Killing the advancement of science on the Web is killing the advancement of science as a whole.
A few weeks ago I attended a panel session at the Commonwealth Club in San Francisco called, “Making the Web work for science.” (Video wrap-up here) Really, the focus was on how to get scientists to put their data up on the Web. Later that evening, I attended a post-reception gathering and briefly spoke with (name shall remain anonymous), the founder of a very popular social news media site; sites like Mixx, Digg, Reddit, Slashdot, Del.ic.ious, Stumbleupon, etc. I told him what I do and his enthusiastic response was, “Fuck the academic publishing industry!”
Now, perhaps his bold statement was due, in part, to the flow of alcoholic beverages at the event. There is more to it than that, however. He is not a trained scientist, but as the founder of a popular social news site, is very well aware of the state of science on the Interwebs. Like many people (notice the avoidance of the worn phrase ‘members of the public’ as if they are unable to contribute to science), he has a personal interest in the domain, and is trying to figure out how his business can enable scientists.
We have to wonder though, why are things so bad with science on the Web, that even an Internet entrepreneur recognizes the problem? And who or what is responsible for this mess?
Two groups are responsible: Scientists and, as you can tell, Publishers. Let’s start with the bad guys first, and then talk about the even more guilty party of bad guys.
Publishers
First, personally knowing several individuals on the inside of the publishing industry, they are decent people. My guess is the same could be said of most in their line of work. Yet, as Friedrich Nietzsche roughly stated, “Madness is something rare in individuals — but in groups, parties, peoples, ages it is the rule.”
Whether it be intentional, or just the nature of having to answer to shareholders, the academic publishing industry has very little incentive to get scientists to really utilize the Web/Internet the way it should be used. And this is hindering the advancement of science.
You think you control science? You think the NIH, Congress, or your respective governments do? Sorry, but it is the publishing industry that is in control.
They decide what is and isn’t fit for publishing. They decide when and where you can self-archive manuscripts and supplementary data. They mesmerize and trap us with the lure of journal impact factors like a baby unable to wean itself. Whether we are conscious of it or not, much of the direction taken in labs isn’t based on pure science, but based on what the editors of a high impact journal will publish.
There is no motive to change up their business model and encourage the open sharing of data on the Web. It took a 2007 law passed by the U.S. Congress to ensure that studies paid for by NIH funds (via US tax dollars) were made publicly accessible 12 months after publication.
The U.S. Congress is being lobbied by publishers to block similar proposals being extended to other research branches, such as NSF. Allan Adler, Vice President of the Association of American Publishers, has stated their reasons “It undermines [the] publishers’ ability to exercise their copyrights in the published articles, which is the means by which they support their investments.” He goes on, “The NIH policy also threatens the intellectual freedom of authors, including their choice to seek publication in journals that may refuse to accept proposed articles that would be subject to the new mandate.”
That last bit is a sprinkling of spin with a dash of scare tactics. And what exactly should we be protecting? An antiquated business model, which hinders the advancement of science?
This past February, I was on a panel discussion at the annual NFAIS conference, a popular forum for academic publishers. The conference theme was on digital natives in science. At one point I was asked (rather rudely) by a rep from a major publisher what exactly the new business model should look like for publishers in an Open Access world. My first thought was, “I don’t care if you find one or not. I’m here to advance science, not your bottom line.”
I was lucky to have escaped that conference with my head still intact. And they should go ask the editors of PLoS for an answer, they’ve certainly seemed to have found a successful OA business model.
There was an excellent article in Times Higher Education written by Zoë Corbyn just last week that dives even deeper into the publisher problem. Funny enough, I had started this article last week as well, which goes to show that this problem is beginning to be on the top of everyone’s minds. It isn’t just the publishers who are to blame though, it is the scientists as well.
Scientists
There seems to be a bit of a grassroots movement amongst younger scientists to move to an Open Access model and to start putting up more data on the Web. This is great! But it isn’t good enough.
Where is the leadership here? Why isn’t the American Academy of Sciences striving for obvious change? Are they just protecting their own lofty positions rather than science? Where are the de facto science celebrities who should be speaking out? Where are the university administrators enacting new hiring policies based not on which journal you publish in, but the science behind the work? Open data on the Web should be encouraged and incentivized by hiring committees rather than discouraged.
Until opinion leaders in the scientific community come out and acknowledge that the way forward is through Open Data and Open Publishing on the Web, there will be little change.
The old standby of “this is how it’s always been done” or “the system is not the best, but it works” is no longer acceptable. The most important facet of science, the scientific method, isn’t being applied to how science gets communicated. We should be ashamed of ourselves.
For more on how the Web will advance science, John Wilbanks, VP of Science at Creative Commons, has some good thoughts on Integrate. Annotate. Federate.
Follow Jason Hoyt, PhD on twitter here. Or, If you are new and wondering how Mendeley works.
Mendeley Blog 
Someone needs to reread their Foucault. Knowlege is about trust and that trust is guaranteed by a process, part of which is funded by the fees we pay publishers. There is now a generation of Students who have had their knuckles rapped for citing Wickipedia: we do not need a whole new class of less than reliable sources for which researchers have have to doublecheck and triplecheck before they cite.
Scientists may be honest but many are not: sufficient for publishers to have to forensically examine, for example, astronomy images for the removal (or addition) of inconvenient stars, or DNA images for the readjustment of inconvenient bands. This sort of editing and verification does not come cheap. The presence of an online article from an established journal allows me to assume some degree of provenance. If it is not from an established journal then I have to add a time wasting layer of research of the article’s origins.
I have elsewhere questioned the assumptions of the ubiquity of the internet in an atmosphere where the main resources consumed to feed the infrastructure, are rising rapidly in price; and the assumptions of freedom of political control are risible. We have already seen the influence of the religeous right on a US administration censoring scientific output. Another 8 years of Republican government immediately following the last one would have set science on a backward path: and I am sure internet restrictions were in their purview. I do not trust governments as far as I can throw them.
We are already seeing governments dismantling their social infrastructures in favout of web based administration and we are also seeing the social consequences: witness the dismantling of the UK Post Office. Yet in a real emergency: the miner’s strikes of the Heath Government is an example, the internet might have great difficulty operating effectively. Whatever online solutions are developed; they must be solidly backed by a distribution infrastructure that will continue to work should the internet break down.
Alarmist, maybe, but there are already online journals that are only accessible using specialist computer applications, albeit delivered by web addons, eg. Internet Archaeology. This is a journal I cannot go into a university library and pull off the shelves to read – and this is a discipline that the very low on the pecking order for funding: internet only access to journals may well exclude the thousands of unfunded students in minor disciplines. This at a time Universities are cutting back on their journal subscriptions to meet smaller budgets.
Note second post contains spelling and gramatical corrections
ARG
Hi Alan, thanks – I’ve updated your comment with your corrections.
I have a couple of articles on specific such gripes at mycogia.com under posts.
Knowlege is about trust and that trust is guaranteed by a process, part of which is funded by the fees we pay publishers.
Quite right; the only question is, what’s a fair level for such fees? As things stand the academic community is being roundly gouged. In addition, the subscription model does not scale: no matter how low the fees, no one can subscribe to 25,000 journals (see Ulrich’s directory). Open access is the only way to guarantee end-user access; not only that, but when publishing fees are paid upfront journals become economic substitutes rather than economic complements, and a truly competitive market can be established.
Whatever online solutions are developed; they must be solidly backed by a distribution infrastructure that will continue to work should the internet break down.
I’m not sure I get this. Under what circumstances, in which the internet had “broken down”, would anyone still be trying to access scholarly journals?
This doesn’t apply to scientists in e.g. developing countries, where the internet is unreliable, but offline backups are not the answer to that supply problem either — better net access is.
Hi Jason, nice post, I particularly liked your quote about business models for publishing “I don’t care if you find one or not. I’m here to advance science, not your bottom line” 🙂
This is as somewhat naive view of academic publishing. Many publishers are not-for-profit societies and university presses. They have mission statements, not shareholders. To characterize them as the bad guys is counter-productive and, well, kinda mean.
These publishers have been providing a valued service for decades or even centuries. The advent of new technologies has changed the media, but not the purpose of responsible publishing. It is up to all of us to find the best way forward so that we can preserve the important aspects of scholarly publishing and further science at the same time.
Thanks for highlighting that, Kristen. Certainly, publishing outfits such as HighWire Press, BioMed Central, and PLoS are blazing a new trail in the digital age that others should emulate.
Also, one reader sent me this via Twitter, “Grant funding committees should require at least one or more articles published in an Open Access journal.” Why not extend that to faculty hiring committees as well?
The view in this blog carries certain ideological and partisan flavour. It is obvious from the frustration and anger of the author.
Is it possible to restate the problem in more objective and unbiased manner? At least that would be the first requirement to move towards “application of scientific method to how science is communicated”.
Mansoor, this was a very partisan post, but that was kinda the point, I think. Why carry on having polite and genteel discourse about this when the whole edifice is crumbling around us?
Mr. Gunn, Thanks for the response. My first name is Masroor.
Anyway, my point was not about being polite or rude, and I do not think this blog has any rude contents. Certainly there may be (or may not be) some problem with scientific publishing, but first we need to accurately identify this. Next is to make sure that solution does not turn out to be worse than problem itself.
At the moment we have a system at hand, and any improvement should evolve out of it. I am more concerned about the radical solutions which may make us worse off.
Mr Bangesh asks if it’s possible to restate the problem in a more objective way. The problem is the growing complexity of all relevant and pressing problems, in comparison to a hundred years ago. The solution the original author proposes, i.e., opening and freeing all data about the problem, has already proven to be a possible solution to complex problems many times, so his choice is a natural one. Mr Bangesh and others, however, do not even appreciate that we live in times that have problems that are vastly more complex than the publisher’s solution can handle—even if it could hundred years ago.
Well, this is not entirely true that I fail to appreciate the complexity. Actually it is the complexity which is preventing us from sharply pinpointing the trouble spot as I am insisting.
There are many issues involved, and it may not be harmful to split the discussion into different threads. For example someone can identify problem the business model of scientific publishing. Another area could be the existing peer review model. I would be delighted if someone enlighten this community with their informed view.
Personally I am more concerned because of yet another aspect of complexity. As the number of disciplines has proliferated many fold in last several decades, the size of community dedicated to any particular discipline is dwindling. It turns out that now it is very difficult to differentiate junk published science from the meaningful published science. Specially when you require (for work in your own field) to judge some findings in your neighbouring field, but you do not have time to work out the quality.
Scientific publishers (or new orgs to replace them) need to serve several important functions… publishing isn’t actually one of them.
What we really need are better categorizing and rating aggregators.
There are much better ways to do this than the publisher model. Something more like a list (a bunch of categorized lists actually) where submitted work is reviewed and possibly elevated to higher (more visible) levels and more general interest categories.
Not really a new idea, but there is no good reason to start with a publisher model and trying to shoe-horn in the real vital functions. Article level impact metrics are a good place to start.