Open access vs. peer review

I wanted to blog about open access issues: I never quite understood what all the hooplah is about - arxiv existed for well over decade, but as any CM physicist can tell you, arxiving a paper doesn't mean much until it went through peer review and got published in at least somewhat respectable journal. To me, peer review combined with some selectivity (not all papers are accepted, even if the work done is correct) is crucial - both in terms of evaluating quality of my own work, but also in trying to keep up with the literature (reading a couple of papers a week from Nature, Science and PRL is a lot easier than combing through hundreds in arxiv or conference proceedings).

But since the peer review process requires participation from editors, they have to be paid somehow - and subscription fees seem as good way as any to do that.

Which brings an interesting questions - how come reviewers don't get paid? Even some nominal amount may be sufficient. Otherwise - and I am not quite sure about how the financial side of this balances out - but it's hard for me to imagine that companies like Elsevier are struggling to break even. Just like with record companies, subscription fees and color print charges (CD sales) don't go back to support hard-working referees and editors (musicians), instead making record executives, err... I mean scientific magazine publishers rich.

Arxiv is a little like napster - a nice concept in theory, but if music (scientific papers) are free, how do we expect musicians (editors/referres) to make an honest living doing their job?

This post was brought in by recent washington post article that argues that all/most papers should be free since taxpayers paid for research to begin with.

I don't have as much problems with this argument, but I do have a lot of problems with the proposed solution - requiring researchers who use taxpayers funds to publish only in open access (=unrefereed) publications means diluting good research with the bad.

A better solution would be to require for the grants to pay publication fees associated with the papers coming out of the research supported by the grant. There obviously have to be an honest cap on the paid amount per paper, otherwise the private journals can arbitrarily jack up the prices. For example, I would gladly pay for editorial (and maybe referee?) contribution, calculated by the amount of time they invested, plus some minimal publication-related charges.

Basically, the model I want is a free, non-for-profit, online journal sort of like elitist section of arxiv, which features a limited number of papers that are carefully selected by editorial staff, and peer-reviewed - perhaps accompanied by a short summary of what was accomplished and why it's an important paper, in broader picture. Because the whole journal is online-only, publication costs are minimal. Costs of maintaining minimal editorial staff that selects the papers and coordinates review process, plus costs of refereeing should be ideally picked up by some government agency. Even though it's not a big deal to ask authors to pay for their papers, I always felt it's backwards - you don't ask for writers to pay additional fees to publish their work, or musicians to pay so that they can perform to a sold-out audience, or professors to pay for a privelege to teach students. So why should scientists pay for having their work known to the world? It's bizarre if you think about it, even though we are all used to it by now.