What’s wrong with PubMed Central?

There is some modest political debate behind the scenes about whether or not Canada should have the equivalent of PubMed Central, an open access repository for peer-reviewed biomedical science publications. The UK has already done so with UK PubMed Central.

Enter Steven Harnard, whose post to American Scientist Open Access Forum was picked up by the excellent Open Medicine Blog.

Harnard’s essay is less an argument than it is an injunction: Those who fund or host research should require scientists to deposit results in their institutions’ open assess archives. Centralized archiving should be avoided because it confuses researchers and therefore hurts the open access movement.

While I’m on board with the first point, his other claim doesn’t make sense. If there are psychological and administrative barriers to open access archiving, they surely apply to both types of repositories.

More to the point, aside from discouraging researchers to self-archive – a claim for which I can find no supporting evidence – what’s wrong with PubMed Central and similar national repositories? Don’t they do a good job at promoting open access? It is my experience that, at least among younger researchers, resources like PubMed Central create expectations that articles be easily found and openly accessible.

The reason for this is simple. Scientists like PubMed Central because it is the WalMart of medical science. It’s an easy-to-search portal linking to easy-to-access content. Better still, its collection is so comprehensive that it provides a measure of assurance something important hasn’t been missed.

Until there’s an easy way to topically search across distributed depositories, centralization is a good idea. Open access isn’t useful if you can’t easily find the information you want – hence the popularity of Google (which isn’t always the best search tool).

It makes sense that open access should not be voluntary. While researchers love using easily searched and openly accessible archives to fill their bibliographies, they are slower to deposit their work in an open access forum once it is published. If those funding and hosting the research care about its visibility, open access should be a condition of funding.

To make things easier, institutions and academic societies should provide administrative support to make open access publication easier for their researchers. They should also require their researchers to archive their research in the open, wherever an audience might be expected to look for it. This includes:

1. researchers’ own websites
2. universities’ faculty, institute and library websites
3. grant funders’ own repositories
4. government’s central archives and national libraries (e.g. PubMed Central, Library of Congress)
5. topical archives (arXive)
6. academic associations websites

Loading them with RSS feeds would be a bonus.

It is easy to see that the proliferation of repositories has the potential to make things more confusing, and it may especially threaten interdisciplinary ventures when researchers might not be aware of one another. That’s why in Canada and elsewhere, government has a role to play in the creation of centralized archives and indexes which make it easier for information to be found.

While I hope that open access spreads from the medical sciences to other areas of knowledge – including the humanities – I doubt Canada will now take up the challenge. The nation seems to be against access to knowledge both inside and outside its borders.

Postscript:

I received an e-mail asking to clarify some terminology.

PubMed is a web-based search interface allowing open access to citations and abstracts in the MEDLINE database (among others). It is searched using the Entrez search engine.

PubMed Central, on the other hand, is an open access repository of peer-reviewed articles.