notes on open scientific publishing

nice article profiling plos & issues relating to public access to publicly-funded research.

“Publisher for the People” by Will Harper, East Bay Express, 2004-09-29

Marc Brodsky, with Association of American Publishers, disputes the virtually indisputable benefits of open publishing of publicly funded research:

Brodsky points out that the government subsidizes other commercial activities where the public is charged for something value-added. “For example,” he says, “when we get medical advances that come out of NIH-funded research, we have to pay for the drugs.”

Hmm. Aside from the retort which springs promptly to my lips — just because we can double-charge the public for research doesn’t mean we should — a few other comments come to mind.

One minor point, and then on to a rant: Brodsky suggests that public funding doesn’t mean that consumer/citizens don’t have to pay in the end. But it’s not like folks won’t be paying for PLoS Biology. It’s just that, as a public-interest organization, the group is trying to peg costs to cost-recovery that relates to its mission — publicizing research. On the other hand, Elsevier, Wiley, etc., are also pegging costs to their mission — but their mission, as for-profit corporate entities, has nothing to do with publicizing research; their mission is making a profit. Isn’t that obvious? Publicizing research — selling the research others have done, rather — is merely the means to the end of making a profit. Subscribers to PLoS Biology are starting out at about $120 per year for the print version. That’s a start at figuring out how to recoup the costs for the print delivery.

Now, the rant.

The process to take basic research and bring a drug to market and into the hands of consumers is a far cry from the process for publishing that basic research.

The pharmaceutical companies have been fond of reminding those who complain about drug costs that when we pay for drugs we’re paying not for the basic research (that was paid for by the NIH). Instead, the drug costs, and the patent protection which enables pharmaceutical companies to maintain artificially high drug costs, are intended to compensate (plus) for the costs of converting basic research into drugs: converting basic research into potential drugs; developing the drugs; researching the drug delivery questions; clinical trials; production costs; advertising & marketing. Many, many potential drugs fall by the wayside somewhere in that process. So companies want a way to recoup their sunk costs, and pay for the failures.

Parts of this system aren’t as efficient as they could be — I immediately think of advertising which is now a major part of the pharmaceutical business. By turning pharmaceuticals into a business, and sinking dollars into advertising to artificially create demand for drugs, entire new levels of inefficiency have been larded into the system.

But while I have my beefs with the pharmaceutical industry-as-business, and while I may question or occasionally suspect that their drug-development costs are inflated, there is no question that there are in fact significant costs which the pharmaceutical industry bears. An incentive structure like patents makes sense in the pharmaceutical industry, even if you don’t ultimately agree that it is the best way to do incentivize the creation of drugs.

Scientific publishing is a whole different ball of wax. The product is very close to the basic research itself (not to minimize the work involved in writing a paper). Many — most — of the refinements from the basic research to the finished product are performed gratis by the same basic research industry. The researcher does the research. The researcher writes the research. Other researchers (“peers”) review the resesarch, comment on it, make suggestions for how it should be shaped, recommend its appropriateness for the journal (subject matter, significance, novelty), screen it for plagiarism and point out prior art that should be cited.

The editorial boards of the journals that decide the overall practice and direction of the journal are also comprised of the research community.

The journal itself provides coordination of the submission and review process; and then gets the work printed and disseminated. Having worked on a newspaper and a journal, I can attest that this is not trivial. However, almost all of the print-and-disseminate portion of this work could be eliminated by electronic dissemination.

An additional piece of the dissemination puzzle — archiving — is done poorly by the present system. Journals provide access to their own archives but not in a unified system. In an attempt to keep a lock on the whole process, in fact, journals restrict access to their archives. charging libraries and database companies exorbitant rates to provide access to even indexes of the contents and even more for full text. This isn’t even a souped-up, XMLL’d version of the full text — at best, journals typically provide text-searchable PDFs of the text, and a searchable database with basic field definitions of the set of articles. Libraries and archives provide the permanent long-term storage and access of all this research, at their own considerable expense. The journals are typically not even provided in a form easily amenable to long-term storage; libraries uniformly send journals out to be bound in hardcover for long-term storage. And the rights that libraries and library users had in a print world have actually diminished in an electronic world: now that full-text data is online, libraries have to pay to maintain access to the archives — even after they have already paid once. So libraries still bear the cost of the archive, but now it’s not a purchase model, but rather on an infinite mortgage model.

So the print-and-dissemination functions of journals are being rendered obsolete by the electronic world. And the archiving and long-term access functions, which have typically been provided by libraries, are now being hindered by journals. That leaves the coordination of the submission and review process as the sole necessary function that journals perform.

Coordination of the submission and review process is, indeed, a useful value-added service. But to compare it to the drug development process isn’t just comparing apples and oranges; it’s more like comparing apples and, say, the production of spider silk from genetically-modified goatmilk.