Tuesday, 20 December 2011

Intellectual-property rights: academic papers

I wrote earlier that I'm opposed to intellectual property rights wherever plausible alternatives exits.  I'll start my review of the alternatives with perhaps the easiest case: academic papers.

My relevant experience is all in science and finance: conceivably things work differently in the humanities.  In the fields I know about, which are overwhelmingly the ones that matter to the vast majority of the population, the way that journals work is that authors submit papers for publication, the editor asks experts in the field to review the paper, the reviewers make recommendations for changes and for or against publication, the authors are invited to make any changes the editor thinks advisable, and the editor eventually publishes the paper or rejects it.  The journal, which has contributed the least to this process, ends up with the copyright to the paper: the author and reviewers work for notional fees or none.

Copyright is therefore no incentive for the production of academic papers.  Its only function is to provide incentives for journals to carry out a filtering process by which readers get an indication of which papers are worth reading, and funding bodies get an indication of whose research is worth funding.  Since there are too many papers to read, and too many researchers to fund, both these filters are valuable.

In practice, authors often circulate their papers to peers before submitting them for publication, both as a courtesy to anyone whose work they cite, and in the hope of getting helpful feedback.  Also,they often  make a version of the paper freely available on their personal websites - this is worth knowing if a paper you want to consult is hidden behind an expensive paywall.  I suppose that journals disapprove of this practice, but think it prudent not to draw attention to it by objecting publicly.

The alternative is simple: authors, as they do now, should consult whomsoever they wish until they think their paper ready for general release.  Then they should publish them on websites dedicated to the purpose.  arXiv does this already for some of the geekier fields.  Here's an outline of how it works, and here are some comments by its founder on its implications for academic publishing.  Here are some thoughts on its disadvantages: none of them seem to me to be fundamental to the question of copyright.  Interestingly, there is no suggestion that prestigious journals in Physics have been unable to operate without exclusive publication rights.

It may be that copyright restrictions are necessary in most cases to make it worthwhile to operate pre-publication peer review: here are some comments by Richard Smith, former editor of the BMJ, on how small a loss it would be to do without.

I submit that medical research in particular would benefit from free publication along the lines of arXiv.  That would get results out quicker, make them easier to consult online, and encourage publication of negative trial results.

If filtering mechanisms are required, something along the lines of Amazon's book-review system would be possible.  The user should have the option to apply weightings to the reviewers, favouring for example ones with high academic titles, or ones whose views, positive and negative, he shares regarding other specified papers.

Let's abolish copyright on academic papers now.  I predict that a few prestigious journals will survive, and the rest will be more than adequately replaced by free on-line publishing.


  1. You've missed open-access publication (e.g. http://www.egu.eu/publications/statement/online-open-access-publishing.html) which solves some of the problems.

  2. One of the problems I have with the existing peer-review process is that although there is review at the time of publication, years later papers which have been proved wrong are still being cited.

    As a researcher, there is no easy way to see the arguments against a given paper, so it's hard to know if a theory has been superseded and so on. The great saviour of all this is a review paper, though they aren't nearly as common as they should be.

    A site a bit like Stack Overflow, where your peers can comment, would be interesting; there is an interesting network effect if you analyse the reputations of the users of such a site - a higher reputation will induce more positive replies, even ignoring the quality of the content.

    One option could be to hide the reputation of a paper's authors until after some threshold is reached.

    I suspect the opportunity to denounce one's rivals and show you know better is enough inducement to peer review, although some catty behaviour may need moderation.