The Best Argument in Favor of Open Access Science is All of Them

Exactly - journals are obsolete. They no longer add value and instead leach money away from the system while at the same time erecting barriers to access.

I once saw a bit of political commentary that said that the political spectrum is less like a line and more like a horseshoe, where the far extremes on both sides are actually closer together than center-right is to center-left. Maybe this is an example of that? In any case, I'm happy to build a coalition with anyone that thinks we can do better than the current system.

I'm sympathetic to this argument, though I disagree. My statement "they don't add value" is perhaps a bit extreme (I've been called on this by other commenters in previous posts), but the truth is that I don't think they add value that's anywhere near commensurate with the costs to research. Personally, I don't even think open access journals go far enough, though they are certainly better.

A system that has worked well for 150+ years can’t suddenly become obsolete.

I don't think it's sudden at all. The internet has been around for quite a while, and just as newspapers, book publishing, the music industry etc are being radically altered, so too must science catch up to modern technology. I would have hoped that science would be on the cutting edge rather than being the slowest to reform. And it's also worth noting that the current models are not really 150 years old. Yes, the first academic journals started around then, but to take one example, peer-review (the way we do it now) has only been standard for 50 years or so. Watson and Crick's double helix paper wasn't peer reviewed, and that turned out fine.

magine if research results were released on blogs such as this for free – do you think that would be a good thing? How cold that possibly be standardized and regulated? The scope for fraudulent data would be so big that no-one would trust anyone else.

Yes, I do think this would be a good thing. It could be standardized and regulated using things like alt metrics, new platforms for open (or at least transparent) peer review and the like. There's not really a framework for this yet because no one has bothered to build one.

But the idea that radically open science would be more prone to corruption strikes me as farcical. What do journals do to prevent fraudulent data? Maybe they can run some fancy algorithms over images or something, but it would be trivially easy to manufacture huge swaths of data that would pass muster with peer review. And the closed nature of publishing makes findings that are incorrect (because of fraud or just because science is science) much harder to overturn. Journals rarely publish negative results, or replication studies, and papers are rarely retracted unless there's blatant wrong-doing.

Sure, people could publish as much BS as they wanted on their blogs, but there are so many journals these days that just about anything can be published anyway. The incentives to not do this would be just as strong, if not stronger, because labs prone to publishing false or misleading data would be rapidly identified and shunned. I think there's even a role for journal-like entities to exist in this sort of ecosystem - curating the best labs and the best stories and charging a small fee for subscription to reviews or lists that they generate. But the cost would be lower, and the incentive to continue to add value would be stronger because they would fail if they did not.

Ultimately, I have faith that the scientific community would have no trouble separating the wheat from the chaff.

I’m afraid they will inevitably become entrenched by capitalism regardless of the socialistic reasoning you put forward in their support.

I actually don't think so. The trouble is, the current model is shielded from market forces because of institutional entrenchment. As long as researchers need journal publications for promotion and grants, they will not seek alternative methods of publishing. This is why I think we need some top-down pressure to start eroding the institutional barriers so that other ideas can flourish.

By the way, full-time editorial staff do far more work than you might think so that you get to read a legible logical piece of research from a non-English speaking researchers.

I'm sure this is true, and in my vision, I see a lot of specialist services springing up to do things like this. Probably with significantly reduced cost.

Finally, I’m also an immunologist by trade (PhD) and enjoy your immunology posts.

Thanks! Be sure to let me know if I screw anything up :-)

“There’s not really a framework for this yet because no one has bothered to build one.”

And why has no-one bothered? Why don’t you and some like-minded people ‘bother’ to change the system if it is so bad?

Imagine a world where your budget for buying books was directly linked to how often you purchased books from a brick-and-mortar store. Now I come along and propose that selling books online would make it easier and cheaper to buy books. "That's nonsense," everyone says, "how could you possibly distribute the books, how would people know what they like if they can't leaf through the pages, why would people be willing to spend money online?"

There were a lot of unknowns when Amazon.com got started, but they've become enormously successful despite many potential obstacles, and their success has even paved the way for whole new technologies and industries (like ebooks). But could Amazon have ever become successful in the world I described? No matter how much more convenient, no matter how much cheaper, no matter how much better for consumers, as long as your income is tied to a particular system, no other models could possibly compete.

The argument "why not try it and see what happens" argument is absurd. Of course it will fail. Why would scientists publish anything online outside the framework of journals when they will get no credit for the work in the current system? Even I, as passionate as I am about the need for change, could not possibly put my research online for all to see. If someone took my knowledge and published it in a journal, I'd lose all credit, it would be harder for my lab to get funding and everyone in it would suffer.

But I have data that it would be useful for my field to know. It could have impact on a lot of the experiments that people are currently doing. It's not quite ready to form a complete story, and we're waiting to publish to see if we can get it into a higher impact journal. But even if it were ready tomorrow, it would likely take at least a year to get it through review and publication - in the meantime, there are people who are doing the wrong experiments and wasting time and money, or the same experiments waisting time and money.

In short, basic business principles would bring us back to where we started.

It's possible that events would play out as you describe (I don't think so, but who can say for sure?). Do you honestly think that we'd still have journals operating with 30% profit margins? Do you think subscription fees would be bankrupting the library budgets of even enormously wealthy institutions like Harvard? Do you think that there are no innovations in distribution or review or editing or meta-analysis or anything that aren't being held back by the monopolies that publishers have?

The current market dictates that science journals are in high demand, so until the critics of this system actually provide a viable alternative

My point is that the focus of the market is distorted. No one denies that journals are in high demand in the current system. We're seeing a huge proliferation of new journals, and scientists will publish in them, because we get money from the state that's based on how much we publish in journals, so we funnel that money into publishing in journals. My point is that what drives the market is publishing, rather than what should actually drive the market: science. There is no market for scientific ideas and data that is separate from publishing, so other models don't stand a chance.

But as you admit yourself there is not even a framework for such an alternative, so at this stage it is simply wrong to imply that journals are the bad guys here and do nothing for science. It is like the inventor of a theoretical teleportation device driving to a conference to tell people that cars are obsolete and add no value, them driving home again.

Cars are useful in a world without teleporters, just as horse carriages were useful in a world without cars and academic journals were useful in world without the internet. I feel like we're in a world where teleporters have been invented, but no one accepts that they would be a better alternative to travel, because people only get paid to travel by car. Again, I accept that journals add some value - my statement that they proved absolutely no value is a bit extreme. Horse carriages still provide value too - it can be pleasant to be pulled around in a park by one. But I'd be bitter if people told me that I had to spend $30,000 on a carriage to get to work, when I know that cars can be cheaper and better.

As you say, publishing the results on a blog will leave you open to others stealing your work. But strangely you seem to think that if everyone published on blogs that this could not happen…?

That's correct. If the standard was open publishing, then credit would be established by publishing openly. The trouble with publishing openly now is that credit is only given to journal-published stories.

Your own actions indicate that there is not actually any viable alternative out there, although reading your blog posts one gets the impression that there is an alternative ready to go and that we should all engage in it immediately.

Sorry, this is not the impression I wish to give. I'm trying to advance the argument that we need to find mechanisms that give credit to data and ideas, in whatever form they are presented, rather than only giving credit to journal-published research. If journals end up being the best solution after all, then I would walk away happy. The trouble is that the journal-publishing model is artificially supported by funding practices - I'm trying to suggest that this needs to be changed.

I notice that in a previous post of yours you mention that Harvard was thinking of boycotting some major publishing companies.

Harvard Libraries were not threatening to boycott any publishing companies, they were encouraging researchers to voluntarily choose to publish in open-access journals. Their argument is that closed journals have prestige, so people continue publishing in them, but that researchers should try to shift prestige to open-access options. Whether this is a principaled stance or a business move, I can't say.

But consider how insulated from costs researchers are. For the most part, labs to do not pay the costs of publishing directly from their grants. Instead, institutional overhead is pulled out, some of which goes to library budgets, which then pay for me to read all of the literature I have access to. Ultimately, the money comes from the same place, but the people making the decisions about where to publish, what journals to review for etc, aren't bearing the cost of those decisions.

You're right that institutions (including Harvard) are given discounts from the list price, but those discounts are often dependent on bundling less desirable journals, keeping them artificially afloat. And the list-prices of journals are rising faster than inflation despite the fact that many of these publishing companies have enormous profit margins.

Sorry Kevin, I didn’t mean to criticize your personal choices of how you publish your research.

No need to apologize. I've been having this discussion a lot lately - I'm happy to take some lumps if I can at least raise awareness of the issue (even if I ultimately don't convince you).