases rarely get picked up, except by the local papers. Nature and Science press releases get worldwide coverage, which is one of the incentives for publishing there, despite their tremendously high rate of retractions due to improperly vetted science.

  4. So do you mean that all publishers don’t add value, or that closed access publishers don’t? Because this is not entirely clear to me from text.

    Notably, Kent Anderson keeps updating (i.e. growing) his list of things that (all) publishers do – that presumably others don’t. I do, of course, appreciate that a. he is not necessarily the kind of role model to quote when discussing pros of OA, and b. some of the things of his list can and are done by others. But surely of the 82 things he lists in the October update there are some that really nobody else wants to do, no? Not saying this can’t change and why people don’t want to do them (they can’t be bothered, they don’t have time, they don’t have money, they don’t think it matters, it goes on).

    What else can (all) publishers do? There is a wealth of new ideas, and I really liked David Mittelman, Laurie Goodman and Razib Khan’s Genome Biology comment on it (I know that there are many people in the community who share these ideas and have plenty more of their own).

    What can scientists do? Keep pushing the publishers.

    Khan et al. comment: http://genomebiology.com/2014/15/12/556

    Disclaimer: I am an employee of BioMed Central. All opinions are my own (although I consider most of the above to be facts rather than opinions).

  5. The question is what do closed access publishers do that open access publishers do not? What do we get for our money (subscription) given the paper journal is now of little value?

  6. Ah, and here we are on the same page. I agree that not much, if anything.

    My guess is that the main argument would be: print. I know the concept is, given the circumstances, antiquated. But this is just about one thing that OA publishers cannot do, because all (?) OA business models rely on the voluminousness of OA publishing to sustain themselves.

    The day we stop printing out every single PDF (OA or not OA), it will completely cease to matter (not that it does matter much now anyway).

  7. I don’t think they do anything that a premier open access journal is categorically incapable of doing. However, this requires some exploration.

    Imagine that Journal X with great prestige spends a certain amount of money and time selecting articles of broad interest (with some normal or slightly elevated risk of retraction). Newspapers know these are likely to be interesting and even important papers, as do scientists in other fields, but both lack the expertise to contextualize the papers. The journal can commission a related expert to comment on the paper, putting it in context and providing additional art and figures to make it more accessible. If publication in Journal X is sufficiently prestigious, maybe the people asked to comment will do tons of work without monetary compensation, but it would be equally possible that a journal could simply pay for the commentary. Either way, the authors of the original article gain a huge amount of value. Almost without saying, there is the initial citation; and then the derivative news pieces. They can use the art in their talks and as a foundation for later review articles. More people understand the work and the interest is drummed up.

    Other journals, which are more focused on technical correctness and not broad interest, offer much less – in fact, most offer nothing at all. Not just open access, but also all those historically closed journals. Part of this is that they have less to gain, because the readers are typically better able to manage the content – they understand the context. Also, they would need to spend more money to get the same effort because commentators are less willing to work for the publicity. Further, Journal X probably publishes a relatively few articles – and relatively frequently, to keep content ‘fresh’ and to keep the audience attention. A broad audience has no stomach for 50 articles at the beginning of the month, but can handle skimming 6-8 a week.

    We could imagine three sample journals that are not presently Journal X. Maybe one is something like the Journal of Applied Astroturf, a specialist journal, respected but not widely read, and closed access. Another is Public Library of Sports Basketball (PLoS B), prestigious and open. The last is Public Library of Sports Triple Play (PLoS 3), technical accuracy only, high volume. Could these respective journals somehow provide the same kind of value as Journal X? For all their authors? For only some?

    Journal of Applied Astroturf is unlikely to reach out to casual browsers because it is closed. However, it might essentially enhance search engine optimization by creating a commentary minireview that attaches the citation to numerous other fields – contextualizing the work. That would drive traffic to both the journal and the article. Right now, we see that some journals try to do this ‘on the cheap’ by requiring ‘general audience abstracts’ from the authors themselves. This is to the publishers’ advantage because it squeezes even more effort out of the author and costs the journal nothing. However, to a certain extent, they get what they pay for. First, those abstracts don’t have prestigious names attached to them, saying nice things about the paper. People read them, I think, with the understanding that the paper does not get a bump in credibility as a result. Also, the authors write them to some extent with their own blinders on – an intelligent commentary by an intelligent peer actually adds something that the authors could not have provided themselves. Finally, these abstracts don’t draw authors to the journals – they presume a certain privilege on the part of the publisher, which may be eroding.

    PLoS B is probably assuming that the traffic it gets is already good enough, buoyed by openness. It might put a link to the homepage for a featured paper, but frankly, most traffic will come from search engines and databases, not from the homepage. There is some benefit to ‘tweeting’ about the article – but if done for every article, that loses its value quickly. Thus, it has relatively little incentive. Actually, if a journal like PLoS B wanted to increase visibility, perhaps it should increase the number of citations per article so that they are more broadly tied into the literature, which I think would raise visibility in the databases. This strategy doesn’t help Journal X because Journal X relies on casual readers and needs to keep page length down. Providing more graphics support in PLoS B would be the biggest gain for the authors, because graphics work is expensive and difficult, and more/better graphics would encourage the media to pick up the stories. It is an aspect of the Journal X strategy that could be duplicated to great effect.

    For PLoS 3, the journal may not be able to afford much investment ‘per article.’ Perhaps here, the authors are trading off other values for avoidance of rejection risk and revision requests. I’m inclined to think that a journal like this would provide the most benefit to the authors not by emphasizing publicity, but by providing grammatical editing, opportunity for on-going revision (like the version control of pre-prints), and rapid turn-around, each of which lower the energy barrier to publication.

  8. Mick,

    The question of open vs closed access is not a simple right vs wrong dichotomy. The main benefit of paywalled closed access model is that these publishers can afford to be selective. That’s what the value added is – selectivity. When I want to get a feel for what important new themes in biomed research are right now, I will skim the table of contents of Nature, Science, Cell, and a few top field-specific journals, and then read a few papers that catch my attention. I cannot possibly do that for every journal. Open access cannot afford to be selective because their business model depends on publishing large volumes of papers. The hybrid PLoS model has so far worked, but only because the top PLoS journals are heavily subsidized by PLoS One. PLoS One, because it is the cash cow of the enterprise, is incentivized to lower its barriers to entry, hence the decrease in its IF. I have not yet seen a respectable OA journal that can “carry its own weight” so to speak, i.e. does not have to be subsidized by a low IF partner. Until OA finds a viable business model that does not depend on publishing large volumes of sub-par science, we will be stuck with paywalled closed access journals. Another issue is that OA depends for its existence on the continual publication of papers. Once the steady stream of new papers runs out, the OA publisher goes bankrupt. Then who will foot the bill for keeping the servers with all these papers on them running? I really wish the OA model was the ultimate solution, but so far it has really not proved to be one.


  9. Explain to me, if the glamour journals are being selective, and this has value, why does impact factor positively correlate with the number of retractions?

  10. Retractions are in no way related to open or closed access. Any high “quality” open access journal will also have a correspondingly higher number of retractions. Selectivity is important because you cannot keep up with all the literature that may have significance for your own research, so you just browse high “quality” journals for interesting themes and techniques. I have found that keyword-based alerts are just not enough to get a broad view of the field. Anyway, if selectivity is not important, than why on Earth does PLoS have PLoS Biology rather than dump all the research – excellent and not-so-much, into PLoS one?

  11. Come on, don’t cherry pick your data. Your first point was that the top closed access journals publish the “best” science; I countered that what they publish is highly enriched for retractions.

  12. pawelniewiadomski

    1st January 2015 at 1:29 am

    What I am saying is simply that the selectivity of so-called top journals represents significant added value to a lot of people, myself included. That is despite all the drawbacks that such selectivity entails (retractions, glamour chasing etc). If selectivity does not hold intrinsic value to you, that’s fine, but it would be unwise to discount it as generally unimportant or even deleterious.

  13. pawelniewiadomski

    1st January 2015 at 1:35 am

    What I am saying is that selectivity has a lot of intrinsic value to a lot of people, myself included. That is despite all its drawbacks (retractions, glamour chasing etc). If it adds no value for you, that’s fine, but I think it would be unwise to discount it as generally useless or even deleterious.

  14. You are assuming that the selectivity is for high quality science; I’m not sure the evidence backs that up. They select for sensationalist science. Sensationalist isn’t the same as good, or high quality.

  15. I was shocked at Leshner’s statement that Science costs $50 million per year to publish. That’s outrageous, and indicates some major inefficiencies somewhere. He’s also ignoring the fact that the public pays something like $200K per study (on average) for NIH-funded work, and probably similar amounts for NSF and other funding agencies. Why should a privately-run journal own the rights to this work? He’s further ignoring the fact that the publishers have stubbornly refused to lower their subscription prices despite the dramatic drop in publishing costs allowed by online publishing and distribution: we just don’t need those print copies any longer.
    Mick, I agree with you: let’s just shut down Science and use the funds to support open access publishing instead. Science is published by AAAS, a nonprofit organization – of which I’m a Fellow btw. AAAS is supposed to promote science, but the publishers of the journal Science seem to be more interested in keeping it locked up. Shame on Alan Leshner.

  16. I’ve written a post on the subject on my blog. Feel free to visit and comment.

  17. Who says that (high-quality) online publishing is cheaper than print? PLoS and others still have huge expenses, and overall do not save money for the research community. All publishers, whether OA or subscription access, make material research results more broadly available than the would otherwise be, so accusing them of hiding results is just laughable. They invest in the distribution, vetting, and preservation of the scholarly record.

  18. There’s also a point of encouraging people to read science papers even if they aren’t scientists. A closed journal automatically shrinks its potential readership (really only Universities are able and willing to pay the subscription fees), and sends the signal that the papers “belong” to the journal rather than to humanity. A closed journal makes it appear to the public that science is stilted and not worth their efforts or attention. Their charges are ridiculous; $25 to read a single paper! I can buy many books on Amazon for that amount!

Leave a Reply

© 2018 Opiniomics

Theme by Anders NorenUp ↑