Monday, April 16, 2012

"Delayed open access"

I've had a few questions from folks who ask how I feel, and how I interpret the ResearchWithoutWalls pledge, in the context of publishers who exploit private rights to publications for some period and open access to them after some period of delay.
My personal position on the matter is based on the principle that research publications that are produced using donations, such as funds donated involuntarily through our taxes or time we donate voluntarily when reviewing, should not be turned into private property.
Strategically, limiting the period of time that a paper is kept closed-access turns the question of open access from one of principles ("is charging someone to access work they voluntarily reviewed appropriate?") into a negotiation ("how long is it appropriate for me to try to monetize donated work?").  I would argue that if publishers are successful in shifting the open-access debate from a question of principle into a question of how much they should be allowed to exploit ownership of publications, the research community will lose.

You see, once they get us to agree on an appropriate amount of time to exploit private ownership, we are explicitly indicating that we believe that taking and exploiting private ownership of donated work is appropriate.  Each one of us may have a different sense of how much value a publisher deserves to profit from private ownership rights, and so publishers will be able to take back more and more rights over time.

Consider, for example, if we went to a model where we agreed to settle for a model in which all publishers could keep exclusive rights to our work for twelve months.  It's unlikely that anyone would express outrage if a publisher (especially a membership-based organization like ACM or IEEE) cited tough economic times and extended private ownership to 18 months.  After all, we'd already conceded that private ownership is appropriate for some period of time.  In increasing that time period, they can argue either that they understand their publishing costs better than we do or, in the case of a membership-based organization, that the funds are for a good cause that only stingy/selfish members of the community would want to take money away from.  A few years later they might ask for two years, and so on.  Once we've said that it's morally appropriate to exploit private ownership, and they assert that a small increase in the private ownership period is necessary, it is unlikely that anyone will feel moral outrage or sense that the short increase is worth fighting against.  That will make it even easier the next time the period is increased.  The equilibrium at which we will find ourselves is the one we've been working so hard to escape.

Tuesday, October 25, 2011

One week in, the ResearchWithoutWalls pledge has 265 signatories: 216 with confirmed emails and 49 who had not clicked on their email confirmation links.  These signatories came from more than 120 domains, the majority of which were universities: 30 under the .edu top level domain, another 28 in an academic domain under a country domain (e.g. or, and others directly under top-level domains for countries (e.g.

Of those 265 signatories, 87 (33%) listed affiliations with conference or journals.  A total of 196 affiliations were listed, 171 (87%) of which were for conferences and journals within the field of Computer Science.  The subfield most represented in Computer Science was security (87 affiliations listed), with a trickle coming in from AI (16), HCI (9), machine learning (9), and systems (8).  The predominance of security suggests that the pledge spreads more effectively person-to-person (both in-person and through social networking sites), or when potential signatories see that people they know have joined, rather than as a result of more general publicity forums such as Slashdot and Digg--not that we didn't appreciate the coverage.

Our most prolific advocate may well have been Matt Welsh of Google.  Thanks in part to his Google+ post supporting the pledge, there are 46 signatories from Google, 16 of whom listed peer reviewing affiliations (roughly the same fraction as the general population of signatories).  Our most prolific peer reviewer is Roger Dingledine, director of the Tor project, who reported peer reviewing for 15 different publications.

Our future success will depend on the grassroots efforts of our signatories to spread the word to colleagues at their institutions and within their research communities.  We will need to dispel common misconceptions about the pledge: that signatories are being asked to shirk work they've already committed to, that signing prevents professors from sending their students' work to restricted-access venues, or that organizations like the IEEE and ACM cannot survive without revenue obtained from selling access to work created at others' expense.  Potential signatories who may be uncomfortable with a particular feature of the pledge wording may need to be reminded that they can rewrite the pledge to meet their personal level of commitment to open access.

While we have a long way to go, there is promising momentum both from our effort and the work of others.  We started the pledge on the first day of the ACM Conference on Computer and Communications Security (CCS).  There are now twelve signatories who ACM CCS has recently relied upon for peer reviews, and our signatories now include one of the CCS 2012 program chairs (his pledge becomes effective at the end of his term).  Ten signatories come from the pool researchers relied upon by the IEEE Symposium on Security and Privacy for peer reviewing, and they include a former program chair.

Other Computer Science researchers (independent of the ResearchWithoutWalls effort) have been tremendously successful in raising awareness in advancing the rights of researchers in our community.  Just yesterday, at the ACM SIGOPS business meeting at the Symposium on Operating Systems Principles (SOSP), Michael Freedman called a straw poll to gauge the community's support for changing the publication policy to allow authors to retain copyright to published research.  The overwhelming majority of attendees supported the measure.

Thanks to the efforts of Michael Freedman, Dan Wallach (who held a similar straw poll at IEEE Security and Privacy), and others, those of us in the research community are recognizing that we share a community value: the belief that conferences and journals should not have ownership rights over the authors who actually performed and wrote up the research to be published.  Please keep working to spread the word so that the the organizations chartered to represent our research community will adopt publishing models that are compatible with the values of the community.

Thank you all again for your support.