This is a response to Peter Suber's article "Removing the Barriers to Research: An Introduction to Open Access for Librarians". In that article, recently published by BioMed Central, Peter advocates self-archiving and open access journals as a means of overcoming the pricing and permission crises that limit access to scientific literature and stifle the progress of academic research.
The main current proposal for increasing the availability of scientific literature is self-archiving: this means of distribution is not intended to avoid peer review, as only papers that have been quality-controlled will be stored in the archives. Peer review is organised by journals, so self-archived papers are all published in journals as well. The traditional journal publishers who charge for access to papers are obviously unwelcoming to the idea of their premium content being made available for free, so prevent self-archiving from happening on a wide scale by retaining the copyright for the papers they publish.
With open access journals, such as those published by BioMed Central (eg the Journal of Biology), the authors retain the copyright for their peer-reviewed papers. The authors are therefore free to distribute this work however they choose, and no charge is made for access. This is possible because distribution and management fees are paid by the authors' institutions.
At the moment many open access journals and archives, due to their novelty, are still lacking the following:
- Wide readership. Researchers need to publish in a proven, long-standing brand-name journal that covers a powerful subject area. eg J Exp Med, J Biol Chem, J Immunol. This should preferably be a journal that's guaranteed to stay around for a long time (ie something like 20 years). Papers should also be indexed in PubMed and other citation databases.
- Technically advanced storage. Not just a PDF preprint the author ran off from Word or LaTeX, but a professionally typeset document including graphics. If it's a PDF, it should have all the added extras, such as hyperlinked references. Machine-readable files (eg XML) should also be available.
- Solid, permanently available distribution (this is one reason I wouldn't trust many universities and smaller institutions to run these archives). It would be ideal if the archive could have dedicated, government-funded taxpayer-accountable hosting.
There is an alternative, illustrated by an analogy for my problem with open access journals: If you've painted a masterpiece (or even a mediocre still life), it costs you nothing to take the painting out to the middle of the local park and leave it there in the rain for everyone to see. On the other hand, you'd much rather have it hanging in a prestigious gallery, assuming that it was good enough to be accepted.
This suggests that while the minimal cost of distribution, ensuring open access, can be met by institutions and archives, a strong commercial opportunity would exist for publishers (or rather, aggregators) to carry out all the other additional-value services. This includes organising peer review, selecting the best papers, compiling topic directories, and building reputation and prestige by virtue of informed selection (see Faculty of 1000 as an example: this subscription service would work much better if the papers they selected were accessible for anyone to read). When anyone can publish, the competition will be to filter and select the best papers fastest, which can only be to the advantage of the end user.
The theme remains: facts should be freely available, but commentary and selection are the added value of a subscription. The current 'publishers' may have to compromise and be able to adapt to this alternative revenue stream, but the benefits of open access to publicly-funded research material are too numerous for scholars and librarians to avoid this opportunity.