There seems to be quite a bit of discussion about the newly announced ACS Publication Agreement policy in the last week. Thanks to all who responded and clarified on the some of the points I raised: Teri Vogel, on the NIH Submission policy information, and Rich Apodaca, who gave 2009 ACS data on income from subscriptions and licenses versus member dues. Also thanks to Peter Murray for providing the ACS policy requiring ACS-endorsed programs to subscribe to Chemical Abstracts (or Scifinder Scholar) online. Overall there seems to be consensus online that the new agreement is an improvement over the previous Copyright Status form, but could go further in supporting author rights, data management issues, and funding mandates. I agree but have some thoughts of my own.
One curious idea is the situation that allows authors to post in institutional repositories with OA policy mandates, but not for non-OA policy institutions. It seems strange to limit posting in this way, especially when there are subject repositories that can post work as well. One of the problems with Chemistry is that there are few subject repository options available. PubMedCentral, Nature Preceedings, and possibly arXiv.org are the main options now, all of which are still considered prior publication with the ACS. Considering the number of places with OA agreements (less than 100 in the US), this new policy would have limited impact on increasing access to the literature.
Another thought is how having an OA mandate may give a chemist an advantage over a colleague at an institution without one. Is this a big advantage? It's hard to say in Chemistry since there are fewer channels available for posting preprints. I will day, given the interest in new free tools like ChemSpider within the chemistry community, there is a restless, increasingly vocal group that will demand more free content and easier access to the literature. I predict the immediate impact might be small but the OA advantage for authors will increase dramatically once these researchers discover and share pathways to this repository content. Perhaps this is what the ACS is afraid of - distribution is almost always more profitable than the initial publication, especially when you can repackage content and resell it.
To me the most interesting part of the agreement is what is now considered OK to reproduce: supplemental materials and possibly data. While this is a positive step, I still thinik their position on data-sharing is not that clear, especially since scientists are certain to have questions in creating Data Management Plans for NSF projects. All publishers (not just the ACS) will need to consider how they will support this within their existing infrastructure and/or provide policies on supporting and sharing data created from publications. Personally I think the ACS isn't sure how to handle this new NSF policy, and while that's OK for now, come January 2011 there will be questions coming from scientists. It begs the question that perhaps the ACS is deliberately framing the conversation to their benefit. I think it's a little too soon to tell.
To me the dealbreaker, and most distressing item of all, is the requirement that authors relinquish copyright for their final manuscript. This is a scholarly society, one of the largest professional organizations in the world, saying to their members (and non-members that choose to publish in their journals) that they can't own their work once they publish it with them. No posting in repositories, except in limited cases and when the ACS can get additional nomey from the author (above and beyond subscription money they already get from libraries and individuals) and throwing a carrot to authors that they can use a table or supplemental data. It makes true innovation and sharing impossible, as the final version is locked up and controlled by the ACS and silences the voices and concerns of individual authors.
This is especially distressing when other areas of the sciences are making very different decisions. In Physics (the AIP) recently decided to make their journal articles freely available to public libraries. Mathematics has already had prominent prize winners refuse to publish in traditional journals because the work was already available in pre-print form. SCOAP3, from the high-energy physics community, is creating an OA model that will make literature from core journals available to all. Even my own profession, librarianship, has made progress in establishing author rights for individual articles. I was able to retain copyright on a co-authored article in College & Research Libraries earlier this year. If these groups can do this so can the ACS.
My final verdict: there's a little bit of good but still a lot potential bad in this agreement. Let's hope the ACS membership and governance will listen to its members and the broader scientific community to realize this position is detrimental to the long-term sharing of information and advancement of science.