This week, the Research Works Act in the United States was defeated when the main sponsor, publisher Elsevier withdrew their support in the face of strong criticism and boycotts from many scientists, particularly some eminent mathematicians. The act would have been another gagging of sorts of scientists, overturning a recent mandate by the National Institutes of Health in which publicly funded research has to be published in open access journals (those that do not charge to view the content). It got me thinking about journals in general and the future of scientific publishing.
First, let’s explain how journals work for the majority of the sciences. Journals are the periodical magazines in which scientific papers are published. Where our science meets the light of day, where it can be discussed and recognised, the implications of our work can be realised and where progress can be made. Our years of hard work are recorded in a few short pages, sitting there in the almost public domain for perpetuity, to stand the test of time. Yet the system is broken, outdated and downright unfair.
Journals make huge amounts of profit from science. Firstly, scientists pay to submit and publish their work to one of the journals. Secondly, the journal has very few staff because the peer-review system means that other scientists review the work for free, while scientists want to be editors as its seen as a very good part of an academic’s CV. Finally, the journals cost money to read, with different titles sold in bundles so that institutions like university libraries have to pay for journals they don’t want or need. Its a very profitable business model.
In addition the whole process of publishing, from submitting to actual publication takes at least six months, often up to a year. This is after many years of research to begin with. In this electronic age, the speed at which journals get from scientist to publication is pretty comparable to the 19th century.
When the National Institutes of Health in the US mandated that taxpayer funded research should be published in what are known as open-access journals, i.e. those that the public can view without paying a fee, you would have thought some sanity had been brought to the world. But not, as it turns out for long.
This policy obviously upset big publishers like Elsevier who decided to take action like only big multinational companies know how, by funding a bill through the United States House of Representatives known as the Research Works Act. It was basically nothing more than an attempt to protect the bottom line of the big scientific publishers (Confusingly the main proponent of the bill was a northern Californian Senator who was one of the main opponents of the SOPA and PIPA bills).
There was uproar, as you might expect, and a petition and a threat of a boycott of Elsevier’s products by scientists, particularly some prominent mathematicans. Elsevier have thankfully seen some sense, and withdrawn their funding from the bill, essentially killing it stone dead. Hooray for some sense in the world.
Not that scientific journals are necessarily evil, they’ve done a lot of good in the scientific world. In particular they safeguard the peer-review process, a central tennet of the scientific method. Sadly, like most big companies, its all about as much profit as possible, and in a world that is needing publishing companies less and less, its hard to see a future for a system that is too slow and far too over-priced.
So where do we go from here, and what is the future for journals in this online, instant access world? Physics and mathematics have started going down a route involving a site called arxiv (pronounced archive because those crazy physicists do love their greek letters). Here, results are published prior to publication, as it were, for discussion, and to circumvent the long waiting. The paper does eventually make it to a journal, through full peer review, but this system allows scientists to see the new breakthroughs as soon as they happen, not a year later, a much better result for speeding up the rate of scientific progress.
Some care of course, has to be taken, as this approach somewhat diminishes the influence of peer-review. The odd crazy theory might pop up from time to time, but it will soon be put it its place.
Such an approach would probably not work for the fields of Medicine or Climate Change for example. In such media scrutinised disciplines, where any proclamation can make headline news, allowing peer-review to be bypassed would be dangerous. It would immediately be exploited by crack-pot ideas and big business would be putting in papers on the wonders of tobacco, new drugs and putting carbon in the atmosphere faster than you could say “but seatbelts save lives”.
Many community led journals are being set-up as alternatives, especially as there is no longer any need for physical copies of any journals as we all read journals as pdfs of our computers anyway. And if you’re going to do a job for free, you might as well not charge anyone. But these journals need time to gain traction, a reputation, and an all important impact factor (kind of like a league table for journals). I personally think this idea is a great one and I hope it works out.
One journal, PLoS, charges loads more than a typical journal for scientists to publish, but because it is then open-access, free to general public, it is hugely popular. Particularly in the field of Medicine where getting information out, especially to the third world where journal subscriptions may not be affordable, even for industry professionals, will save lives.
Perhaps the most radical idea I’ve heard is from the computer science industry. Amazingly computer science is currently one of the most backwards of scientific disciplines when it comes to publishing. Most papers are published through conference proceedings (a big book of everything presented at a conference) rather than a journal. These conferences pride themselves on their high rejection rate, often accepting only 10-15% of submissions. It gives them the air of exclusivity and makes publication far more sought-after. After-all, the more papers a conference turns down, the higher their reputation becomes. Conferences actually compete with each other to have the highest rejection rate.
All this does is to preserve the status quo, to allow progress only by small obvious increments. Radical ideas don’t get in, no matter how good they are. Countless good papers, even exceptional ones make no progress for years. A good example is one of the most cited papers in engineering (let alone computer science) in 2005 was held up in this system for 7 years, only seeing the light of day in 2004 when the author, David Lowe, went to a lower quality journal. It was on the SIFT method of detecting features in images and has now become the industry standard in computer vision.
So what some computer scientists are proposing is this: http://yann.lecun.com/ex/pamphlets/publishing-models.html. In summary it proposes making papers available for comment immediately. Comments are made, revisions are made, and the whole process is transparent from start to finish. And here’s the crux, the comments get rated too, so that the good, constructive, informed, comments are given a more positive weighting. Good reviewers would get good reputations and their thoughts would become sought after. This system would have a lot of withdrawals and retractions, a lot of crack-pot ideas would have to be discussed, but the idea of this open forum is something that is quite exciting. Its amazing, ambitious, and rather crazy.
Personally I think it’s a little too much, I prefer the idea of more community led journals (though these would not be without their problems such as egos). It would never work, for the reasons mentioned above, in the fields of Medicine or Climate Change. But just perhaps, there’s a little kernel in there, a brilliant idea, that could change the way science is published.