Patent reform, publication and repositories

Patent reform has been percolating in Congress for quite a few years now, and I have to admit to being caught off-guard when I saw the announcement that a comprehensive reform package had passed in the U.S. Senate by an overwhelming majority.  This story about the bill (which has not been passed in the House) set off an animated discussion between David Hansen, the intern in Duke’s Scholarly Communications office this year, and me regarding the issue the article raises about creating a rush to publication and whether there was a potential impact, if the law changes in this way, on disciplinary and institutional repositories.

This post is necessarily tentative, since neither David nor I are patent experts by any means.  Any reader who can correct or clarify our tentative conclusions is welcome to do so.  But based on our initial discussions, I think we are in fact likely to see more pressure to publish quickly and that that pressure could give some repositories a more prominent role in communicating scientific research.

Two facts about the patent system and the bill in Congress are relevant here.

First, under both the current system and the proposed new one, an invention must be “novel” to receive a patent.  To show that an invention is novel, patent applicants provide a list of “prior art” that shows what their invention is based on and establishes that it represents some new idea or “creative spark.”  If the prior art anticipates the new invention too completely, such that the new twist seems obvious, that will defeat the application for a patent.  An inventor herself, however, is allowed to publish a report of her new discovery, usually in a journal article, and for a period of 1 year that publication will not count as prior art such as to render the invention non-novel. Essentially, the inventor has a one year window to file a patent application after publication of the information about her own discovery.

Second, the new system embodied in the Senate bill would change the priority for who gets the patent in a particular invention when there are rival claimants.  The current system in the U.S. awards the patent to the first person to invent the object of the application, regardless of who filed the first application.  This seems fair, of course, but it results in significant problems of proof and makes the patent application system lengthy and expensive.  The new bill would adopt the system used in most of the rest of the world, where the patent goes to the first person to file an application.

When we understand these two facts, it is easy to see why this proposed change could lead researchers to want to publish their results even more quickly.  In the past, delay did not matter as much because if I was the first to invent, I would get the patent in preference to someone who filed before me.  Under the new system, an earlier filer could defeat my patent.  If I published my results in an article, however, that article can serve as prior art that would defeat the other claimant’s application.  It would not be prior art for the purpose of my application, however, as long as I filed within one year.  Thus there could be pressure to publish articles “defensively,” to undermine any applications that beat my own to the Patent Office mailbox.

Also, because of the way academic rewards are structured, academic inventors often want to publish articles about their research even before the invention is finalized in a way that justifies a patent application.  That one year window for inventor-authors has served this perceived need to get a peer-reviewed, tenure-supporting article into print even before the application was filed.  Under a “first to file” system, however, the whole process may get telescoped.  Since the filing date would matter, researchers might want to publish more quickly and file more quickly in order, again, to defend against another claimant who might also be planning to file.

Assuming there is this added pressure to publish quickly when a patent application is in the offing, disciplinary and institutional repositories may have an important role to play.  Researchers already complain that the process of publishing in a journal takes too long, so that publications are really just a formal record of research that is often 6 months to a year out-of-date.  The added pressure of defending against a rival patent application would seem to make this delay even less acceptable.  Pre-print repositories may be the solution, where a record of research can be “published” such as to serve as prior art to defeat anyone else’s patent application while waiting for the formal process of journal publication to proceed.  Since even reports in a single printed dissertation have been held to be prior art for this purpose, I have no doubt that a pre-print in ArXive or some other disciplinary repository would also serve the purpose, as would institutionally-managed repositories.

If the patent law changes and researchers really do start to feel this added pressure, librarians may serve an important role in directing them to appropriate institutional or disciplinary repositories where their pre-prints can “hit the streets” as quickly as possible.  And repository managers will need to be sure that they can turn these deposits around in a way that helps our researchers protect their rights.