Wednesday, December 9, 2009

Quickie: Let it be Free

I was thinking about how the "scientists" in the Climategate scandal are hiding their raw data. If the science is really objective, then by its own merits other researchers will reach the same conclusions.

If you're afraid that they won't, then you're hiding something.

As such, no scientific researcher should ever hide their data. Ever! Not only that, but their work should be out there too -- formulas, theories, methodologies, everything -- to be judged not only by their peers, but also by the public at large. If science is a quest for the truth, then the truth has nothing to fear from inspection.

But I realize that it is not just our disgraced climate researchers who have adopted policies of secrecy. You can't, for example, get most of the academic work in Computer Science unless you join the ACM. In fact you can't get most works in academia unless you join some tight-nit clique. It's all closely guarded.

Long ago, I could understand how these organizations needed to charge money in order to recoup the costs of publishing on paper. It makes sense that they shouldn't take a loss. But amazingly that changed in the Information Age, and instead of offering free downloads it appears that most organizations choose to monetize their IP (business speak for "go for the profits").

You could make some well thought-out arguments about how the funds collected go back into the organizations to help spread their good works, but oddly one can easily suspect that most of the income is really used to protect the revenue generation (business speak for "stuff that causes profits") and pay salaries.

Academia -- data, papers, methodology, everything -- should be free and done out in the public. It should be open, and visible. Approachable. Companies hide their research because they are locked in competitive battles. Universities (and other research institutes) are not companies.

Research is not competition, it should be co-operative. And it shouldn't matter who contributes (particularly when we live in a day and age were so many people are both eager to help, but also have the available time).

I should never be told that I can't download a paper unless I pay a big sum of money. It should be free.

2 comments:

  1. Paul, you oversimplify a little here. Scientific publishing has never been about the problem of printing the papers, but about the filtering of them.

    The peer review process is supposed to work thus: Scientist A submits her paper, and B, C and D review it. B thinks it's OK, but B and D have a problem with the methodology, which was recently shown to be inaccurate. So the paper is rejected, pending editing. The process requires the time of B, C and D, even though the paper was ultimately rejected.

    You can see how the work would mount up, particularly as the number of rejections per accepted paper increases.

    On the other side of the coin, a journal could publish N papers in each volume. So if M papers are submitted, there is some acceptance ratio r=N/M. The point of getting a paper accepted is broadly to prove its validity, and the lower the acceptance ratio of the journal, the more prestige it has.

    The end result, as we see in the scientific communities today, is that there is a rough hierarchy or ranking of journals, from the oldest and most prestigious (Nature et al) to the niche, anything-goes journals.

    While the process has its problems - cost, unwillingness of scientists to bother reviewing, limited level of reviewing - it is certainly not about restricting access to information. I would hardly class a group including every research university in the world, and many R&D departments as a clique.

    I don't believe scientific publishing can or should just carry on as before, because a more Web2.0-style system with open feedback could easily replace it. And, of course, the old journals will have difficulty adapting to that change, as all institutions do.

    But I would say that scientific publishing, and the peer review process, has long been a part of open and meritocratic science, making as much information available as possible whilst remaining fiercely competitive.

    This is not the first science scandal, and it won't be the last. But if your work has any importance, your competitors will attack any weakness they can find. Every scientist wants their results to match their private (or public) hypotheses, so there is always bias in selection. But if you outright lie about anything important, you will be found out.

    I would say that the story of science is broadly one of "the truth will out", and the level of scandal about a bit of lying is a good indicator that the spirit of truth lives on.

    ReplyDelete
  2. Hi Phil,

    Thanks for commenting. I do realize that there is a lot of work going on in the peer review and the editing process. I really won't want to deny that any funding (do reviewers get paid?) or limit the number of journals. The more options someone has to publish, the better it is.

    Still, most of the money for this likely comes from the government, and the researchers themselves. I'm sure the public at large doesn't actually contribute a significant portion of the funds.

    I'd just like to see that the 'final' stage of the process is that the papers and data (all of it) are made public, and that (like openSource) there is always a way that outside people can get to it. While you're working on it, it should be private, but once it is published, then it should be freely accessible.

    I don't know how it works on the inside, but the public perception of most academic disciplines is that they are very tightly nit groups (one or two, and a few stragglers). That might just be a stereotype.

    Personally I'd say the history of Science is all about various different groups fighting to control the flow of information. Information is power. :-)

    Paul.

    ReplyDelete

Thanks for the Feedback!